ISPOR 4th Asia-Pacific Conference Research Abstracts

9 downloads 32945 Views 687KB Size Report
Both cost and health outcomes were discounted at 3%. RESULTS: The ...... domains. The discriminant validity between the two questionnaires was acceptable.
Volume 13 • Number 7 • 2010 VA L U E I N H E A LT H

ISPOR 4th Asia-Pacific Conference Research Abstracts

COST-EFFECTIVENESS STUDIES CE1 COST-EFFECTIVENESS ANALYSIS OF GLUCOSAMINE SULPHATE FOR THE TREATMENT OF OSTEOARTHRITIS IN THAILAND Chaiyakunapruk N1, Saokaew S2, Pansang S1 1 Naresuan University, Muang, Phitsanulok, Thailand; 2Naresuan University Phayao, Muang, Phayao, Thailand OBJECTIVES: Recent evidences have shown that glucosamine sulphate (GS) is a potentially effective treatment for osteoarthritis; however, it is unclear whether its use is cost-effective from the perspective of Thai Ministry of Public Health (MOPH). This study aims to estimate the cost-effectiveness of glucosamine sulphate (GS) compared with current care (CC) in osteoarthritis (OA) patients in Thailand. METHODS: A Markov model was used to simulate 10,000 hypothetical OA cohorts. The model comprised four health states including OA without total knee replacement (TKR), OA with TKR, OA after TKR, and death. Transition probabilities and health state utilities were obtained from published literature. Drug cost was obtained from the Drugs and Medical Supplies Information Center, MOPH, while the cost of TKR was based on the reimbursement rate provided by National Health Security Office. The model estimated quality-adjusted life-years (QALYs), costs and incremental cost-effectiveness ratio (Cost/QALY gained) over life time horizon. Probabilistic sensitivity analyses (PSA) were undertaken. Both cost and health outcomes were discounted at 3%. RESULTS: The estimated QALYs were 17.88 and 17.79 QALYs for GS and CC, respectively. The lifetime direct medical costs for GS (generic brand), GS (original brand) and CC were 123,679 THB, 279,988 THB, and 57,863 THB, respectively. Compared to CC, the incremental cost per QALYs gained was 774,125 THB/QALY for GS generic brand and 2,612,605 THB/QALY for GS original brand. Based on a cost-effectiveness cut-off of 100,000 and 300,000 THB/QALY, the probability that GS (generic brand) being cost-effective was 38.6% and 50.0%, respectively. Using GS generic brand would be cost-effective if the GS price is lowered about two-thirds of the current price. CONCLUSIONS: Using GS in OA patients may not be cost-effective from the perspective of Thai Ministry of Public Health. Policymakers may consider using such value for money information for aiding policy decision-making. CE2 COST-UTILITY ANALYSIS OF INTENSIVE LIFESTYLE INTERVENTIONS OR METFORMIN VERSUS STANDARD CARE IN THE PREVENTION OF TYPE 2 DIABETES IN HIGH-RISK SUBJECTS IN AN AUSTRALIAN SETTING. A SIMULATION ANALYSIS BASED ON THE LONG-TERM RESULTS OF THE DIABETES PREVENTION PROGRAM AND DIABETES PREVENTION PROGRAM OUTCOMES STUDY Tucker DM, Palmer AJ Menzies Research Institute, University of Tasmania, Hobart, Tasmania, Australia OBJECTIVES: The 3-year Diabetes Prevention Program (DPP) and 7-year follow-up Diabetes Prevention Program Outcomes Study (DPPOS) demonstrated that both intensive lifestyle interventions (ILI) and metformin led to reductions in the incidence of type 2 diabetes (T2D) versus standard care (control) in overweight or obese subjects with impaired glucose tolerance (IGT). Our aim was to project the long-term costutility of T2D prevention in an Australian setting, based on clinical results and resource utilization from the DPP + DPPOS. METHODS: We developed a semiMarkov, second-order Monte Carlo model to project the 10-year clinical and cost results of the DPP + DPPOS to patient lifetimes. Four health states were modeled: normoglycemia (NG); IGT; T2D and dead. Subjects started in IGT and progressed to T2D or NG, depending on the treatment received. State-specific mortality rates for NG, IGT, or T2D were used. We incorporated direct medical costs (from official Australian published sources and the reimbursement perspective) and Australian utility and probability data. Total lifetime costs (C), quality-adjusted life-years (QALY) and incremental C/QALY-gained were calculated for ILI or metformin versus control. Outcomes were discounted at 5% annually. Validations and sensitivity analyses were performed. RESULTS: Delayed onset of T2D led to mean (standard deviation) QALY-gained of 0.38 (0.05) and 0.12 (0.04) years for ILI or metformin versus control, respectively. Costs savings of $282 (4222) per patient and cost increases of $1116 (4338) were projected for ILI or metformin versus control, respectively. ILI was dominant to control. C/QALY-gained for metformin versus control were $8757. Probability of acceptance at a willingness to pay threshold of $50,000 were 100% and 85% for ILI or metformin respectively. Results were most sensitive to probabilities of developing T2D, and costs of implementing the interventions. CONCLUSIONS: Lifetime projection of the DPP + DPPOS results found ILI dominant and metformin to be cost-effective versus control by current Australian standards.

A502

CE3 COST-UTILITY ANALYSIS OF CHOLINESTERASE INHIBITORS IN THE TREATMENT OF MILD TO MODERATE ALZHEIMER’S DISEASE Turongkaravee S Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Nonthaburi, Thailand OBJECTIVES: To evaluate the cost-utility of cholinesterase inhibitors (i.e., donepezil, rivastigmine, and galantamine) compared with no drug treatment in patients with mild to moderate Alzheimer’s disease (AD) based on the governmental and societal perspectives and to estimate the budget impact when providing the cost-effective treatment. METHODS: A Markov model was used to evaluate the cost-utility of drug treatment options versus no drug treatment. Input parameters on the clinical efficacy of cholinesterase inhibitors were obtained from systematic reviews and meta-analysis of international literature. The costs associated with AD were obtained from primary data collection at a university hospital and literature reviews. All costs were presented in 2008 Thai (THB). Probabilistic sensitivity analysis was performed to assess uncertainty surrounding model parameters. RESULTS: Based on both governmental and societal perspectives, galantamine was the most cost-effective compared with other drugs and no treatment option. The ICER at the base case scenario (i.e., patients with 60-year-old who had AD disease for 1 year, an Alzheimer Disease Assessment ScaleCognitive (ADAS-cog) score of 17, no presence of psychotic and extra-pyramidal symptoms (EPS) was 229,367 THB and 157,247 THB per QALY gained using the governmental and societal perspectives, respectively. Moreover, when providing galantamine for patients with EPS or patients with ADAS-cog score ranging from17 to 30 or patients having psychotic symptoms, galantamine was still more cost-effective compared with the base case scenario. Furthermore, the additional budgets required for providing galantamine for all Thai patients with mild to moderate AD at the first year were 12,768 million THB. CONCLUSIONS: At the ceiling threshold of three times of gross domestic product (GDP) per capita (300,000 Baht per QALY), providing galantamine would be a cost-effective intervention for mild to moderate AD patients in Thailand. Galatamine would also be cost-effective at the threshold of onetime of GDP per capita (100,000 Baht per QALY) for patients with EPS. CE4 ECONOMIC EVALUATION OF REHABILITATION SERVICES FOR INPATIENT WITH STROKE IN TWO REGIONAL HOSPITALS IN THAI CONTEXT Khiaocharoen O1, Teerawattananon Y2, Pannarunothai S3, Riewpaiboon W4 1 Naresuan University Faculty of Medicine, Pitsanuloke, Thailand; 2Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand; 3Naresuan University, Muang, Phitsanulok, Thailand; 4The Program of Health Promotion for People with Disabilities in Thai Society, Nonthaburi, Thailand OBJECTIVES: To evaluate cost-effectiveness and cost-utility of rehabilitation for stroke patients under Thai settings. METHODS: This was a prospective observational cohort study with a 4-month follow-up. Subjects were 207 first stroke patients in two regional hospitals. Data was prospectively collected from July 2008 to May 2009. Top-down and bottom-up costing approaches were employed using societal perspective. The Barthel index (BI) was used to evaluate functional status and the EQ-5D was used to assess patients’ quality of life. Effectiveness was defined as improving functional status and QALY. Multivariate analysis of variance, longitudinal logistic model, and multiple regressions were employed. Cost-effectiveness ratios per disability averted score and QALY gained were presented. A probabilistic sensitivity analysis was performed for uncertainty analysis. RESULTS: Compared to the control group, BI and QALY of patients with rehabilitation were significantly improved (P < 0.01). The incremental cost-effectiveness ratio of providing of rehabilitation services for stroke patients was 73,191 Baht per QALY. Cost-effectiveness acceptability curves ensure that the rehabilitation services are likely to represent good value for money at the ceiling ratio of 200,000 Baht per QALY. CONCLUSIONS: The rehabilitation services for disabled stroke survivors were cost effective under the Thai health-care setting.

© 2010, International Society for Pharmacoeconomics and Outcomes Research (ISPOR)

1098-3015/10/A502 A502–A567

4th Asia-Pacific Abstracts PODIUM SESSION I: CANCER STUDIES CN1 CERVICAL CANCER SCREENING PROGRAM IN THAILAND: ASSESSMENT OF SERVICE COVERAGE AND DETERMINANTS OF PROGRAM UPTAKE Sirisamutr T, Butchon R, Putchong C, Praditsitthikorn N, Insrisawang L, Suksomboon N, Tantivess S, Teerawattnanon Y Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Nonthaburi, Thailand OBJECTIVES: In Thailand two approaches for cervical cancer screening, namely Pap smear and visual inspection with acetic acid (VIA), are introduced in the public sector. This study aims to assess the service coverage and the factors associating with screening test seeking behavior in two provinces, namely Chiangmai and Nakon Si Thammarat. METHODS: A cross-sectional household survey was conducted in 2009 in the two study provinces, among women aged 30–60 years old. A total of 1600 women were randomly selected by stratified four-stage sampling. Of these, 1577 were interviewed (98% respondent rate). Descriptive statistics and logistic regression were used in the data analysis. RESULTS: This study suggests that the coverage rates of cervical screening tests in the past 5 years were relatively high; 76% in Chiangmai and 70% in Nakon Si Thammarat. In both provinces, Pap smear was more commonly introduced than the VIA around two to three times. The most common reasons for not seeking screening tests was “the absence of symptoms,” followed by “the lack of time” and “feeling shy toward health workers.” A multivariate logistic regression analysis indicates that supporting and impeding factors of the screening service seeking behavior were significantly associated with screening test. Women with the following characteristics were more likely to seek the tests than others: age 40–50 years, agriculture occupation, child-bearing experience, cervical cancer history in family member, and ever exposing to the information regarding cervical cancer and screening tests. Meanwhile, respondents who were less likely than others to seek the screening services included cigarette smokers. CONCLUSIONS: There were several factors associating with the service seeking practices among women in these settings. In order to increase the service uptake, it is suggested that extensive education program concerning cervical cancer and screenings should be provided with the aim to abolish the misunderstandings and increase awareness among target population. CN2 A SYSTEMATIC REVIEW AND META-ANALYSIS OF ADJUVANT CHEMOTHERAPY FOR STAGE III COLON CANCER Lerdkiattikorn P1, Chaikledkaew U2, Kingkaew P3, Teerawattananon Y4 1 Social Administrative Pharmacy Division, Department of Pharmacy, Faculty of Pharmacy, Mahidol University, Bangkok, Thailand; 2Division of Social and Administrative Pharmacy, Bangkok, Thailand; 3Health Intervention and Technology Assessment Program (HITAP), Ministry of Public Health, Nonthaburi, Thailand; 4Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: This study aimed to assess clinical efficacy of three adjuvant chemotherapy regimens, namely 1) Mayo clinic regimen (5-fluorouracil and leucovorin, 5FU/ LV); 2) FOLFOX4 (oxaliplatin plus 5-FU/LV); and 3) oral capecitabine for treatment of patients with stage III colon cancer. METHODS: A systematic review of randomized controlled trials (RCTs), and meta-analysis of RCT of adjuvant chemotherapy for patients with stage III colon cancer were included by searching through the Medline and Cochrane databases. The clinical efficacy studies of three adjuvant chemotherapy regimens on improving survival outcomes of patients with stage III colon cancer were included. Indirect or mixed-treatment comparison meta-analysis with fix effect model was applied to combine results of several studies. The meta-analysis was carried out using Bayesian approach and WinBUGS14 software program. The summary efficacy of adjuvant chemotherapy were presented as odds ratio (OR) and its 95% confidence interval (CI). To test the variation of study outcomes between studies, heterogeneity test was also applied. RESULTS: Total of 714 abstracts were reviewed and four eligible studies related to adjuvant chemotherapy for patients with stage III colon cancer were included in the meta-analysis. Two studies compared oral capectitabine with 5-FU/LV, while one study compared FOLFOX4 with 5-FU/LV. Indirect comparison was used to compare FOLFOX 4 and oral capecitabine. When compared to 5-FU/LV, FOLFOX4 and oral capecitabine could significantly reduce the risk of death by 23% (OR = 0.77, 95% CI = 0.68–0.86) and 16% (OR = 0.84, 95% CI = 0.72–0.98), respectively. Moreover, the OR of mortality among patients treated by oxaliplatin plus 5-FU/LV was 0.92 (95% CI = 0.76–1.11) compared to capecitabine. CONCLUSIONS: Of three regimens for patients with adjuvant chemotherapy stage III colon cancer, FOLFOX4 could significantly yield the longest patient survival, followed by capecitabine and 5-FU/LV. However, FOLFOX4 did not significantly reduce mortality events compared with capecitabine. CN3 QUALITY OF LIFE IN ADVANCED CANCER PATIENT—COMPARISON OF PATIENT-REPORTED OUTCOME (PRO) AND PROXIES ASSESSMENT Choi J1, Miyashita M2, Kim B3 1 National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea; 2 Tohoku University School of Health Sciences, Sendai, Japan; 3Hanyang University, Seoul, South Korea OBJECTIVES: While Quality Of Life (QOL) in subjects suffering from advanced cancer patient has been studied using a variety of generic or specific instruments, only

A503 very few studies have analyzed the agreement between patients and proxy ratings on patients’ QOL. The objective of this study was to compare PRO of quality of life and proxy assessment using EORTC QLQ PAL-15. METHODS: We administered the EORTC QLQ PAL-15 to 32 patients and their own family, nurse and doctor, respectively as proxies of patient. The QLQ PAL-15 is a 15-item shortened version of the EORTC QLQ-C30 cancer-specific health-related quality of life measure consisting of two functional scale (physical and emotional), seven symptom scale (fatigue, pain, nausea and vomiting, dyspnea, appetite loss, insomnia, constipation), and single-item scale to assess quality of life. The analyses focused on intraclass correlation coefficients (ICCs) to comparing the ICC 95% lower confidence interval with critical value 0.70 and Pearson’s correlation coefficients. RESULTS: Agreement between patients and proxies on the scales was excellent for physical function (ICC = 0.889) and fatigue (ICC = 0.739). Emotional function, emesis, pain, appetite loss, constipation, constipation and quality of life scale was fair agreement (ICC range from 0.471 to 0.739). Dyspnea (ICC = 0.301) and insomnia (ICC = 0.097) was poor agreement between PRO and proxies assessment. There were higher correlation with family than other proxies with patient in emotional function (r = 0.791, P < 0.001), insomnia (r = 0.774, P < 0.001), nausea and vomiting (r = 0.646, P < 0.001), appetite loss (r = 0.638, P < 0.001), dyspnonea (r = 0.402, P < 0.005). There were higher correlation with nurse than other proxies with patient in physical function (r = 0.791, P < 0.001) and constipation (r = 0.540, P < 0.001). There were higher correlation with doctor than other proxies in pain(r = 0.494, P < 0.001), and fatigue (r = 0.406, P < 0.005). CONCLUSIONS: The agreement between PROs and proxies assessment in QOL assessment is different by symptom and function. Family caregiver were more agreement than nurse and doctor. We need to paid attention to proxy assessment more carefully. CN4 CAN WE CORRECTLY TARGET HIGH-COST THERAPIES? THE DIAGNOSTIC ACCURACY OF HISTOLOGY IN DIFFERENTIATING BETWEEN SQUAMOUS AND NON-SQUAMOUS NON-SMALL CELL LUNG CANCER Paech D, Weston A Health Technology Analysts Pty Ltd, Sydney, NSW, Australia OBJECTIVES: The importance of identifying non-small cell lung cancer (NSCLC) histological subtype has increased recently because of the need to target high cost therapeutic agents to the correct patient population. This systematic review was undertaken to examine the diagnostic accuracy of histology in differentiating between subtypes of NSCLC, specifically the ability to differentiate squamous from nonsquamous histology. METHODS: A systematic literature search was undertaken to identify studies that evaluated the reproducibility of histologic diagnosis by pathologists in their reporting of NSCLC subtypes. Studies were screened systematically using a priori defined eligibility criteria. The National Health and Medical Research Council (NHMRC) diagnostic levels of evidence were applied, and quality assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool. Data were extracted and re-analyzed to permit comparison of agreement in non-squamous and squamous cell carcinoma via 2 × 2 tables. Pooled diagnostic performance measures including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) were calculated, with the PPV being considered the most relevant indicator for the treating clinician. RESULTS: Out of 1480 articles identified through the literature search, seven were eligible for inclusion. The PPV of non-squamous histology was consistently high in each individual study with only a small amount of variation between studies. This resulted in a very high pooled PPV of 95.3% (94.0–96.3%) for the primary meta-analysis. CONCLUSIONS: The high individual and pooled PPV suggests that pathologists can reliably differentiate between non-squamous and squamous NSCLC. This is important in guiding oncologist decision-making in choosing the most appropriate therapy for their patients. It is also significant from a purchaser perspective; because it will limit cost leakage from prescriptions mistakenly given outside the drugs approved indication (i.e., to the incorrect NSCLC patient group) due to misdiagnosis.

PODIUM SESSION I: HEALTH TECHNOLOGY ASSESSMENT STUDIES HT1 EARLY ALERT SYSTEMS FOR NEW PHARMACEUTICALS—DO THEY HAVE AN IMPACT ON PHARMACEUTICAL REIMBURSEMENT DECISIONS?: A CROSS-NATIONAL COMPARISON Hiller JE, Mundy L University of Adelaide, Adelaide, South Australia, Australia OBJECTIVES: With the successful incorporation of horizon scanning (HS) into HTA it has been postulated that HS of pharmaceuticals may inform pharmaceutical assessment for public reimbursement in Australia. This paper intends to examine the role of HS for pharmaceuticals, the effect that HS may have had on the introduction of new drugs onto the health market, and assess whether HS for pharmaceuticals would improve access to new drugs in Australia and New Zealand. METHODS: The EuroScan database of HS agencies was searched for pharmaceuticals assessed in 2004 by the National Horizon Scanning Centre (UK) and the Canadian Agency for Drugs and Technologies in Health. Time taken to licensing and public reimbursement or access approval, if given, in the UK, Canada, Australia and New Zealand was ascertained. RESULTS: Of 21 drugs identified by HS in 2004 by the NHSC, 11 received licensing approval and the National Institute for Health and Clinical Excellence has

A504 delivered guidance on five of these, with three being recommended. In Australia, 10/21 of these drugs had licensing approval and seven underwent assessment by the Pharmaceutical Benefits Advisory Committee. All seven received approval for listing on the public reimbursement scheme. Similarly, CADTH assessed 16 drugs by HS in 2004. Of these, 10 received licensing approval but only one gained approval to be listed by the Canadian Expert Drug Advisory Committee. Of these 16 drugs, 14 had licensing approval in Australia and 12 underwent assessment by the PBAC, with only two being declined approval for listing on the public reimbursement scheme. CONCLUSIONS: The current process of pharmaceutical assessment appears to give Australian patients timely access to publicly funded pharmaceuticals and that the introduction of HS for new and emerging drugs would neither decrease the time to access for patients or better inform policymakers than the system already in place. HT2 COST-EFFECTIVENESS OF INTERVENTIONS FOR REDUCING ROAD TRAFFIC INJURIES RELATED TO DRIVING UNDER THE INFLUENCE OF ALCOHOL Ditsuwan V1, Veerman JL2, Bertram M2, Vos T2 1 Thaksin University, Phatthalung, Thailand; 2The University of Queensland, Brisbane, Queensland, Australia OBJECTIVES: To determine the cost-effectiveness of interventions to reduce road traffic injuries caused by driving under the influence of alcohol in Thailand. METHODS: This study used generalized cost-effectiveness analysis. We calculated costs from a health sector perspective. The time-horizon for intervention implementation was 1 year with health outcomes measured over a lifetime. The model covers road traffic crash victims who were injured, disabled, or died due to road traffic crashes. We obtained proportions of alcohol-related crashes, by age and sex of the victims, from the Thai Injury Surveillance system. Random breath testing (RBT), selective breath testing (SBT) and mass media campaigns compared to a “do nothing” scenario. We calculated intervention costs and cost offsets of prevented treatment costs in 2004 Thai Baht and measured benefits in terms of disability-adjusted life-years (DALYs) averted. RESULTS: The study found similar incremental cost-effectiveness ratios (ICERs) of SBT and RBT. SBT was cost saving while RBT had an ICER of 1000 Baht (95% uncertainty interval [UI]: cost saving to 7200) per DALY averted if costoffsets are included. The ICER of mass media campaigns was 8300 (95% UI: cost saving to 45,100). A combination of SBT or RBT and mass media campaigns is very cost-effective compared to a willingness to pay threshold of one times gross domestic product per capita (110,000 Baht) in 2004. CONCLUSIONS: The result shows all interventions are cost-effective, but wide variation is seen in the uncertainty ranges. Sobriety checkpoints and mass media campaigns reduce burden of alcohol-related RTI by 38,000 to 41,000 DALYs. HT3 COST-EFFECTIVENESS OF BLOOD PRESSURE LOWERING WITH A FIXED COMBINATION OF PERINDOPRIL AND INDAPAMIDE IN TYPE 2 DIABETES MELLITUS: A TRIAL-BASED ANALYSIS USING THE ADVANCE STUDY Clarke PM1, Glasziou P2, Alexander J3, Chalmers J4, Patel A5 1 University of Sydney, Sydney, NSW, Australia; 2University of Oxford, Oxford, Oxfordshire, UK; 3University of Queensland, Brisbane, Australia; 4George Institute, Sydney, Australia; 5 George Institute, Sydney, Australia OBJECTIVES: To determine the cost-effectiveness of routine administration, irrespective of blood pressure, of a fixed-dose combination of perindopril and indapamide to patients with Type 2 diabetes. METHODS: Prospective cost-effectiveness analysis within a 20 country randomized trial of 11,140 patients with Type 2 diabetes randomized to perindopril plus indapamide or placebo. We calculated cost per death averted at 4.3 years average follow-up and estimated cost per life-year gained by extrapolation. RESULTS: The ADVANCE trial showed a 14% relative risk reduction in all cause mortality (P < 0.03) and an 18% relative risk reduction (0.8% absolute reduction) in cardiovascular mortality (P < 0.03). Hospital admissions for coronary heart disease and coronary revascularization were reduced by 5%. Per patient perindopril-indapamide cost AUS$1368, but reduced total hospitalization costs by AUS$410 and other medication costs (mainly other blood pressure lowering drugs) by AUS$332. Quality of life, measured by the EQ-5D, was 0.80 (on a 0–1 scale); with no difference between groups. Absolute reduction in all-cause mortality was 1.1%, giving a cost per life saved of AUS$49,200, and life-time extrapolation estimated a cost per life-year saved of AUS$8,470 (discounted at 3%). CONCLUSIONS: The combination of perindopril and indapamide in patients with type 2 diabetes reduces all-cause mortality and appears cost-effective. HT4 HEALTH TECHNOLOGY ASSESSMENT DATABASE IN THAILAND 1 2 2 2 2 2 Kapol N , Lochid-amnuay S , Chalongsuk R , Maitreemit P , Sribundit N , Amrumpai Y 1 Silpakorn University, Nakorn Pathom, Thailand; 2Silpakorn University, A. Muang, Nakhon Pathom, Thailand BACKGROUND: Health technology assessment is a pivotal tool to assist policymakers to decide whether a health technology is efficient. However, the accessibility of health technology assessment information in Thailand is problematic. Attempts to develop Thai health technology assessment database has been raised. The database aims to increase accessibility and utility to policymakers, researchers and practitioners. From January 2008, Thai health technology assessment database has been available online at http://www.db.hitap.net. OBJECTIVES: The objectives of this study are to

4th Asia-Pacific Abstracts determine current contents of the database and evaluate the usefulness of the database. METHODS: A review of the Thai health technology assessment database was conducted to determine the main characteristics of studies, the number of included studies, the number of database members, and number of visitors. Moreover, the use of the database was evaluated by an electronic mail survey to 260 website members. RESULTS: By the end of 2009, the database consisted of 732 articles including three types of studies, 155 full economic evaluation studies, 471 randomized control trials studies and 106 quality of life studies. The number of database members is 711. The number of visitors increases from 1123 to 1717 hits per month in 2008 and 2009, respectively. For the survey results, of 260 members, 41 respondents (15.77%) replied to the electronic mail survey. Most of them used the database for studying and/ or conducting a research. The usefulness of the database was found most in 65.86% of respondents. CONCLUSIONS: Thai health technology assessment database provided useful information to users. However, some types of studies should be included to support potential users’ needs. Moreover, the registration system for the members should be further improved to follow up and effectively disseminate information to users.

PODIUM SESSION I: INFECTIOUS DISEASE STUDIES IN1 ECONOMIC AND CLINICAL BURDEN OF PNEUMOCOCCAL DISEASES AND ACUTE OTITIS MEDIA IN TAIWAN: A NATIONWIDE POPULATION-BASED DATABASE ANALYSIS Chang CJ1, Wang PC2, Huang YC3, Wu BS4 1 Chang Gung University, Taoyuan, Taiwan; 2Cathay General Hospital, Taipei, Taiwan; 3Chang Gung Memorial Hospital, Taoyuan, Taiwan; 4National Yang-Ming University, Taipei, Taiwan OBJECTIVES: Streptococcus pneumonia can cause invasive diseases such as meningitis and bacteremia, and noninvasive diseases such as pneumonia and acute otitis media (AOM), leading to high morbidity and mortality in infants and the elderly. Limited data are available for the burden of these diseases in Taiwan. This study investigates economic and clinical burden of pneumococcal diseases and AOM in Taiwan. METHODS: A retrospective population-based National Health Insurance Reimbursement Database (NHIRD) study was performed to estimate the incidence and direct medical costs of pneumococcal meningitis/bacteremia/pneumonia and AOM from 2002 to 2007 from healthcare provider perspective, including only NHIcovered costs. RESULTS: The incidence of pneumococcal meningitis is 0.1 per 100,000 and highest at age 90% of being cost-effective compared with enoxaparin at a low willingness-to-pay threshold of CNY 20,000 per QALY gained. CONCLUSIONS: Compared with enoxaparin, rivaroxaban improved patients’ health outcomes and produced overall cost savings in VTE prevention after TKR in China. CV4 THE CLINICAL EFFECTIVENESS OF 64-SLICE OR HIGHER COMPUTED TOMOGRAPHY ANGIOGRAPHY AS AN ALTERNATIVE TO INVASIVE CORONARY ANGIOGRAPHY IN THE INVESTIGATION OF SUSPECTED CORONARY ARTERY DISEASE Paech D, Weston A Health Technology Analysts Pty Ltd, Sydney, NSW, Australia OBJECTIVES: This systematic review was conducted for The New Zealand Ministry of Health to summarize recent evidence pertaining to the clinical effectiveness of 64-slice or higher computed tomography angiography (CTA) in patients with suspected coronary artery disease (CAD). If CTA proves to be a successful diagnostic performance measure, it could prevent the application of invasive diagnostic procedures in some patients. This would provide multiple health and cost benefits, particularly for under resourced District Health Boards where invasive coronary angiography is not always available. METHODS: A systematic method of literature searching and selection was employed with searches limited to December 2006 to March 2009. Included studies were quality assessed using NHMRC diagnostic levels of evidence and a modified QUADAS tool. Individual and pooled diagnostic performance measures (i.e., sensitivity, specificity, positive predictive value (PPV), negative predictive

A507

4th Asia-Pacific Abstracts value (NPV) and overall diagnostic accuracy) were calculated at the patient, vessel and segment level. RESULTS: This systematic review included 28 studies. The base case meta-analysis at the patient-level indicated a sensitivity of 98.2%, specificity of 81.6%, PPV of 88.9%, NPV of 96.8%, and diagnostic accuracy of 91.6%. In all vessels, the pooled sensitivity was 95.0%, specificity 85.2%, PPV 69.4%, NPV 97.9%, and diagnostic accuracy 87.7%. At the individual artery level, overall diagnostic accuracy appeared to be slightly higher in the left and right coronary artery and slightly lower in the left anterior descending and circumflex artery. In all segments, the sensitivity was 91.1%, specificity 94.3%, PPV 65.7%, NPV 98.9%, and overall diagnostic accuracy 94.0%. CONCLUSIONS: The high sensitivity observed in this update indicates that CTA can effectively identify the majority of patients with significant coronary artery stenosis. The high NPV at the patient, vessel and segment level establishes CTA as an effective noninvasive alternative to ICA for the exclusion of stenosis.

PODIUM SESSION II: DIABETES STUDIES DB1 EVALUATING THE COST-EFFECTIVENESS OF THERAPY CONVERSION FROM BASAL INSULIN TO BIPHASIC INSULIN ASPART 30/70 IN PATIENTS WITH TYPE 2 DIABETES IN CHINA: A MODELING STUDY OF LONG-TERM COSTS AND HEALTH OUTCOMES Chang J1, Sun F1, Li H2 1 Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China; 2China Pharmaceutical University, Nanjing, Jiangsu, China OBJECTIVES: To evaluate the long-term cost-effectiveness of switching from basal insulin to biphasic insulin aspart 30 (BIAsp30) in patients with type 2 diabetes (T2D) in China. METHODS: A published and validated computer simulation model of diabetes (CORE Diabetes Model) was used to project long-term (30 years) of health and economic outcomes. Simulated cohorts and treatment effects were derived from a 16-week, multi-center, and single-arm trial-NCT00669864 which investigated the efficacy and safety of BIAsp30 ± Metformin in T2D patients inadequately controlled with basal insulin. Two subgroups of basal insulin treatment were categorized as insulin glargine (IGla) ± Metformin and neutral protamine hagedorn (NPH) insulin ± Metformin. The market retail prices of medications were calculated to estimate treatment costs. The diabetes management and complications costs were obtained from Chinese published data. An annual discounting rate of 3% was used for both costs and health outcomes. One-way sensitivities analysis was performed. RESULTS: Therapy conversion to BIAsp30 was projected to improve life expectancy significantly in comparison with IGla (0.347 ± 0.245 years), and NPH (0.452 ± 0.242 years). Transfer to BIAsp30 was associated with improvements in 0.327 quality-adjusted life-years (QALYs) over IGla, and 0.393 QALYs over NPH. Therapy conversion to BIAsp30 reduced medical costs by Chinese Yuan (CNY) 46,540 per patient compared to IGla. However, it increased CNY 19,525 compared to NPH and was associated with an incremental cost-effectiveness ratio of CNY 49,730 per QALY gained. CONCLUSIONS: Therapy conversion from basal insulin to BIAsp30 in T2D patients in China was associated with improvements in life expectancy and QALYs. Transfer to BIAsp30 was cost-saving treatment strategy in T2D patients managed with IGla, and would be considered cost-effective in T2D patients managed with NPH, given a willingness-to-pay threshold of CNY 75,375 per QALY (three times GDP per capita in 2009) gained in China. DB2 TRANSLATION AND VALIDATION OF MICHIGAN DIABETES KNOWLEDGE SCALE INTO MALAYSIAN VERSION Al-Qazaz HK1, Hassali MA1, Shafie AA1, Sulaiman SA1, Sundram S2 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Hospital Balik Pulau, Balik Pulau, P.Penang, Malaysia OBJECTIVES: To translate the Michigan Diabetes Knowledge Scale (MDKT) into the Malaysian language, and to examine the psychometric properties of the Malaysian version of the MDKT among patients with type 2 diabetes, including its validity and reliability. METHODS: After obtaining permission, a standard “forward-backward” translation procedure was used to create the Malaysian version of the MDKT from the original English version. A convenience sample of 307 outpatients with type 2 diabetes was identified between May and October, 2009. All data were collected from the Penang General Hospital, Penang, Malaysia. Instruments consisted of the Malaysian version of MDKT and a socio-demographic questionnaire. Medical records were reviewed for hemoglobin A1C (HbA1C) levels and other clinical data. Reliability was tested for internal consistency using Cronbach’s α coefficient. Validity was confirmed using known group validity. RESULTS: Employing the recommended scoring method, the mean ± SD of MDKT scores was 7.88 ± 3.01. Good internal consistency was found, (`’s α = 0.702), the test–retest reliability value by using Spearman’s rank correlation was 0.894 (P < 0.001). For known group validity, a significant relationship between MDKT categories and HbA1c categories (×2 = 21.626; P ≥ 0.001) was found. CONCLUSIONS: The MDKT can be used for diabetes knowledge assessment in diabetes. The findings of this validation study indicate that the Malaysian version of the MDKT is a reliable and valid measure of medication adherence which can now be used in clinical and research practice.

DB3 LIFETIME CLINICAL PROJECTIONS FOR OVERWEIGHT OR OBESE SUBJECTS WITH IMPAIRED GLUCOSE INTOLERANCE BASED ON THE LONG TERM RESULTS OF THE DIABETES PREVENTION PROGRAM AND DIABETES PREVENTION PROGRAM OUTCOMES STUDY Palmer AJ, Tucker DM Menzies Research Institute, University of Tasmania, Hobart, Tasmania, Australia OBJECTIVES: Metformin and intensive lifestyle interventions (ILI) were shown to reduce incidence of type 2 diabetes (T2D) versus standard care in overweight or obese subjects with impaired glucose tolerance (IGT) in the Diabetes Prevention Program (DPP) trial and Diabetes Prevention Program Outcomes Study (DPPOS), a total follow-up of 10 years. Our aim was to project the lifetime clinical outcomes to be expected from T2D prevention in high-risk subjects treated with standard care, metformin or ILI, based on the results from the DPP + DPPOS. METHODS: A semiMarkov, second-order Monte Carlo model was developed to project the 10-year clinical results of the DPP + DPPOS to patient lifetimes. Specifically, we calculated years free of T2D, cumulative incidences of T2D, and nondiscounted life expectancies in subjects who were initiated on diabetes prevention regimens based on metformin, ILI or standard care. Four health states were modeled: normoglycemia (NG); IGT; T2D and dead. Subjects started in IGT and progressed to T2D or NG, at rates dependent on the treatment received. State-specific mortality rates for NG, IGT or T2D were used. Univariate and probabilistic sensitivity analyses were performed. RESULTS: For standard care, metformin or ILI, mean (standard deviation) number of years free of T2D were 9.47 (0.08), 11.98 (0.09), 15.17 (0.11) years respectively. Cumulative incidences of T2D were 89.7% (0.2), 83.7% (0.2) and 73.4% (0.3%) for standard care, metformin or ILI respectively. Mean life expectancies from baseline age of 50 years were 27.64 (0.14), 27.95 (0.12), 28.33 (0.11) years for standard care, metformin or ILI respectively. Results were most sensitive to the relative risk reduction in the incidence of T2D and relative risks of mortality in the T2D state versus IGT state. CONCLUSIONS: Substantial improvements in lifetime clinical outcomes can be expected in high risk subjects treated with metformin or ILI to delay or prevent the onset of T2D. DB4 ECONOMIC EVALUATION OF THIAZOLIDINEDIONES AS ADD-ON THERAPY FOR TREATMENT OF TYPE 2 DIABETIC PATIENTS IN THE TAIWANESE NATIONAL HEALTH INSURANCE SYSTEM Hsiao FY1, Mullins CD1, Huang WF2 1 University of Maryland School of Pharmacy, Baltimore, MD, USA; 2Institute of Health & Welfare Policy, National Yang-Ming University, Taipei, Taiwan OBJECTIVES: The cost-effectiveness of adding thiazolidinediones (TZDs), rosiglitazone or pioglitazone, to metformin in treating type-2 diabetes mellitus was assessed from a Taiwanese national health insurance perspective. METHODS: This analysis was based on patient-level data extracted from the 2000–2005 Taiwan’s National Health Insurance (NHI) databases. Type 2 diabetic patients who had their first ambulatory visits with a diagnosis of diabetes mellitus and had received consecutive metformin treatments between 2001 and 2005 were identified. Clinical effectiveness, a proxy of glycemic control (time to insulin dependence), and direct medical cost also were estimated from the NHI databases. Incremental cost-effectiveness ratio (ICER) was calculated and expressed as cost per delayed year to insulin dependence. RESULTS: The use of TZDs as add-on therapy compared non-TZDs add-on therapy was associated with a delay in time to insulin dependence, rosiglitazone was associated with an additional 151 days (0.41 years) and pioglitazone was associated with an additional 101 days (0.28 years) of delay in insulin dependence. During the follow-up period, total mean medical costs were higher in patients who received an add-on rosiglitazone (New Taiwan dollars (NT) 153,162) or pioglitazone (NT 139,931) compared to add-on non-TZDs (NT 113,492) and the additional medical costs were driven primarily by diabetic medication cost and outpatient visit costs. Combining the cost and effectiveness results, the ICER showed that the additional total medical costs of add-on rosiglitazone or pioglitazone were comparable, with ICERs of 95,874 and 95,485 NT dollars per year delay in insulin dependence, respectively. CONCLUSIONS: This analysis suggests that add-on rosiglitazone or pioglitazone improves glycemic control but also increases direct medical costs compared with add-on nonTZDs when used in type-2 diabetic patients. In terms of the incremental medical costs associated with these clinical benefits, add-on rosiglitazone or pioglitazone are similar in the National Health Insurance system in Taiwan.

PODIUM SESSION II: DATABASE STUDIES DS1 ESTIMATING ADHERENCE AND PERSISTENCY OF ANTIDEPRESSANTS USING THE KOREA NATIONAL HEALTH INSURANCE CLAIMS DATABASE Jung SY, Song H, Shin S, Shin E, Park J1, Ahn J National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea OBJECTIVES: To investigate adherence and persistence of antidepressants (ADs) among the patients with depression in Korea. METHODS: Using the Korean Health Insurance Review & Assessment Service (HIRA) claims database (2006–2008), patients aged 18–84 with at least one inpatient or two outpatient diagnoses of depres-

A508 sion (ICD 10 = F06.3, F31.3, F31.4, F32, F33, F34.1, F38.1, F41.2) within 1 year before the first observed prescription of ADs between January 2007 to June 2008 (Index period: 18 months) were identified. ADs were categorized as tricyclic antidepressants (TCAs), selective serotonin reuptake inhibitors (SSRIs), and new antidepressants (NADs; SNRIs and others). For each AD category, adherence by Medication possession ratio (MPR) of 180 days post-index period and persistence by duration of uninterrupted therapy (DUR) with a 14-day permissible gap were estimated. Additional subgroup analyses by non-adherence (MPR < 75%) and by early dropout (DUR < 90 days) were included. Multivariable logistic regression analysis was performed to further investigate the relationship between AD categories and adherence (persistence). RESULTS: Of 318,490 patients satisfied the selection criteria, mean MPR was 0.52 (SD = 0.34) and mean duration of uninterrupted therapy of were 118.08 days (SD = 67.67). Non-adherence proportion was 65.9%, and early dropout rate was 36.7%. Patients prescribed by psychiatrists (N = 208,791) showed significantly higher MPR (0.54 vs. 0.40; P < 0.001) and persistence (125.43 vs. 102.67; P < 0.001). When adjusted for age, gender, pre-/post- inpatient or outpatient visits, Chalson’s comorbidity index, hypertension, and hyperlipidemia, with TCA as reference, patients prescribed NADs (aOR = 1.75 [1.70–1.80]) and SSRIs (aOR = 1.56 [1.52–1.60]) were associated higher adherence, whereas SSRIs were associated higher (aOR = 1.03 [1.01–1.05]) but NADs were associated lower (aOR = 0.93 [0.91–0.96]) persistence. CONCLUSIONS: Contrary to the previous literature, Korean adherence and persistence of ADs were not significantly lower than the United States or Europe. In addition, NADs showed an advantage in terms of adherence and SSRIs had an edge in terms of persistence. DS2 TREND OF ANTIHYPERTENSIVE MEDICATIONS AND RISK OF STROKE AMONG HYPERTENSIVE PATIENTS IN TAIWAN: BASED ON THE NHI DATABASE Chu NF Shuang-Ho Hispital, TMU, School of Public Health, NDMC, Taipei, Taiwan, Taiwan OBJECTIVES: Hypertension (HTN) is an important and prevalent risk factor for cardiovascular diseases. In Taiwan, the prevalence of hypertension is around 21.4% and only about 40% patients with acceptable range of blood pressure control. The purpose of this study is to evaluate the trend of antihypertensive medications and also examine the risk of stroke among these HTN patients from 2001 to 2004 in Taiwan. METHODS: Based on the National Health Insurance (NHI) database, there are 1,931,436 patients with ICD-9 coding of hypertension, we randomly selected 10% hypertension outpatients (194,547 subjects) from the NHI claim database, after excluded those visited less than three times or without anti-HTN medication, 135,660 patients in the finally analyses. These subjects follow their anti-HTN medications and occurrence of stroke for a 4-year period. RESULTS: Among these patients, 27.6% used only one medication and 37.0% with three or more medications for HTN. There are 9530 (4402 males and 5128 females) patients had both drug information on 2001 and 2004. The prevalence of stoke was 7.3% in 2001 and increased to 13.6% in 2004. The average out-patient medical costs for hypertension were increased from NT$15,778.7 to 19,712.1 dollars (at an exchange rate of 32 NT$ to 1 US$) from 2001 to 2004. Using Cox proportion hazard model, gender (male) and age were important factors to predict the occurrence of stroke. Comparing with beta-blocker, diuretics had higher relative risk for stoke but the risk was attenuate after adjusted for other confounders (HR = 1.3, 95% CI 0.7−2.6). CONCLUSIONS: From this study, the out-patient medical costs for hypertension were increased from 2001 to 2004 in Taiwan. However, the occurrence of stroke among hypertensive patients was also increased during this period. There was no significant protective effect for stroke among hypertensive patients using different anti-HTN medications. Further studies may be indicated to evaluate the cost-effectiveness of different medications on hypertensive-related disorders. DS3 A COMPARISON OF SIGNAL DETECTION PERFORMANCE BETWEEN REPORTING ODDS RATIO AND BAYESIAN CONFIDENCE PROPAGATION NEURAL NETWORK METHODS ON ADVERSE DRUG REACTION SPONTANEOUS REPORTING DATABASE OF THE THAI FDA Bunchuailua W1, Zuckerman I2, Kulsomboon V3, Suwankesawong W4, Singhasivanon P5, Kaewkungwal J6 1 Faculty of Pharmacy Silpakorn University, Muang, Nakhon Pathom, Thailand; 2University of Maryland, Baltimore, Baltimore, MD, USA; 3Chulalongkorn University, Bangkok, Thailand; 4Food and Drug Administration, Muang, Nonthaburi, Thailand; 5Faculty of Tropical Medicine Mahidol University, Ratchawithi, Bangkok, Thailand; 6Mahidol University, Bangkok, Bangkok, Thailand BACKGROUND: Several statistical methods have been applied to detect signals in spontaneous reporting databases. The Thailand Food and Drug Administration (Thai FDA) uses the reporting odds ratio (ROR) method for signal detection because of its ease of implementation and interpretation. The performance of different methods needs to be explored to determine need for modifications to the Thai FDA signal detection system. OBJECTIVES: To examine the concordance between the ROR and the Bayesian Confidence Propagation Neural Network (BCPNN) methods in identifying adverse drug reaction (ADR) signals using the Thai FDA database. METHODS: The two methods were retrospectively applied to identify ADRs reported with antiretroviral (ARV) drugs using the dataset from 1990 to 2006. The criteria of lower limit of 95% confidence interval of ROR >1, 3 or more cases; and information component (IC) two standard deviations >0, were used to identify signals for ROR and

4th Asia-Pacific Abstracts BCPNN, respectively. The sensitivity, specificity and agreement of timing of signal detection were measured for the concordance of the ROR in respect to the BCPNN. RESULTS: Using the BCPNN as a reference method, the ROR has high sensitivity and specificity. For the agreement of timing of signal detection between the ROR and the BCPNN, signals detected by both methods were in agreement on the first time of signal detection for 76.92%. CONCLUSIONS: ROR and BCPNN are comparable in identify signals of potential ADRs. Comparisons using other drug classes will provide additional insight into the performance of these two methods. DS4 HERBAL PRODUCTS SAFETY: AN APPLICATION OF HEALTH PRODUCT VIGILANCE CENTER DATABASE Saokaew S1, Suwankesawong W2, Permsuwan U3, Chaiyakunapruk N4 1 Naresuan University Phayao, Muang, Phayao, Thailand; 2Food and Drug Administration, Muang, Nonthaburi, Thailand; 3Chiang Mai University, Muang, Chiang Mai, Thailand; 4Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: The use of herbal products continues to expand rapidly across the world. The concerns of safety of herbal products have been raised. Health Product Vigilance Center (HPVC) database (Thai Vigibase) is the national existing health products surveillance system which can be used to detect signal of adverse events (AEs) in patients receiving herbal products. This study aims to describe and characterize reported AEs in patients receiving herbal products in Thailand. METHODS: Thai Vigibase data from February 2000 to December 2008 involving AEs of reported herbal products were included. The database contained patients’ demographic, AEs associated with herbal products, seriousness, causality and quality of reports. Descriptive statistics were used for data analyses. RESULTS: A total of 593 reports with 1868 AEs were reported during the study period. The age range of reported cases was 1 to 86 years (mean, 47 years). Female patients were frequently reported (72%). Reports involving turmeric were most frequently reported. Gastro-intestinal problems were the most common clinical outcomes among herbal products associated with adverse events. Severe adverse events such as Stevens Johnson syndrome, anaphylactic shock and exfoliative dermatitis were also found. CONCLUSIONS: HPVC database is a potentially effective tool for herbal adverse events monitoring. Encouraging reporting can contribute to improve awareness among health personnel and patients about the benefit-harm profile of the herbal products.

PODIUM SESSION II: DRUG USE STUDIES DU1 THE IMPACTS OF ABUSE OF ANTIBIOTIC THERAPIES IN CHINA Yang L Peking University, Beijing, China OBJECTIVES: In the 1980s, China launched market-oriented reforms. Public hospitals were encouraged to make their own incomes with the aim of mobilizing medical workers and improving hospital efficiency. Less government funding resulted in deficits for public health institutions, which forced hospitals to generate their own revenue by aggressively selling drugs, especially antibiotics. This study was designed to evaluate the impact of inappropriate antibiotic use on inpatients’ cost during the hospitalization. METHODS: One thousand cases with antibiotic treatment from 10 hospitals of five provinces in China in 2006. We created multivariate linear regression model for hospital cost and logistic regression model for rationality evaluation of antibiotic use. RESULTS: Finally we collected 964 valid cases. Rate of inappropriate antibiotic use was 58.4%. Costs of inpatients with inappropriate antibiotic use was as 2.75 times as the ones with appropriate use (P < 0.001). Risk factors included antibiotic prophylaxis (OR = 2.929), operations (OR = 2.44), long hospital stay (OR = 1.021 for every prolonged day) and regional factors. Protection factor was in tertiary hospital (OR = 0.510). CONCLUSIONS: This study concluded that inappropriate antibiotic use contributes to inpatients’ high cost. Efforts to control misuse of antibiotic should be pursued. DU2 THE UTILIZATION OF PROPHYLACTIC ANTITHROMBOTIC AGENTS AFTER MAJOR ORTHOPEDIC SURGERIES—A POPULATION-BASED STUDY IN TAIWAN Yu YY1, Chen LC1, Cheng LJ2 1 Kaohsiung Medical University, Kaohsiung, Taiwan; 2University of Macau, Macau, Macau OBJECTIVES: Antithrombotic agents have been recommended to prevent venous thromboembolism (VTE) after major orthopedic surgeries, yet neither effectiveness evidence nor guidance for prophylactic antithrombotic agents is available in Taiwan. This study aims to evaluate current utilization of prophylactic antithrombotic agents after total hip and knee replacement (THR/TKR). METHODS: This retrospective cohort study used a Taiwanese National Health Insurance dataset including 1,000,000 beneficiaries randomly sampled in 2005 from nationwide population and followed longitudinally from 1995 to 2008. Adults hospitalized for THR or TKR were identified by procedure codes from April 2003 to June 2008. Patients with VTE, bleeding events, or antithrombotic prescriptions within 90 days preadmission were excluded. Eligible patients’ characteristics, medication during operation, and antithrombotic prescription up to 30 days post-admission were collected, presented in descriptive statistics, and stratified by different operations. RESULTS: Of all, 1026 THR (41.03% women) and 2401 TKR (75.51% women) were identified, with mean age of 56.36 ±

A509

4th Asia-Pacific Abstracts 14.50 and 69.07 ± 8.28 years, and average length of stay for 9.41 ± 4.45 and 9.17 ± 4.55 days, respectively. Most procedures were conducted at medical centers (45.52% and 37.57%). The most frequent diagnoses for THR are avascular necrosis (45.22%) and osteoarthritis (44.05%); while for TKR is osteoarthritis (96.42%). Only 40 THR (3.90%) and 160 TKR (6.66%) had ever been prescribed antithrombotic agents after surgeries, and mostly used aspirin alone (18 THR and 95 TKR). CONCLUSIONS: According to this nationwide individual patient dataset, antithrombotic agents are not commonly used after THR and TKR in Taiwan, but aspirin is the most frequently prescribed of all. These findings deviate from the recommendations of current international guidelines. To establish Taiwanese own guidance, it is necessary to further investigate the effectiveness and safety of prophylactic antithrombotic agents after THR and TKR, and explore the influences of reimbursement policy and genetic factors on these outcomes. DU3 ASSESSMENT OF THE KNOWLEDGE AND PERCEPTIONS OF THE MALAYSIAN COMMUNITY PHARMACISTS REGARDING PHARMACOVIGILANCE: FINDINGS FROM AN INTERVENTIONAL STUDY Elkalmi RM1, Hassali MA1, Izham M2 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia OBJECTIVES: To assess the knowledge and perception of community pharmacists in Malaysia toward reporting of adverse drug reactions (ADRs). METHODS: A crosssectional study employing pre and post intervention design was undertaken with a sample of 42 community pharmacists to explore their knowledge of and the perception toward the pharmacovigilance and ADRs reporting. Changes in scores before and after the completion training program in knowledge and perception were measured. Statistical analysis: The data were analyzed by using SPSS version 17.0 software package (SPSS Inc., Chicago, IL, USA). Both descriptive and inferential statistics were used when ever appropriate. Frequencies and percentages were used to describe the respondents’ demographic information and professional characteristics. Wilcoxon Signed ranks test, Paired sample t test and The McNemar χ2 test were applied to compare the differences in knowledge and perception before and after the educational program whenever appropriate. RESULTS: Forty-two community pharmacists (CPs) were enrolled in this study. A comparison of CPs knowledge before and immediately after implementation the education program showed significant differences, the participated community pharmacists’ mean scores in knowledge of pharmacovigilance and adverse drug reactions reporting significantly increased compared to the baseline (z = −5458, N = 42 P < 0.001). Following the intervention, only few (3, 7.2%) respondents reported that they do not have any ideas on how to report ADR to the relevant authorities in Malaysia. Upon completion of the education program, a large proportion of the pharmacists (45%) believed that they were confident in their knowledge in detection and performing ADRs reporting (z = −2.866, N = 42, P = 0.004). CONCLUSIONS: The finding of the study indicates the need for designing educational programs for the community pharmacists, in order to improve the level of knowledge toward the pharmacovigilance and ADRs reporting process locally. DU4 COST-EFFECTIVENESS ANALYSIS OF TRAVOPROST FOLLOWED BY FIXED COMBINATION TRAVOPROST/TIMOLOL TREATMENT SEQUENCE (T-TTFC) COMPARED TO LATANOPROST FOLLOWED BY LATANOPROST/TIMOLOL FIXED COMBINATION (L-LTFC) FOR PATIENTS WITH OPEN ANGLE GLAUCOMA OR OCULAR HYPERTENSION IN MALAYSIA AND INDIA Taylor M1, Tan R2 1 York Health Economics Consortium, York, UK; 2Alcon Laboratories R&D International, Singapore OBJECTIVES: To assess the cost-effectiveness of T-TTFC compared with L-LTFC treatment sequences for patients with open angle glaucoma or ocular hypertension (OHT) using a payer perspective in Malaysia and India. METHODS: A Markov model simulated the disease progression of open angle glaucoma and ocular hypertension (OHT) patients in India and Malaysia. Transition probabilities were obtained from Denis (2008), and extrapolated from two phase III clinical trials (Topouzis, 2007; Netland; 2001). Probabilities to develop new visual field defect (VFD) was estimated from literature (Denis, 2004). Utilization patterns were derived from local hospitals and literature. Unit costs were obtained from list prices. The time horizon was 5 years and sensitivity analyses were performed on key parameters. RESULTS: At 5 years, fewer T-TTFC patients switched to third line treatment, and 65.5% of patients showed no disease progression (no new VFDs) compared to 61.4% for L-LTFC Thus initiation of treatment in an OHT patient with T followed by TTFC if needed would avoid one incidence of VFD in every 24 incident cases (Number Needed to Treat—NNT). In Malaysia, T-TTFC results in a cost-savings of 1286 RM (US$378) against L-LTFC. In India over a 5-year period, T-TTFC results in a cost savings of 11,184 INR (US$243) against L-LTFC. Sensitivity analyses have demonstrated the robustness of these finding. CONCLUSIONS: From the payer perspective, T-TTFC yielded costsavings and fewer treatment switches. Poor IOP control results in therapy changes and increases the probability of patients experiencing new VFDs (Denis 2004), which in turn adversely impacts patient quality of life. Our savings estimates are conservative because potential cost savings, as a consequence of fewer ophthalmologist visits, particularly for patients who use private healthcare facilities, have not been captured. With these findings policy-makers and clinicians should consider the use of T-TTFC for glaucoma patients as potential cost-saving medications.

PODIUM SESSION II: HEALTH EXPENDITURE STUDIES HE1 A CROSS-COUNTRY COMPRASION OF HEALTH PERFORMANCE BETWEEN TAX-FUNDED AND SOCIAL HEALTH INSURANCE SYSTEM: THE CHANNEL VIA PREVENTIVE CARE Chen Q, Li L Peking University, Beijing, China OBJECTIVES: Tax-funded system and the social health insurance (SHI) system are two main choices for health-care system reform. This research investigates the differences of health outcome and health expenditure between these two systems. The channels of such differences are also valuable for policymaker. We further explore the channel via public health and preventive care investment that leads to these differences. We are also interested in the impact of fiscal and political decentralization on health systems. METHODS: We merge the data from a variety of sources including WHO, the World Bank, OECD together with database prepared by independent researchers like Polity IV and Government Performance Indicator. After controlling the variables indicating the social-economic situations, we implement the Pooled-OLS Method and Fix-Effect Method for panel data. Then check the model specification by testing the autocorrelation of the residuals. RESULTS: When the share of SHI in public health expenditure increases, the total health expenditure (THE) measured as its share in GDP increase dramatically. However, if there is local responsibility, this negative impact will be mitigated. What’s more, the increase of SHI share also plays down the life expectancy, but with local responsibility the negative impact can get amended. It is worth the attention that the public health performance, measured by the share of expenditure in THE and a series of public health behaviors including the immunization rates and infectious disease incidences, presents a consistent pattern. That is, with the increase of the SHI share, the public health performance is weakened, but with local responsibility, the performance will be improved to some extent. CONCLUSIONS: Tax-funded system is more efficient in the sense of cost-benefit, and public health and preventive care investment are important channels of these different performances between the two types of health-care system. HE2 NEW DRUGS AND THE GROWTH OF HEALTH EXPENDITURE: EVIDENCE FROM DIABETIC PATIENTS IN TAIWAN Liu YM1, Hsieh CR2 1 National Cheng Kung University, Tainan, Taiwan; 2Academia Sinica, Taipei, Taiwan OBJECTIVES: The aim of this paper is to investigate whether the adoption of pharmaceutical innovation increases the overall expenditure on health care by focusing on diabetic patients. METHODS: By examining the National Health Insurance sampling longitudinal claims data between 2000 and 2004, we use a new class of drugs, namely, thiazolidinediones (TZD), as an example to investigate the effect on health expenditure of prescribing new drugs to patients by decomposing the impact into treatment substitution and treatment expansion effects. The difference-in-difference approach was used to estimate the effects of adopting new drugs by comparing the difference between the users of TZD drugs (the treatment group) and the nonusers of TZD drugs (the control group) as well as the difference between the pre and post periods. RESULTS: Our results indicate that the introduction of new drugs mainly impacts the outpatient drug expenditure and does not give rise to any offsetting effect on other outpatient and inpatient health expenditures. This suggests that the adoption of pharmaceutical innovation in treating diabetic patients is expenditure-increasing. In addition, we find evidence that the treatment substitution channel has a more significant impact on health expenditure than the treatment expansion channel. CONCLUSIONS: An important policy implication for our finding is that new prescription drugs do not pay for themselves, suggesting that the justification for increasing health expenditure on the treatment of diabetes is not conditional upon a lowering in the demand for other types of health-care services. By contrast, it is conditional upon the increased health benefits per se. HE3 ANALYSIS OF AN AFFORDABLE PRICE FOR PERFORMANCE-BASED RISK SHARING SCHEME: OMALIZUMAB ADAPTATION IN KOREA Kim S, Kim J, Kim J, Kim C, Shin K, Kim K, Kim H Novartis Korea, Seoul, Seoul, South Korea OBJECTIVES: The innovative but expensive biologic agents are introduced in recent years, Korea government is meeting with difficulty in funding for these agents due to limited health-care budget. Pay for performance as risk sharing scheme has been adopted to overcome this situation in other countries. In this article we tried to find an affordable price based upon a budget neutral concept and adapted to risk sharing of Omalizumab. METHODS: Omalizumab, for treatment of moderate to severe allergic asthma, is only treated for responders at 16 weeks. We modeled the performance-based risk sharing along four assumptions: 1) no additional national healthcare budget, budget of risk sharing ≤ budget of no risk sharing; 2) positive manufacturer’s profit; 3) reimbursed only for responders at 16 weeks; and 4) manufacturer funding to responder in screening period. We also simulated the model by changing of input parameters. RESULTS: We defined and analyzed the function of risk sharing scheme based on national health-care budget and manufacturer’s profit. Lower boundary of affordable price was derived from the minimum of acceptable predicted sales volume for manufacturer and upper was calculated by a point of budget neutral. Lower boundary was highly depends on unit selling price, production cost

A510 and projected vials. Upper boundary was influenced more by total target patients, response rate, and expected price with no risk sharing. CONCLUSIONS: The concept of performance-based payment is a risk sharing between the payer and manufacturer for high-quality of new agent, and improves to patient’s quality of life within the available national health-care budget. We show that the range of affordable price is calculated by a manufacturer’s profit and neutral point of health-care budget in the performance-based risk sharing and discuss potential challenges. HE4 THE EFFECT OF PRICE CONTAINMENT ON THE TREND OF PHARMACEUTICAL EXPENDITURE FROM 1999 TO 2007 IN TAIWAN Lang HC1, Tan CH2 1 National Yang-Ming University, Taipei, Taiwan; 2National Taiwan University, Taipei, Taiwan OBJECTIVES: To control the high pharmaceutical expenditure, the Bureau of National Health Insurance introduced the generic grouping method to reduce price variation among off-patent drugs since 2000. The purpose of this study was to examine the effect of the price adjustments on the trend of pharmaceutical expenditure. METHODS: We used BNHI monthly claim data from 1999 to 2007 and incorporated drug registry file, drug price file and 2009 WHO ATC/DDD file. To further analyze the effect of generic grouping, we classified the branded drug into three categories patented branded drug, off-patent branded drug with no competitive generic drug and off-patent branded drug with competitive drug. RESULTS: We found that from 1999 to 2007, the annual growth rate of pharmaceutical expenditure was 6.0%; outpatient 6.3% and inpatient 5.5%. During the study period, the sales of the patented branded drugs rose from 20.5% to 30.9% with the highest increasing rate. The sales of the off-patent branded with competitive generic drugs and generic without the BA/BE testing drugs fell from 21.1% to 18.6% and from 39.3% to 28.4%, respectively. The sales volume in terms of DDD also has the same pattern as the drugs sales. Regarding the impact of drug price adjustment on Hospital, we found the share of drug claim by medical centers rose from 35.5% to 41.3%; regional hospitals from 25.3% to 30.2%, however local hospitals and primary health care decreased their share. In terms of firms, the IRPMA group held the highest increasing rate on share and importers held the lowest. CONCLUSIONS: The generic grouping price adjustment policy has effectively reduced the off-patent branded drug expenditure. However, total drug expenditure didn’t decrease due to the increased expenditure of the new drugs with high prices.

POSTER SESSION CANCER – Clinical Outcomes Studies PCN1 THE EVALUATION OF ADR ONLINE REPORTING OF RADIOPAQUE AGENTS IN A UNIVERSITY HOSPITAL IN TAIWAN Huang CY, Hsieh YW, Lin WL, Wu HC China Medical University Hospital, Taichung, Taiwan OBJECTIVES: Radiopaque agents induce ADR frequently. But the numbers of ADR reporting are less than actual happening because of heavy workload on medical staffs. We created a simple and convenient online reporting system to replace paper report in our hospital. The aim of this study was to assess the variation of reporting rate. METHODS: In the system, the cases reported by physician including medication used, time of ADR come up and description of ADR event. We downloaded the data of patients who administrated radiopaque agents retrospectively as the base of study group, and observed the reporting rate of online reporting system from 2008 July to 2009 June (i.e., intervention group), comparing it with the paper report in 2007. RESULTS: There are five radiopaque agents with four nonionic and 1 ionic contrast media (e.g., Gadobenate Dimeglumine, Iodixanol, Iohexol, Ioversol and Urografin) used in our hospital. We received 20 ADR reports within 22,754 prescriptions of radiopaque agents from January to December in 2007 and 62 ADR reports in 29,840 prescriptions from 2008 July to 2009 June. The reporting rate elevated 2.3 times (0.09% to 0.21%) when we shifted paper reporting to online reporting system. In the intervention group, more than half ADR reports were rash and no serious side effects (e.g., nephrotoxicity, anaphylactic shock) had been reported. CONCLUSIONS: ADR report is important in patient safety and pharmacovigilance issues. In clinical practice, staff may miss to report ADR because of enormous amount of work load. If the reporting system is more user-friendly, it would be closer to the true incidence of adverse events which have been reported. PCN2 ASSESSMENT ON THE DECREASING USE OF NARCOTIC ANALGESICS BY USING CANCER PAIN ADJUVANTS Chen CH1, Chen KT2, Chen HY3, Hou MF4, Huang YB4, Chan AL1 1 Chi Mei Medical Center, Tainan, Taiwan; 2Chi-Mei Medical Center, Tainan, Taiwan; 3Chi-Mei Medical Center, Liou Ying, Tainan, Taiwan; 4Kaohsiung Medical University Chung-Ho Memorrial Hospital, Kaohsiung, Taiwan OBJECTIVES: Narcotic analgesic was usually used to relieve cancer pain in the past decades. Recently, co-analgesics with adjuvant drugs have become an essential way and successful alternatives for holistic care. The aim of this study is to assess the outcome of using cancer pain adjuvants. METHODS: This is an observational study.

4th Asia-Pacific Abstracts Cancer patients with pain score > 3 were recruited from three medical centers in Taiwan. Numeric Pain Rating Scale, Wong-Baker Face Pain Rating Scale and Brief Pain Inventory were used to evaluate the differences of pain control between two groups, addition of pain adjuvants and narcotics. The monthly consumption of narcotics, the frequency of breakthrough pain and adverse event were recorded for analysis by Paired t-test. RESULTS: A total of 120 patients were eligible in this study. Fifty-five (45.8%) and 65 (54.2%) patients received narcotic analgesics and pain adjuvants for pain control, respectively. The satisfaction of patients in both groups on pain control improved 2.03 ± 1.87 points in the narcotics group and 2.14 ± 1.72 points in concomitant adjuvants group without significant difference. (F1,118 = 0.1024, P = 0.7495) The times of breakthrough pain was significantly decreased in adjuvants group as compared to narcotics group(−1.47 ± 2.23 vs. −0.48 ± 1.80; F1,118 = 7.3556, P = 0.0077). CONCLUSIONS: No significant difference in total consumption of narcotics and pain control was founded. The frequency of breakthrough pain is likely to be improved in the concomitant adjuvants as compared in the narcotics group. The study found the safety of adjuvant therapy do no harm to palliative patients. PCN3 RANDOMIZED, PLACEBO, CONTROLLED, DOUBLE-BLIND TRIAL OF MELATONIN IN CHOLANGIOCARCINOMA PATIENTS Manprasert W1, Johns NP1, Sookprasert A1, Kantpittaya J1, Johns J1, Konsil J1, Weerapreeyakul T2, Phongthai P2, Porasutpathana S1, Laolukana W2 1 Khon Kaen University, Muang, Khon Kaen, Thailand; 2Khon Kaen Hospital, Muang, Khon Kaen, Thailand OBJECTIVES: Cholangiocarcinoma is a leading cause of cancer death in the Northeast of Thailand. Currently, chemotherapy and supportive care are the main treatments in non-resectable cholangiocarcinoma patients. Many adverse events (AE) from disease and treatments have been reported that affect quality of life (QOL) of the patients. This study evaluated the efficacy of melatonin compared to placebo on the QOL, AE and survival time. METHODS: The study was a randomized, double-blind, placebo, controlled trial. The patients received treatment following the protocol of each hospital, and the study drug. Patients were randomized by mixed-block randomization stratified by hospital and type of treatment. The treatment group received melatonin (20 mg/day) and the control group received placebo. Patients started taking the study drug on the first day of the treatment and continued for 3 months. QOL was assessed using Thai Functional Assessment of Cancer Therapy-Hepatobiliary (FACT-HEP) and AE were assessed using Common Toxicity Criteria Adverse Events (CTCAE). RESULTS: There were 30 patients recruited in the study, divided into 15 patients in each group. Baseline characteristics of the two groups were not different. The melatonin group had a higher but not significant percentage of improvement in FACT-Hep scores than the control in both the first (20% vs. 6.7%) and second (20% vs. 6.7%) month. Median survival was longer in the melatonin group (160 vs. 130 days, P > 0.05). In addition, there were fewer reports of Grade 3–5 AE in the melatonin group than placebo in terms of anorexia (11% vs. 50%), fatigue (11% vs. 50%), nausea/vomiting (0% vs. 10%) and weight loss (0% vs. 20%). CONCLUSIONS: The combination of melatonin with standard treatment did not prolong overall survival. However, the melatonin treatment can decrease AE and maintain quality of life of non-resectable cholangiocarcinoma patients. Further studies with larger samples are needed. PCN4 CANCER SYMPTOM SURVEY AND RELIABILITY ANALYSIS Li P Peking University School of Oncology, Beijing, China OBJECTIVES: Through symptom survey in cancer patients to find out the symptom distribution and patients treatment require. Also validate set of symptom items that are most commonly used in Traditional Chinese Medicine (TCM). METHODS: Use M. D. Anderson Symptom Assessment Inventory (MDASI) that previously validated Chinese version. Add 10 TCM items. We also asked the questions about using TCM request in cancer care. A multi center cross-sectional study using a convenience sample of 340 patients were conducted in 3 hospitals in Beijing and Dalian. Statistical use Spss and Excel software. Self evaluate questioner with Cronbach’s Alpha score. RESULTS: The most severe symptoms were fatigue 89.4%, sleeping disturbance 74.4%, moth drier 72.9%, and poor appetite 72.9%, forgotten 71.2%. The influence of cancer symptoms to work, mood and activity were 89.7%, 82.6%, 78.6% separately. The quality of life was affected severely. 80% patients want to regulate the body with TCM. Almost 100% patients hope to know the knowledge of integrated TCM and western medicine, and the effect of TCM treatment cancer and health recover. Cronbach’s Alpha score indicated acceptable internal consistency both the MDASI and TCM items, 0.86 for MDASI, 0.78 for TCM, 0.9 for MDASI_TCM 23 items. CONCLUSIONS: Fatigue, sleep disturbance, moth drier, poor appetite and forgotten are more severe symptoms in our cancer patient’s survey. Greatly affect the quality of life of patients. There are an extremely request that TCM holism regulation in cancer patients. The MDASI and its TCM model could a critical tool to measure the effectiveness of TCM in cancer symptom management.

A511

4th Asia-Pacific Abstracts PCN5 SOFRAFENIB VERSUS SUNITINIB IN METASTATIC RENAL CELL CARCINOMA: INDIRECT COMPARISON ANALYSIS Leung HW1, Chan AL2 1 Taipei Medical University- Shuang Ho Hospital, Taipei, Taiwan; 2Chi Mei Medical Center, Tainan, Taiwan OBJECTIVES: To evaluate the clinical effectiveness of sorafenib and sunitinib in metastatic renal cell carcinoma (RCC) by using indirect comparison meta-analysis. METHODS: Systematic literature search of Medline, Embase, Cochrane controlled trials register. All randomized clinical trials of sorafenib or sunitinib versus interferon alfa for treating metastatic renal-cell carcinoma were included. Study selection, data extraction and quality assessment were performed by two reviewers with disagreements being resolved by consensus. The effects of sorafenib and sunitinib on progression-free survival were compared indirectly using indirect treatment comparison program, with interferon alfa (IFN) as a common comparator. RESULTS: Two studies were included. Median progression-free survival was prolonged with the treatment of sunitinib (11 months) compared to interferon alfa (5 months). For the comparison of sorafenib and interferon-alfa, the median progression-free survival was similar (median PFS: 5.7 months vs. 5.6 months). Indirect comparison suggests that sunitinib is not superior to sorafenib for prolongation of progress free survival (hazard ratio 0.37; 95% CI: 0.236–0.58, P = 0.0189). CONCLUSIONS: There is no significant evidence to suggest that treatment with sunitinib has clinical advantages over treatment with sorafenib in patients with metastatic RCC. PCN6 RISK OF BREAST CANCER AMONG USERS OF POSTMENOPAUSAL HORMONE REPLACEMENT THERAPY IN TAIWAN Shen WC1, Lin MS2, Bai CH3, Chen WC4, Tang CH4 1 Taipei Medical University and Wan Fang Hospital, Taipei, Taiwan; 2National Taiwan University and National Taiwan University Hospital, Taipei, Taiwan; 3Taipei Medical University/Shin Kong WHS Memorial Hospital, Taipei, Taiwan; 4Taipei Medical University, Taipei, Taiwan OBJECTIVES: To determine whether the association between the different dosage of hormone replacement therapy (HRT) and the incidence of breast cancer (BC) in postmenopausal women with HRT formulation. METHODS: Patients who had at least one outpatient visit for postmenopausal syndrome (ICD-9-CM code 627) with estrogen prescription in Taiwan National Health Insurance (NHI) claims database during 1999– 2006 were identified as the study cases. There were 883,052 women identified from the dataset. The index date was defined as the date of the first menopausal visit with estrogen prescription during the study period. To identify any BC events, each case was tracked from the index date until December 31, 2006 or death, whichever came first. Women without events were censored on December 31, 2006. Survival analysis was performed to assess whether cumulative estrogen dosage and combined progesterone were independent risk factors of BC. RESULTS: A total of 5524 cases of BC were identified during the study period. Women with higher dosage of estrogen had significantly higher risk of BC than women with lower dosage (HR = 2.23; P < 0.0001). The risk of BC was even higher when progesterone was combined with estrogen (HR = 1.08; P = 0.036). Women aged 60–69 (HR = 0.87, P = 0.002) and >70 (HR = 0.66, P < 0.0001) had lower risk of BC, compared with women aged = 1 HCC claim (index = first claim), age > = 18, and no other cancer diagnoses. For each procedure code, all payments (2009 USD) were summed across the cohort and divided by the total patientmonths to estimate the per patient-month (PM) cost. The PM costs were ranked to identify procedures with the highest payer budget impact. The proportion of patients utilizing each procedure was calculated to evaluate whether PM costs were driven by a minority of patients. RESULTS: The study sample included 2927 patients: mean age 50.4 years, 57% male, and median 9 months follow-up. The inpatient procedures with the highest budget impact were: liver transplant and/or intestinal transplant ($526/ PM, 4.6% of patients), tracheostomy with mechanical ventilation ($204/PM, 0.2%), disorders of liver except malignancy/cirrhosis/alcoholic hepatitis ($96/PM, 5.1%). For hospital outpatient procedures, they were: magnetic resonance imaging (MRI) of the abdomen ($18/PM, 15%), computed tomography of abdomen ($13/PM, 17%), and transcatheter occlusion/embolization ($8/PM, 4%). For office outpatient procedures, they were: low complexity visits ($26/PM, 70%), moderate complexity visits ($18/ PM, 18%), and MRI of abdomen ($8/PM, 7%). Patients with > = 1 sorafenib prescription ( = 1 HCC claim, > = 1 HCC post-index claim, and no other cancers. Controls were matched on age, sex, and race. Costs (2008 USD) for medical care services (inpatient, outpatient, emergency room [ER], long-term care [home health, nursing home, hospice]) were analyzed as first-year costs and follow-up adjusted costs (per-patientper-month [PPPM]). All costs were compared using rank sum tests. RESULTS: The study identified 126 HCC cases and 126 controls: mean age 49 years, 51% male, mean follow-up months: 9.2 cases/11.9 controls, and deaths: 29.4% cases/1.6% controls. First-year costs were 1.5 to 8 times higher in cases versus controls (long-term care $3752 vs. $2554, ER $126 vs. $51, outpatient $4143 vs. $1412, inpatient $12,425 vs. $1595, all with P < 0.012). PPPM costs were 3 to 27 times higher in cases versus controls (long-term care $626 vs. $233, outpatient $884 vs. $119, ER $34 vs. $4, inpatient $3738 vs. $139, all with P < 0.001). Total costs were two times higher for HCC cases in the first year ($29,795 vs. $13,151, P < 0.001) and six times higher as PPPM ($6376 vs. $1125, P < 0.001). No patients had a sorafenib prescription. CONCLUSIONS: First-year medical care costs were substantially higher for the HCC patients, and even higher when adjusted for follow-up on a PPPM basis. Future therapies that improve survival and disease control may enable payers to reduce monthly costs, and use the savings to treat other patients. PCN10 ESTIMATION OF THE HEALTH AND COST BURDEN OF HPV-RELATED DISEASES IN THAILAND Termrungruanglert W1, Havanond P1, Khemapech N1, Lertmaharit S1, Pongpanich S1, Khorprasert C1, Kitsiripornchai S2, Jirakorbchaipong P2, Taneepanichskul S1 1 Chulalongkorn University, Bangkok, Thailand; 2MSD (Thailand) Ltd., Bangkok, Thailand BACKGROUND: Cervical cancer is the first leading female cancer, especially, in developing countries. The two available HPV vaccines protect against HPV types 16/18 responsible for majority of cervical cancer. The quadrivalent vaccine also protects against HPV types 6/11 that cause majority of genital warts. It is important

A512 to examine health and cost burden of HPV-related diseases to understand the potential impact of introducing HPV vaccine. OBJECTIVES: We aimed to evaluate the longterm health and cost burden associated with cervical cancer, cervical intraepitherial neoplasia (CIN) and genital warts from health-care provider perspective in Thailand. METHODS: We developed a state-transition Markov model to simulate the epidemiology of stages of cervical cancer, CIN and genital warts in a hypothetical cohort of 100,000 12-year-old girls. Costs included diagnosis and treatment costs of HPV related diseases. Probabilities at each chance node in the model were derived from the Thailand health-care context. RESULTS: The highest incidence of CIN and genital warts was observed among women aged 20–30 years. For cervical cancer, the highest incidence was observed among women aged 45–55 years. Death rate was estimated at 2%, 8%, 84%, and 94% for cervical cancer stage IA1, IA2-IIA, IIB-IVA, and IVB, respectively. The estimated mean direct cost per patient with cervical cancer stage IA1, IA2-IIA, IIB-IVA, IVB, CIN1, CIN2/3 and genital warts were US$1277, US$3020, US$12,506, US$10,019, US$167, US$1551, and US$111, respectively. The overall lifetime costs were estimated at US$26.7 million for a cohort of 100,000 women, which corresponded to approximately US$132.0 million for the current entire cohort of 12-year-old girls in Thailand. CONCLUSIONS: HPV- related diseases impose significant health and cost burden in Thailand. The potential impact on HPV-related diseases of a national immunization program with HPV vaccine should be examined to inform the policy discussions around the HPV vaccination program in Thailand. PCN11 PREVENTION OF OXALIPLATIN HYPERSENSITIVITY REACTION AND COST SAVING Huang SF, Chen CH, Chan AL Chi Mei Medical Center, Tainan, Taiwan OBJECTIVES: Hypersensitivity reactions have been reported in Oxaliplatin for the treatment of advanced colorectal cancer (mCRC). The reported incidence of hypersensitivity reactions (HSRs) is approximately 12%, with 1–2% of patients developing grade 3 or 4 in severity which may postpone length of stay and incurred extra cost for the treatment of hypersensitivity reactions. METHODS: This is a retrospective observational study. Medical records of hospitalized patients with mCRC, who treated with FOLFOX regimen and occurred mild to severe HSRs were identified and reviewed by oncology pharmacist from January 2004 to February 2010. Direct medical cost for the treatment of HSRs and the extra days of hospitalization were calculated. We compared the difference of direct medical cost for patients whose oxaliplatin was discontinued (group A) and those oxaliplatin was continued (group B) for treating mCRC after the computer detected patients’ allergy and the signal was appeared on the screen of physician-order-entry-system. Student t test was used for data analysis. RESULTS: A total of 442 patients with the diagnosis of mCRC and had reported moderate to severe HSRs during the treatment of FOLFOX regimen through our computerized decision supporting system (patients allergy history alert system) were included in this study. Twenty-four of 442 patients reported mild to severe hypersensitivity reactions caused by oxaliplatin. The total direct medical costs for 24 patients were NT$864,352 dollars. The direct medical cost to manage HRSs in group A was higher than that in group B. CONCLUSIONS: Oxaliplatin induced severe hypersensitivity reaction is scare, but it may cause an extra direct medical costs if it is appearance. To develop a computerized alert signal is helpful and is likely to save costs. PCN12 COST-EFFECTIVENESS OF THE ONCOTYPE DX® ASSAY IN AUSTRALIA: AN EXPLORATORY ANALYSIS O’Leary B1, Foteff C2, Byron K3, Chang C4, Chao C4, Ng C5, Skrzypczak S4 1 Covance Pty Ltd, North Ryde, NSW, Australia; 2Covance Pty Ltd, North Ryde, NSW, Australia; 3Gribbles Pathology, Clayton, VIC, Australia; 4Genomic Health, Inc., Redwood City, CA, USA; 5Genomic Health, Inc., Hong Kong, HKSAR, Hong Kong OBJECTIVES: Oncotype DX is a molecular diagnostic assay that measures quantitative expression of 21 genes within a breast tumor sample. The result is reported as a recurrence score (RS) that correlates with the risk of 10-year recurrence. The potential cost impact on chemotherapy treatment and an exploratory cost utility analysis were undertaken from the Australian health-care system perspective. METHODS: Input on the proportion of patients treated with chemotherapy and treatment regimens were obtained from an expert panel and a supplementary survey of Australian clinicians (oncologists and surgeons, n = 12). Data on the proportion of patients who would forgo chemotherapy based on knowledge of the RS and the incidence and cost of adverse events were obtained from published literature. RESULTS: The clinician input indicated that 33% of node negative and 84% of node positive women receive adjuvant chemotherapy on average. The most common treatments for node-negative patients were AC (anthracycline and cyclophosphamide) (77%) and FEC100 (fluorouracil, epirubicin and cyclophosphamide) (16%) and for node-positive patients FEC-D (FEC plus docetaxel) (54%), AC + paclitaxel (26%) and TAC (docetaxel plus AC) (14%). Published switch rates away from chemotherapy are 20% for node negative and 24% for node positive patients. The cost saving due to a reduction in chemotherapy was estimated to be A$2264 per woman tested. After consideration for the cost of the assay (A$4200) and a published utility rate of 0.5, a A$/QALY gain was estimated at A$9986. CONCLUSIONS: Knowledge of the Oncotype DX RS has a cost-offset due to the reduction in chemotherapy and is likely to be cost-effective. These benefits reflect the quality of life and survival benefits of a more targeted approach to treatment decision-making. Further analysis is warranted to include the

4th Asia-Pacific Abstracts potential costs of relapse avoided by use of the assay and any patient indirect costs in Australia. PCN13 COST-BENEFIT ANALYSIS OF A 21-GENE RECURRENCE SCORE FOR EARLY STAGE BREAST CANCER IN SINGAPORE de Lima Lopes G1, Chien R2, Hornberger J3 1 Johns Hopkins Singapore International Medical Centre and Johns Hopkins University School of Medicine, Singapore; 2Cedar Associates LLC, Menlo Park, CA, USA; 3Cedar Associates LLC and Stanford University, Menlo Park, CA, USA BACKGROUND: Breast cancer represents a significant burden of illness worldwide. Cost of cancer management is a key policy concern. The adoption of 21-gene breast cancer recurrence score (RS) was projected to be cost-effective. OBJECTIVES: To assess cost-benefit of the 21-gene RS for lymph node negative, estrogen receptor positive early-stage breast cancer (ESBC) in Singapore from a patient’s perspective. METHODS: We adopted a validated Markov model to calculate the cost implications of RS for an early-stage breast cancer patient. The probability of individual’s risk of recurrence, chemotherapy benefits and decision impact of RS were derived from existing studies. The model accounted for both direct and indirect costs associated with adjuvant chemotherapy. Direct costs included chemotherapy drugs, supportive care drugs, administration, and management of adverse events adjusted by incidence, and cost of recurrence. Indirect costs included productivity loss during chemotherapy and distant recurrence. Chemotherapy regimen distribution and costs were obtained from medical oncologists at a private and a public cancer centers in Singapore. RESULTS: The pretest probabilities of risk of recurrence per woman as low, intermediate or high were 49.8%, 30.3%, and 19.9%, respectively. The average direct potential savings per patient tested in Singapore Dollars (SGD) for chemotherapy drug, supportive care, management of AE and administration were $2942, $1077, $169, and $1340, respectively. Per patient saving from productivity loss during treatment was $468. At manufacturer price, the model projects immediate realized savings of $430 with the adoption of the RS. The model also projects $1331 direct savings, and $2418 indirect savings from preventing distant recurrence. Multiple sensitivity analyses demonstrate the robustness of the savings with adoption of RS under a variety of conditions. CONCLUSIONS: For women with ESBC in Singapore, the RS is a cost-saving treatment decision tool. PCN14 ADDITIONAL HEALTH AND ECONOMIC IMPACT OF THE BIVALENT VERSUS THE QUADRIVALENT HPV VACCINE IN TAIWAN: RESULTS OF A PREVALENCE-BASED MODEL Ho T1, Van Enckevort PJ2, Demarteau N3 1 GlaxoSmithKline, Taipei, Taiwan; 2GlaxoSmithKline Biologicals, Singapore; 3GlaxoSmithKline Biologicals, Wavre, Belgium OBJECTIVES: Two human papillomavirus (HPV) vaccines currently exist: a bivalent targeting oncogenic high-risk HPV-16/18 and a quadrivalent targeting high-risk HPV16/18 and low-risk HPV-6/11. Based on data in their respective trials, the bivalent vaccine is likely to have higher efficacy against non-vaccine oncogenic HPV-types (cross protection). The quadrivalent has an additional effect against genital warts. The differences in the annual impact of both vaccines on precancerous lesions, genital warts (GW) and cervical cancer (CC) were evaluated. METHODS: A static prevalencebased model was developed, which estimates the differences in lesions, GW, CC, and costs prevented between the two vaccines over a one-year period at steady state (i.e., when all women are vaccinated). The base case analysis was performed from a healthcare payer’s perspective, including only National Health Insurance covered costs. In addition, an analysis from a societal perspective was performed, including also indirect costs and out-of-pocket payments. Epidemiological and cost data were obtained from published sources. Efficacy figures were based on the latest results from each vaccine’s respective clinical trial. Univariate sensitivity analyses were conducted, in which parameters were varied 20% up and down from the baseline value. RESULTS: Under this model, the bivalent vaccine would result in an additional reduction of 1486 ASCUS; 1768 CIN1; 1230 CIN2/3 and 166 CC cases per year, while the quadrivalent vaccine would result in an additional 7726 GW cases prevented per year. The additional annual costs averted with the bivalent vaccine was estimated at US$ 1.3 million and US$ 1.9 million from a health-care payer’s and societal perspective, respectively. This outcome is most sensitive to cross protection figures. CONCLUSIONS: Vaccination with the bivalent vaccine is expected to have a higher impact than the quadrivalent vaccine on the prevention of precancerous lesions and CC in Taiwan, and would lead to more cost averted. PCN15 COST ANALYSIS BETWEEN TYROSINE KINASE INHIBITOR VERSUS DOCETAXEL IN ADVANCED NON-SMALL CELL LUNG CANCER Topibulpong N, Tanasanvimon S, Parinyanitikul N, Sriuranpong V Chulalongkorn University, Bangkok, Thailand OBJECTIVES: To determine the cost-effectiveness of the second line treatment of advanced non-small cell lung cancer (aNSCLC) with epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) comparing to docetaxel under health-care provider perspective. Only direct medical cost was considered. Outcomes included the patients’ survivals. METHODS: We collected data of aNSCLC patients treated at the Medical Oncology Unit during the year 2009 in a prospective manner. Patients were followed until death. Direct medical costs included medicine cost, laboratory cost,

4th Asia-Pacific Abstracts routine service cost, and cost of adverse event treatment. RESULTS: EGFR-TKI treatment provided a better progression-free survival and overall survival compared with docetaxel. Except rash, dry skin and dry eyes, the incidence of other adverse events in patients treated with EGFR-TKI was lower than docetaxel. However, the monthly treatment cost and total treatment costs of EGFR-TKI were higher than docetaxel at 37,017.55 baht/month and 451,071.08 baht respectively. The incremental cost-effectiveness ratio (ICER) was 631,426.01 baht/life-year-gained (LYG). CONCLUSIONS: Though EGFR-TKI has a more favorable clinical outcome than docetaxel, the additional cost per LYG is significant. This ICER is greater than the threshold suggested by the WHO, three times of GDP/capita (300,000 baht), which may consider to be not cost-effective. Both docetaxel and EGFR-TKI are the effective and available treatments for aNSCLC in Thailand. When making a decision and policy regarding the treatments of aNSCLC, this economic evidence should be taken into account along with other clinical aspects. PCN16 COST-EFFECTIVENESS ANALYSIS OF K-RAS TESTING AND CETUXIMAB FOR METASTATIC COLORECTAL CANCER IN JAPAN Shiroiwa T1, Motoo Y2, Tsutani K3 1 Ritsumeikan University, Kusatsu, Shiga, Japan; 2Kanazawa Medical University, Uchinada, Japan; 3 The University of Tokyo, Tokyo, Japan OBJECTIVES: Cetuximab, a monoclonal antibody directed against the epidermal growth factor receptor, improves progression free survival and overall survival for metastatic colorectal cancer (mCRC) patients. However patients with K-ras mutation don’t benefit from cetuximab. METHODS: We performed cost-effectiveness analysis of K-ras testing and cetuximab treatment as last-line therapy for mCRC patients. In our analysis, we considered three groups: A) cetuximab treatment for patients with wild-type tumors and best supportive care (BSC) for mutant patients with K-ras testing; B) cetuximab treatment for all the patients without K-ras testing; and C) BSC for all the patients without K-ras testing. The cost-effectiveness of three comparison groups was calculated: group A versus B (cost-effectiveness of K-ras testing), group A versus C (cost-effectiveness of cetuximab treatment with K-ras testing), and group B versus C (cost-effectiveness of cetuximab treatment without K-ras testing). Constructed three-state Markov model was used to predict expected costs and outcome of each group. Outcome was based on the report by Karapetis et al. (2008), who retrospectively analyzed NCIC CTG CO. 17. Only direct medical costs were included from the perspective of Japanese health-care payer. Three percent discount rate was used for both costs and outcome. RESULTS: K-ras testing for cetuximab treatment reduced costs by JPY 500,000 per patients. Budget impact was estimated to be JPY 3 to 5 billion. However, the ICER of cetuximab with K-ras testing was more than JPY 10 million per life-year gained (LYG) although it was smaller than ICER of cetuximab without K-ras testing. Even if the mean utility score was 0.7, the ICER was approaching JPY 20 million per QALY. CONCLUSIONS: When the cetuximab treatment is administered as last-line therapy for mCRC patients, K-ras testing is recommended. However, the ICER of cetuximab treatment is too high, even if the treatment is limited to patients with wild-type K-ras. PCN17 ECONOMIC EVALUATION OF PEMETREXED IN PATIENTS WITH PREVIOUSLY TREATED ADVANCED NON-SMALL CELL LUNG CANCER IN JAPAN Kamae I1, Nakahara N2, Moriwaki K3, Enatsu S4, Nambu Y4 1 Keio University Graduate School of Health Management, Fujisawa, Japan; 2Eli Lilly Japan K.K., Kobe, Japan; 3Kobe University Graduate School of Medicine, Kobe, Japan; 4Eli Lilly K.K. Japan, Kobe, Japan OBJECTIVES: To elucidate the more cost-effective dosing strategy for the administration of pemetrexed in the second-line treatment of patients having been treated for advanced non-small cell lung cancer (NSCLC). METHODS: Validation of the model was based on efficacy and safety data from a phase II clinical study conducted in Japan of two pemetrexed doses: 500 mg/m2 (P500) versus 1000 mg/m2 (P1000). Cost-effectiveness and cost-minimization analyses were performed to estimate the impact of P500 versus P1000 on expected life-years and costs from the perspective of the Japanese national health insurance (NHI) system. Costs covered additional treatment and test which are needed due to the treatment of pemetrexed and adverse events. A subgroup analysis was conducted to compare the cost-effectiveness of pemetrexed in patients with different NSCLC histologies. RESULTS: P1000 exhibited a negative life-years gained (LYG) of −0.115 at an incremental cost of US$13,674 compared to P500, implying simple dominance of P500 over P1000. The incremental cost-effectiveness ratio (ICER) of pemetrexed therapy with P500 was estimated at US$19,479/ LYG, compared with the same therapy in squamous cell carcinoma, when it is performed in non-squamous cell carcinoma. Results of a probabilistic sensitivity analysis demonstrated an almost 100% likelihood of the ICER of pemetrexed with P500 for non-squamous cell patients falling below the threshold of US$50,000/LYG. CONCLUSIONS: P500 is a more cost-effective dosing regimen than P1000 for pemetrexed chemotherapy in patients with NSCLC. Also, results suggest that pemetrexed is more cost-effective in NSCLC patients of non-squamous histology than of squamous histology from the payer’s perspective.

A513 PCN18 ECONOMIC IMPLICATIONS OF THE FIRST-LINE TREATMENT OF ADVANCED RENAL CELL CARCINOMA IN THAILAND: A COSTEFFECTIVENESS ANALYSIS Topibulpong N, Tanasanvimon S, Parinyanitikul N, Vinayanuwattikun C, Sriuranpong V Chulalongkorn University, Bangkok, Thailand OBJECTIVES: To determine the economic implications of agents used to treat metastatic renal cell carcinoma (mRCC). METHODS: A lifetime Markov model was developed to simulate the patterns of disease progression and to determine the outcomes under treatment of available medicines including interferon-alpha (IFNa), sunitinib, sorafenib, and bevacizumab. Costs were measured based on the perspective of a health-care provider. Resource utilization derived from retrospective chart reviews of 14 mRCC patients treated at the King Chulalongkorn Memorial Hospital, Bangkok, Thailand. Costs and survival benefits were discounted annually at 3%. RESULTS: The major assumption of the study was that the projected PFS and overall survival were longer for sunitinib than the other comparative treatments based on results of published studies. Additional treatment after failure of one tyrosine kinase inhibitor (TKI) in the model was limited which fit well with the practical scenario in developing country. Cost-effectiveness analyses demonstrated that IFNa had the most favorable cost-effectiveness ratio at 1.3 million Baht/QALYs followed by sunitinib, sorafenib, and bevacizumab at 1.7, 1.9, 9.3 million Baht/QALYs respectively. However, sunitinib had a better incremental cost-effectiveness ratios (ICERs) over IFNa at 2.0 million Baht/PFY, 2.7 million Baht/LYG, and 3.6 million Baht/QALY gained. Sensitivity analysis showed that the most sensitive parameters for sunitinib were the overall survival efficacy and costs. CONCLUSIONS: Our study reveals that sunitinib has the good cost-effectiveness profiles as one of the first line treatment of mRCC. With current study model, it may be applicable to the situation in developing country with limit availability of TKIs in the treatment of mRCC. However, sunitinib’s ICER comparing to IFNa is still beyond the ICER threshold for a developing country. This cost-effectiveness evidence should be taken into account in conjunction with other clinical, societal and ethical considerations in the treatment of mRCC. PCN20 COST-EFFECTIVENESS ANALYSIS OF A FOBT-BASED COLORECTAL CANCER SCREENING PROGRAM Pizzo E1, Bracci E2, Vagnoni E2, Wilschut J3, van Ballegooijen M3 1 Imperial College London, London, UK; 2University of Ferrara, Italy, Ferrara, Italy; 3Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands OBJECTIVES: Colorectal cancer (CRC) is one of the most common forms of cancer in western countries and represents the second leading cause of cancer mortality in Europe (AIRTUM, 2009). Early detection and removal of cancerous lesions can reduce the incidence of CRC, its mortality and improve patients’ quality of life (Taupin et al. 2006). The main literature on this topic refers to USA and few studies have been conducted in Italy to date (Zappa et al. 1997; Tappenden et al. 2007). Aim of the paper is to shed some light on the effectiveness and costs of screening programs in the Italian health-care system, presenting the results of a cost-effectiveness analysis of a CRC screening program in Italy. METHODS: We use as case-study a Regional CRC screening program to determine the full costs and the effectiveness of the adopted techniques, FOBT combined with colonoscopy. The costs involved in each phase of the program are evaluated using a micro-costing analysis. Effectiveness is valued in terms of early detected lesions and years of life gained. Cost and effectiveness data are used to estimate the costs for year of life gained, using a MISCAN-COLON Model© to simulate and compare two alternative scenarios, with or without the screening program. RESULTS: The preliminary results show that the screening will prevent almost 2.0 deaths (11.2%) per 1000 screened individuals, corresponding to 19.4 years of life gained in 30 years with an incremental cost effectiveness ratio of f2400 for life-year gained. CONCLUSIONS: The results outpace those of previous studies (Sonnenberg, 2000), signalling an increasing effectiveness of CRC screening program. Besides, the paper highlights the importance of implementing a screening not only for the effects that prevention can have in clinical terms, but also for the economic impact of such a policy in relation to the long-term sustainability of healthcare systems. PCN21 COST-EFFECTIVENESS ANALYSIS OF ADJUVANT HORMONAL TREATMENTS FOR POSTMENOPAUSAL HORMONE-RECEPTOR POSITIVE EARLY BREAST CANCER IN KOREAN CONTEXT Lee HJ, Lee TJ, Yang BM Seoul National University, Seoul, South Korea OBJECTIVES: This study aims to evaluate the cost-effectiveness of two aromatase inhibitors for adjuvant treatment of postmenopausal hormone receptor positive early breast cancer, and to address the most reasonable treatment option when population is stratified by nodal status. METHODS: A Markov model was developed defining six Markov states based on breast cancer progression. Annual probabilities of recurrence by adjuvant treatment, anastrozole, letrozole, and tamoxifen were estimated from published studies in overall population, node negative, and node positive group. Costs of defined breast cancer events were measured by micro-costing methods based on 2009 National Health Insurance Fee Schedule and the third Clinical Guideline of Breast Cancer Treatment. Anastrozole and letrozole were compared with tamoxifen, respectively, using the same Markov model. Incremental cost-effectiveness ratios (ICERs) in overall population and each subgroup were estimated. RESULTS: When

A514 anastrozole was compared with tamoxifen, anastrozole was more effective and costly than tamoxifen costing additional KRW 22,461,689 per QALY. Letrozole showed similar incremental cost of KRW 21,004,142 per QALY. In the node negative group, anastrozole was the most cost-effective with incremental cost of KRW 19,717,770 per QALY, while letrozole was the most cost-effective with incremental cost of KRW 8,150,512 per QALY in node positive group. Sensitive analysis showed that these results were robust. CONCLUSIONS: Subgroup analysis clearly demonstrated which treatment was superior among aromatase inhibitors. Such a demonstration was not confirmative in the cases of overall population. The implication of this study is that the decision maker should be careful when generalizing the cost-effectiveness results. The stratified analysis in this context may help reach a reasonable decision on resource allocation. PCN22 ECONOMIC ANALYSIS OF CAPECITABINE PLUS OXALIPLATIN (XELOX) VERSUS FLUOROURACIL/LEUCOVORIN PLUS OXALIPLATIN (FOLFOX) IN THE TREATMENT OF ADVANCED GASTRIC CANCER IN CHINA Chen W Fudan University, Shanghai, China OBJECTIVES: The objective of the study was to examine the direct medical cost of XELOX (capecitabine plus oxaliplatin) compared to FOLFOX (fluorouracil/leucovorin plus oxaliplatin) for the treatment of advanced gastric cancer in China. METHODS: Since the equal efficacy was already demonstrated by the published literature and local clinical guideline, cost minimization analysis was performed to compare the direct medical costs of XELOX and FOLFOX for the treatment of advanced gastric cancer. The direct medical costs were associated with the drug costs, drug administration costs, hospitalization costs and adverse events management costs. The costs were calculated based on a questionnaire survey from a clinical expert panel of 20 gastrointestinal surgeons and medical oncologists. RESULTS: According to the result of clinical expert panel survey, the median treatment duration of XELOX and FOLFOX was six cycles and nine cycles, respectively. The drug cost of XELOX regimen was CNY 33,948 (US$4992), higher than FOLFOX by CNY 14,160 (US$2082). However, the cost increment of XELOX regimen was offset by the higher drug administration cost (deviation CNY 9985), hospitalization cost (deviation CNY 4734) and adverse events management cost (deviation CNY 1828) of FOLFOX regimen. As a result, XELOX showed a significant overall cost savings of CNY 2386 (US$351) compared with FOLFOX. CONCLUSIONS: According to the study, XELOX is cost saving in comparison with FOLFOX for the treatment of advanced gastric cancer in China, especially in the chemotherapy administration and hospitalization utilization. PCN23 ECONOMIC EVALUATION ON LIQUID-BASED CYTOLOGY IN THE CERVICAL CANCER SCREENING PROGRAM IN TAIWAN Chow IHI1, You SL2, Pwu RF3, Tang CH1 1 Taipei Medical University, Taipei, Taiwan; 2Academia Sinica, Taipei, Taiwan; 3Division of Health Technology Assessment, Taipei, Taiwan OBJECTIVES: To estimate the clinical and economic effect of Liquid-Based Cytology (LBC) compared with Conventional Papanicolaou Smear (CP) screening program in Taiwan. METHODS: A decision analytic model was used to simulate the natural history of cervical cancer in Taiwan health-care system. This study adopts a Ministry of Health perspective in cost-effectiveness analysis to compare a CP strategy every year with eight different screening strategies. These strategies comprise three screening tools (CP, LBC with Filter Base Technology, and LBC with Cell Enrichment Technology), and three screening intervals (annually, every 3 years, and every 5 years). Outcomes are life expectancy, quality-adjusted life-years (QALYs), lifetime costs, and incremental cost-effectiveness ratios (ICER). Future costs and QALYs were discounted at an annual rate of 3%. One-way sensitivity analyses are conducted to assess parameter uncertainty. RESULTS: Compared with the current screening strategy (an annual CP), LBC with Filter Base Technology and LBC with Cell Enrichment Technology strategies every 3 or 5 years are less costly while the QALYs lost are trivial. When three times GDP per capita is used as the decision threshold, switching cervical cancer screening tools from Pap annually to LBC with Filter Base Technology annually strategy or LBC with Cell Enrichment Technology annually strategy are cost-effective, with ICER equaling NT$233,000 or NT$272,000 per QALY gained, respectively. CONCLUSIONS: Results from this study showed that it may be desirable to adopt LBC as primary screening tool and may to extend screening interval annually to every 3 or 5 years in cervical cancer prevention. PCN24 A POPULATION-BASED DECISION ANALYTIC MODEL FOR ASSESSING THE HUMAN PAPILLOMAVIRUS VACCINATION PROGRAM Liao CH1, Pwu RF2, Chow IHI3, Tang CH3, You SL4, Chen CA5, Tarn YH6 1 National Taiwan University, Taipei, Taiwan; 2Division of Health Technology Assessment, Taipei, Taiwan; 3Taipei Medical University, Taipei, Taiwan, 4Academia Sinica, Taipei, Taiwan; 5National Taiwan University Hospital, Taipei, Taiwan; 6Center for Drug Evaluation, Taipei, Taiwan OBJECTIVES: Sexually-transmitted Human PapillomaVirus (HPV) is the major cause of cervical cancer. For its greatest benefit, HPV vaccine is recommended for young girls before there is a possible exposure to HPV through sexual contacts. To assessing the potential benefits of HPV vaccination program, it is crucial to collect the comprehensive information of infectious status and sexual behavior among adolescents. This study aims to implement the simulation technique to overcome the information barrier

4th Asia-Pacific Abstracts in Taiwan. METHODS: A population-based decision analytic model was established to simulate the health histories of 15 birth cohorts till they are aged 12 to 26. In the starting-point of simulation, these cohorts are aggregated to form a fully representative population in 2009. To capture the potential effects of catch-up vaccination programs, we computed the initial distribution of health states among each cohort by Markov model with cervical cancer natural history in Taiwanese women. We further validated the model by comparing the actual epidemiologic data and simulated results. The model was applied to estimate the incremental cost-effectiveness ratio (ICER) of the different vaccination programs from the perspective of health care. RESULTS: The simulated age patterns of HPV prevalence was corresponding to the observed age patterns of sexual experience. Based on the model, we found that inclusion of a HPV vaccination program for the cohort aged 12 can be considered as cost-effective (ICER value 1,010,000 NTD/QALY) comparing with annual Pap smear program only. Catch-up vaccination program for the cohorts aged 12–26 can be considered costeffective (ICER value 570,000 to 1,300,000 NTD/QALY) comparing with no intervention. Results of sensitivity analysis suggest the robustness of the analysis. CONCLUSIONS: We have developed a model incorporating the local health and medical conditions. The population-based model is helpful for assessing a complicated public policy involved several generations. It provides plentiful information for estimating the benefits and budget impact of public policy. PCN25 COST-EFFECTIVENESS ANALYSIS OF ERLOTINIB VERSUS DOCETAXEL AS A SECOND- OR THIRD-LINE TREATMENT OF NON-SMALL CELL LUNG CANCER IN KOREA Kim YG1, Lee EK2 1 SookMyung Women’s University, Seoul, South Korea; 2Sookmyung Women’s University, Seoul, South Korea OBJECTIVES: Lung cancer has a high mortality and is associated with a substantial financial burden. Cost-effectiveness estimated from foreign studies are usually not directly comparable between countries without adapting by applying the appropriate national value set. This study was performed to evaluate the cost-effectiveness of erlotinib versus docetaxel in patients with advanced non small cell lung cancer(NSCLC) who have failed previous chemotherapy in Korea. METHODS: A Markov model simulated to access the clinical and economic impact of erlotinib or docetaxel over 2 years from society perspective. Progression free survival(PFS) and overall survival(OS) data were derived from clinical trials, and utility/disutility data were collected in published study. Life-year gained(LYG) is projected based on clinical outcomes and converted to quality-adjusted life-years (QALYs) by utility weight. Both direct medical costs (e.g., drugs, visits, monitoring, etc. ) and nonmedical costs were calculated. Costs and outcomes were discounted at an annual rate of 5% and incremental cost-effectiveness ratio (ICER) consists of the difference in cost divided by the difference in QALYs. We performed sensitivity analysis to evaluate uncertainty in the results. RESULTS: After 2 years follow-up, the total costs per patient was lower with erlotinib(11,234,900KRW, 1USD = 1156.53KRW as of February 2010) than with docetaxel (12,105,719KRW). More QALY per patient would be obtained with erlotinib than with docetaxel (0.268 and 0.213, respectively). Thus, ICER for erlotinib compared to Docetaxel is −15,829,756KRW per QALY. The results of the sensitivity analysis showed no significant difference. CONCLUSIONS: This study suggests, with its underlying assumptions and data, the use of erlotinib as second- or third-line treatment for advanced NSCLC would not only save costs but also improve outcomes compared with docetaxel in Korea. PCN26 PHARMACOECONOMIC EVALUATION OF NILOTINIB IN TREATING TAIWAN PATIENTS WITH CHRONIC MYELOID LEUKEMIA (CML) Ko BS1, Tang JL1, Kuo MC2, Shih LY2 1 National Taiwan University Hospital, Taipei, Taiwan; 2Chang Gung Memorial Hospital, Taipei, Taiwan BACKGROUND: CML accounts for 15% of adult leukemia, with incidence ~1 per 100,000. Imatinib mesylate is an oral medication for chronic phase (CP), accelerated phase, or blast crisis phase of CML. For CML patients showing resistance to imatinib, dose escalation is a treatment option. Nilotinib is a second generation tyrosine kinase inhibitor (TKI), rationally designed to preferentially target BCR-ABL, and has demonstrated efficacy and safety in patients resistant or intolerant to imatinib. OBJECTIVES: To estimate the total treatment costs and quality-adjusted life-years (QALY) gained for patients receiving high-dose imatinib (HDI) ≥600 mg/day because of resistance or nilotinib 800 mg/day using a Markov model to project long-term outcome. METHODS: Model parameters were obtained from medical records of patients treated with HDI in Taiwan, along with published clinical trial data for HDI and nilotinib, and local drug and treatment costs. RESULTS: At the end of July, 2008, 45 patients who have received imatinib ≥600 mg/day were included in the study and their charts were reviewed. During the entire treatment period of HDI and nilotinib, anemia was the most common treatment-related adverse event (22.2% vs. 15.6%). The model projected longer expected life-years gained by patients receiving nilotinib therapy (11.72 years vs. 9.31 years), and patients treated with nilotinib are expected to experience longer QALY than HDI (9.6 years vs. 7.47 years). Estimated cost per QALY was 1,189,150 (NTD/year) for nilotinib and 1,169,547 (NTD/year) for HDI. CONCLUSIONS: Under the assumption of similar drug price for nilotinib and HD imatinib, the benefit of using nilotinib is obvious for the better clinical outcome achieved in terms of lower adverse event rate, better QALY, with acceptable economic impact.

4th Asia-Pacific Abstracts The decision of moving therapy from HD imatinib to nilotinib in CML patients with resistance to standard dose imatinib is a reasonable option. PCN27 THE COST OF CHEMOTHERAPY SIDE EFFECTS Pearce AM, Haas M, van Gool K, on behalf of the EMCaP Investigators University of Technology, Sydney, Ultimo, NSW, Australia OBJECTIVES: While many new cancer treatments are expensive, their improved side effect profile may reduce overall costs to the health-care system compared to traditional chemotherapy. The objective of this work was to identify a benchmark approach to the estimation of costs for chemotherapy side effects, which could then be incorporated into estimates of overall chemotherapy costs. METHODS: A systematic review of clinical and economic literature was undertaken using multiple databases and search terms for chemotherapy, adverse events, and cost. Eligible articles were reviewed for methods, quality and cost outcomes. Results were combined using descriptive techniques to identify best practice methodologies for estimating resource use and costs associated with managing chemotherapy side effects. RESULTS: From 4985 titles, 59 eligible articles were identified. Studies were divided into model based analysis (n = 33) and empirical studies (n = 26). The perspectives, approaches, and methods for estimating quantities and value of resources consumed were highly variable. These disparate methods contributed to highly divergent estimates of cost when specific side effects were examined (eg $446 to $17,889 for neutropenia, 2009 AUD). A clear benchmark approach to the economic analysis of chemotherapy side effects was not available. Thus, general methodological recommendations were made to guide development of reliable, valid and transparent estimates of the costs of chemotherapy side effects, based on best practice modeling techniques. The recommendations address issues such as the use of 1) the health-care system perspective; 2) clinical guidelines and patterns of care studies to identify resources; 3) modeling and empirical methods to obtain accurate costing information; and 4) appropriate modeling to allow incorporation into broader economic evaluations. CONCLUSIONS: These recommendations provide a framework for developing cost estimates for specific adverse events. The EMCaP Project will utilize this framework to incorporate estimates of adverse event costs into broader economic evaluations of alternative chemotherapy protocols.

CANCER – Patient-Reported Outcomes Studies PCN28 TRANSLATION AND VALIDATION OF EORTC QLQ–C30 INTO INDONESIAN VERSION FOR CANCER PATIENTS IN INDONESIA Perwitasari DA1, Atthobari J2, Dwiprahasto I2, Hakimi M2, Gelderblom H3, Nortier JWR3, Guchelaar HJ3, Kaptein AA3 1 Ahmad Dahlan University, Yogyakarta, Indonesia; 2Gadjah Mada University, Yogyakarta, Indonesia; 3Leiden University Medical Center, Leiden, The Netherlands OBJECTIVES: To validate the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire—C30 (EORTC QLQ-C30) in an Indonesian version. The standard procedure of forward-backward translation was adhered to in the translation procedures, pilot testing, review, and reliability and construct validity evaluation. METHODS: Data were collected from cancer patients in the Oncology Department of Dr. Sardjito Hospital, Yogyakarta, Indonesia, who were treated with a cisplatin at the dosage ≥50 mg/m2 as monotherapy or in combinations. We used the Short Form-36 (SF-36) to assess the construct validity of our translated questionnaire. RESULTS: About 85% patients (n = 17) were able to fill out the questionnaires by themselves and there were no missing responses to the questions in the pilot study. One hundred twenty-eight patients with different cancer diagnoses were recruited in the validation process from March 2009 to November 2009. The internal consistency was above 0.70 for the SF-36 scales physical function, social function, pain, physical role, and emotional role; values >0.70 were observed in the Indonesian version of EORTC QLQ-C30 scales. Floor and ceiling effects were seen more evidently in the physical, social dimensions, pain and fatigue than in single items. Both of the questionnaires showed that different diagnoses were associated with a similar impact of quality of life in sensitivity tests. Correlations between Indonesian version of EORTC QLQ-C30 and SF-36 dimensions were moderate: between 0.18 and 0.48 for the five domains. The discriminant validity between the two questionnaires was acceptable (between 0.02–0.47). CONCLUSIONS: The Indonesian version of the EORTC QLQ C-30 can be used as a questionnaire to assess quality of life in Indonesian cancer patients. PCN30 ESTIMATING WOMEN’S PREFERENCES FOR CERVICAL CANCER SCREENING PROGRAMS USING CONJOINT ANALYSIS Cheng YS1, Jen LC2, Tang CH3, Huang CH3 1 Taipei Medical University Hospital, Taipei, Taiwan; 2National Taiwan University, Taipei, Taiwan; 3 Taipei Medical University, Taipei, Taiwan OBJECTIVES: Compared with the 3- to 5-year coverage rates of 75% in many European countries, the 3-year coverage rate of 51.5% for Pap smear screening in Taiwan has not achieved a satisfactory level. Previous studies have showed that patients’ preferences were important information for improving the participation rate of cancer screening. This study aimed to understand women’s preferences for cervical

A515 cancer (CC) screening tests with various features. METHODS: Face-to-face patient interviews were carried out with 212 women aged 30–55 at the gynecological outpatient clinics in a regional hospital in Taipei, Taiwan, during January and April 2009. The participants were asked to rate 12 pairs of hypothetical test scenarios comprised of different combinations of six attributes including accuracy of the test, price, screening administers, place of screening and frequency of screening. The relative importance of test attributes was calculated and a cluster analysis was done to identify subgroup with differential preferences. RESULTS: Accuracy, price, and place of screening (P < 0.001) were found to be the most important attributes in determining women’s preferences. Cluster analysis revealed no significant differences in subgroup of women with different health-promoting lifestyles. CONCLUSIONS: The most preferred screening programs was “not screening at home,” “high accuracy of test,” and “low price” among women residing in a metropolitan city like Taipei. Future direction of study on preferences for CC screening should be targeted at women with low utilization of Paps smear screening to better inform the decision-makers in designing more effective CC prevention program. PCN31 USING DISCRETE CHOICE EXPERIMENTS TO ESTIMATE THE MARGINAL WILLINGNESS TO PAY OF INSURANCE PREMIUM FOR LUNG CANCER TREATMENT IN KOREA Jo C Hallym University, Chuncheon, South Korea OBJECTIVES: This study, using DCE method, figures out the characteristics of the decision-making for lung cancer treatment and investigates the attributes affecting the respondents’ choice. Also it ascertains marginal willingness to pay and relative preferences for lung cancer treatment among the general population of 600 respondents in Korea. METHODS: In the survey, the respondents are asked to choose more than one discrete choice option, resulting in multiple observations for each individual. For each pair-wise comparison of choice set, the respondent will make a choice among three alternatives; A, B, or opt out. Thus, the nested-logit model using full maximum likelihood allows us to empirically estimate multi-levels of dependent variables. For the robustness check of our empirical results, we try considering the nature of distribution of error terms in the utility function in several ways. The survey questionnaire includes four attributes associated with lung cancer in Korea (incidence rates, survival rates, treatment costs, and monthly insurance premium), socio-economic status, antecedent variables, and questions regarding risk averseness and subjective health evaluation. RESULTS: The estimates of MWTP between survival rate and monthly insurance premium and MTWP between total treatment costs for lung cancer and monthly insurance premium, by employing “Hybrid Conditional Fixed Effects Logit Model” to figure out the existence of heterogeneity of any observed and unobserved components, are reflecting ranges of 261 KRW~286 KRW and 3414 KRW~4538 KRW, respectively. CONCLUSIONS: Compared to female counterparts, male respondents have higher MWTP of monthly insurance premium for two related attributes. Besides, married and younger respondents, with higher income and education have more MWTP compared to their respective counterparts. One interesting point is that dependents’ MWTP of survival rate is higher than that of insurance premium payers although MTWP of total treatment costs for dependents is significantly lower than premium payers’ MWTP. PCN32 USING DISCRETE CHOICE EXPERIMENTS TO ESTIMATE THE MARGINAL WILLINGNESS TO PAY OF INSURANCE PREMIUM FOR STOMACH CANCER TREATMENT IN KOREA Jo C Hallym University, Chuncheon, South Korea OBJECTIVES: This study, using DCE method, figures out the characteristics of the decision-making for stomach cancer treatment and investigates the attributes affecting the respondents’ choice. Also it ascertains marginal willingness to pay and relative preferences for stomach cancer treatment among the general population of 600 respondents in Korea. METHODS: In the survey, the respondents are asked to choose more than one discrete choice option, resulting in multiple observations for each individual. For each pair-wise comparison of choice set, the respondent will make a choice among three alternatives; A, B, or opt out. Thus, the nested-logit model using full maximum likelihood allows us to empirically estimate multi-levels of dependent variables. For the robustness check of our empirical results, we try considering the nature of distribution of error terms in the utility function in several ways. The survey questionnaire includes four attributes associated with stomach cancer in Korea (incidence rates, survival rates in 5 years after treatment, total treatment costs, and monthly insurance premium), socio-economic status, antecedent variables, along with questions regarding risk averseness and subjective health evaluation. RESULTS: The estimates of MWTP between survival rate and monthly insurance premium and MTWP between total treatment costs and monthly insurance premium, by employing “Hybrid Conditional Fixed Effects Logit Model” to figure out the existence of heterogeneity of any observed and unobserved components, are reflecting reasonable range of 176 KRW~194 KRW and 5408 KRW~6945 KRW, respectively. CONCLUSIONS: Compared to female counterparts, male respondents have higher MWTP of monthly insurance premium for two related attributes. Besides, currently married respondents, with higher income, and higher educational attainments have more MWTP compared to their respective counterparts. One interesting point is that dependents’ MWTP is higher than that of insurance premium payers even after controlling for any other variables.

A516 PCN33 USING DISCRETE CHOICE EXPERIMENTS TO ESTIMATE THE WILLINGNESS TO PAY FOR CANCER TREATMENT IN KOREA: A GENERAL POPULATION STUDY Jo C Hallym University, Chuncheon, South Korea OBJECTIVES: This study, using DCE method, figures out the characteristics of the decision-making for cancer treatment and investigates the attributes affecting the respondents’ choice. Also it ascertains marginal willingness to pay and relative preferences for cancer treatment among the general population. METHODS: In the survey, the respondents are asked to choose more than one discrete choice option, resulting in multiple observations for each individual. For each pair-wise comparison of choice set, the respondent will make a choice among three alternatives; A, B, or opt out. Thus, the nested-logit model using full maximum likelihood allows us to empirically estimate multi-levels of dependent variables. For the robustness check of our empirical results, we try considering the nature of distribution of error terms in the utility function in several ways. The key assumption of logit models is the independence of irrelevant alternatives (IIA), which results from the assumption that other alternatives are independent. The validity of IIA assumption can be simply conducted by a Hausman-McFadden test. We also consider an alternative specification where error terms are independently, but not identically distributed. And finally stand errors are calculated by using bootstrapping and compared to the previous results. RESULTS: In this DCE study, evidence of reliability was found at both input and output level. The estimates of MWTP between survival rate and monthly insurance premium and MTWP between total treatment costs and monthly insurance premium, by employing “Hybrid Conditional Fixed Effects Logit Model” to figure out the existence of heterogeneity of any observed and unobserved components, are reflecting reasonable range of 817 KRW~1324KRW, and 23,690 KRW~38,139 KRW, respectively. CONCLUSIONS: Compared to female counterparts, male respondents have higher MWTP of monthly insurance premium for two related attributes. Besides, currently married respondents, with higher income, and higher educational attainments have more MWTP compared to their respective counterparts. PCN34 USING DISCRETE CHOICE EXPERIMENTS TO ESTIMATE THE MARGINAL WILLINGNESS TO PAY OF INSURANCE PREMIUM FOR LIVER CANCER TREATMENT IN KOREA Jo C Hallym University, Chuncheon, South Korea OBJECTIVES: This study, using DCE method, figures out the characteristics of the decision-making for liver cancer treatment and investigates the attributes affecting the respondents’ choice. Also it ascertains marginal willingness to pay and relative preferences for liver cancer treatment among the general population of 600 respondents in Korea. METHODS: In the survey, the respondents are asked to choose more than one discrete choice option, resulting in multiple observations for each individual. For each pair-wise comparison of choice set, the respondent will make a choice among three alternatives; A, B, or opt out. Thus, the nested-logit model using full maximum likelihood allows us to empirically estimate multi-levels of dependent variables. For the robustness check of our empirical results, we try considering the nature of distribution of error terms in the utility function in several ways. The survey questionnaire includes four attributes associated with liver cancer in Korea (incidence rates, survival rates, treatment costs, and monthly insurance premium), socio-economic status, antecedent variables, and questions regarding risk averseness and subjective health evaluation. RESULTS: The estimates of MWTP between survival rate and monthly insurance premium and MTWP between total treatment costs for liver cancer and monthly insurance premium, by employing “Hybrid Conditional Fixed Effects Logit Model” to figure out the existence of heterogeneity of any observed and unobserved components, are reflecting ranges of 212 KRW~294 KRW and 3458 KRW~4777 KRW, respectively. CONCLUSIONS: Compared to female counterparts, male respondents have higher MWTP of monthly insurance premium for two related attributes. Besides, married respondents, with higher education have more MWTP compared to their respective counterparts. One interesting point is that MWTP of high income group is lower than that of low income counterparts. This might be related to a lower alcohol consumption of high incomers who are less prone to liver cancer compared to low incomers.

CANCER – Health Care Use & Policy Studies PCN35 PRIMARY HEPATOCELLULAR CARCINOMA PATIENTS’ REPEATED MEDICAL UTILIZATION AND EXPENDITURE FOR CONFIRMATION IN TAIWAN Huang HH, Yang MC National Taiwan University, Taipei, Taiwan OBJECTIVES: Hepatocellular carcinoma is one of the most prevalent cancers in Taiwan. Since Taiwan’s National Health Insurance (NHI) system neither prohibit patients from choosing health-care facilities nor limit number of outpatient visits. Therefore, patients being diagnosed as having hepatocellular carcinoma may seek second or even third doctor to confirm their condition. Thus, the purpose of this study was to examine the effects of disease confirmation behavior on the medical utilization

4th Asia-Pacific Abstracts and expenditure for patients with hepatocellular carcinoma. METHODS: Data came the Longitudinal Health Insurance Database, which consisted of one million subjects of randomly selected samples from the entire NHI enrollee profile for 2005 to 2007. We selected patients with hepatocellular carcinoma by using the ICD-9-CM codes (155, 155.0, and 155.2). Medical utilization and expenditure of those patients within 3 months after first being diagnosed as having hepatocellular carcinoma were extracted from the data set. RESULTS: There were 1282 new hepatocellular carcinoma patients in our sample and 258 patients (20.12%) were identified as having the behavior of repeated disease confirmation. Total repeated physician fee and repeated examination were NT$1,272,765 and NT$3,783,983, respectively. (US$1 = NT$31.86) The expenditure of examination in the first-repeated confirmation hospi Consultant in Medical Oncology, and Assistant Director for Clinical Research, Johns Hopkins Singapore International Medical Centre, Singapore tal was higher than that of first-diagnosis hospital (P < 0.01). The expenditure was highly related to having Charlson comorbidity index score 2 or higher, received chemotherapy, embolization, surgical operation and confirmation in medical centers. CONCLUSIONS: Within 3 months after first time being diagnosed, 20.1% of hepatocellular carcinoma patients used repeated medical services to reconfirm their disease. The expenditure of examination in the repeated visit was higher than that of the first time being diagnosed. Further investigation is needed to compare the long-term health status of patients with and without repeated confirmation behavior. PCN36 DRUG UTILIZATION AND PRICE TRENDS OF ORAL-CHEMOTHERAPY MEDICATIONS IN MEDICAID 1991–2008 Bian B, Guo JJ University of Cincinnati College of Pharmacy, Cincinnati, OH, USA OBJECTIVES: Numerous oral chemotherapies have been approved and widely used for the treatment of different cancers. The objective of this study is to examine drug utilization and price trends for oral chemotherapies in US Medicaid programs. METHODS: A total of 18 oral anti-cancer drugs were included in this research. A retrospective, descriptive time-series analysis was performed using the National Medicaid pharmacy claims database from 1991 to 2008. The quarterly prescription numbers and reimbursement amounts were calculated over time by summing data for individual drug products. The quarterly per-prescription reimbursement as a proxy for drug price was computed for the price of each oral chemotherapy drug. RESULTS: Total oral chemotherapy prescriptions increased from 176,000 in 1991 to 259,000 in 2004. For individual drugs, the two top-sell drugs are Vesanoid® and Hydrea® in terms of prescription numbers. In 2006 the number of prescriptions dropped to 164,535 due to the implementation of Medicare Part D, and increased to 188,689 in 2008. Expenditure for oral chemotherapy drugs increased from $9.3 million in 1991 to $112.3 million in 2005 then dropped to $100 million in 2008. Medicaid expenditure for oral anti-cancer drugs reached to the peak in 2005 quarter 4, which was almost $29 million, then the expenditure dropped to $26.5 million in 2008 quarter 4. Average oral chemotherapy drug payment per prescription also increased dramatically from $1269 in 1991 to $14,297 in 2008. The price of newer oral chemotherapies is more expensive than older ones, and increased over time regardless other brands’ competitions. CONCLUSIONS: Increased utilization of oral chemotherapy drugs was paralleled with its increased expenditure, which may be due to their efficacy and safety profiles. Increased price for oral chemotherapies may be due to biomarkers specifications with limited patient pool or the shrink demand. PCN37 GAUGING THE ROLE OF HTA IN REIMBURSEMENT DECISION-MAKING ACROSS SEVEN MARKETS IN ASIA-PACIFIC Dummett H, Shankland BDT Double Helix Consulting, London, UK OBJECTIVES: HTA is at different stages of development across the Asia-Pacific region, from the highly developed system in Australia at one extreme to China at the other, where the impetus to formalize HTA exists but guidelines and a dedicated agency do not. Understanding the weight of HTA as a driver of reimbursement decision-making both now and in the next 5 years is crucial for informing reimbursement strategy. METHODS: A total of 20 HTA reviewers and academic health economists were interviewed across Australia, China, Korea, Malaysia, the Philippines, Taiwan and Thailand to understand the parameters of the HTA system, the importance of different stakeholders within the process and the decisions influenced by HTA. RESULTS: HTA systems within Asia-Pacific exist at all stages of the HTA development continuum. At one end sit two single national payer systems, Australia and Korea, in which demonstration of cost-effectiveness is a formal prerequisite for reimbursement. At the other extreme sit China and Malaysia in which no formal role for HTA yet exists, although the speed and direction of HTA development is likely to differ. In between sit the Philippines, Taiwan and Thailand win which the role of HTA is becoming more formalized, albeit largely as one among several criteria in decisionmaking. CONCLUSIONS: HTA is developing rapidly within the markets we considered, rendering the submission of economic evidence mandatory in most markets. But a formal role for cost-effectiveness evaluation does not always translate into costeffectiveness being the basis for decision-making. Understanding the nuances of where HTA sits in the reimbursement system and how it is applied in practice in each market is essential for maximizing the possibility of favorable reimbursement.

4th Asia-Pacific Abstracts PCN39 TREATMENT PATTERNS AMONG PATIENTS WITH ADVANCED MELANOMA: A RETROSPECTIVE LONGITUDINAL STUDY O’Leary B1, Colman S1, Ruth A2, Kotapati S3, Cook G2 1 Covance Pty Ltd, North Ryde, NSW, Australia; 2Bristol-Myers Squibb Pharmaceuticals, Noble Park, VIC, Australia; 3Bristol-Myers Squibb Pharmaceuticals, Wallingford, CT, USA OBJECTIVES: A range of treatments have been used for late stage metastatic melanoma including chemotherapy, immunotherapy drugs (given alone or in combination), palliative surgery and radiation therapy. Nevertheless, current treatment practices in Australia for patients with advanced melanoma are poorly defined. As such, a retrospective longitudinal survey was conducted to gain an insight into the treatments currently used in this patient group. METHODS: Three Australian public hospitals provided patient level data. Inclusion criteria encompassed patients with a history of an unresectable stage III/IV melanoma (advanced melanoma) at initial diagnosis and had at least 6 months follow-up between January 1, 2005 and December 31, 2006. Information on demographics, disease characteristics, adverse events, survival and treatment history was obtained from medical records of all eligible patients. Descriptive statistics were performed for the four most frequently used first-line treatments. Survival estimates were assessed using Kaplan Meier methods. Comparisons of survival by treatment choice were assessed using a log-rank test. RESULTS: In total, 51 patients were recruited with a mean age of 61.6 years and the majority Caucasian (96.1%). Twelve patients (23.5%) were stage III and 23 (45.1%) stage IV. Specific staging of remaining 16 (31.4%) advanced melanoma patients was unknown. The main first-line treatments were dacarbazine (17.6%), fotemustine (15.7%), radiotherapy (23.5%), and surgery (23.5%). Patients treated with surgery were younger and mainly stage IV at diagnosis. Surgery was associated with longer survival (P = 0.0016); however, a causal relationship cannot be inferred. There was a tendency for fewer patients to receive surgery for second-line treatment with slightly more receiving radiotherapy. Chemotherapy was associated with nausea while hematological events were common in fotemustine patients. CONCLUSIONS: This study found that four main treatments are used for unresectable stage III/IV melanoma: dacarbazine, fotemustine, radiotherapy and surgery. The data suggests that surgery was associated with longer overall survival compared to the alternative therapies.

CARDIOVASCULAR DISORDERS – Clinical Outcomes Studies PCV1 COST-EFFECTIVENESS OF ACUTE ISCHEMIC STROKE IN CHINA: MARKOV MATRIX ANALYSES Wang M Shenyang Pharmaceutical University, ShenYang, Liaoning, China OBJECTIVES: To measure economic consequences of NBP (Butylphthalide) for acute ischemic stroke in China. METHODS: A multi-center, randomized, double-blind control study was conducted, 220 patients were admitted into the treatment group and control group, the main regimen is A1 (Butylphthalide sodium chloride injection and Butylphthalide sodium chloride capsule), A2 (Butylphthalide sodium chloride injection), B (Ozagrel injection). We explored the MRS (modified Rankin score) to establish Markov state transition model, with six different health states. The cost and effectiveness data were come from the studies based on Chinese acute ischemic stroke studies. Based on transferred Markov matrix, we will obtain the time period for disease steady-state and predict long-term costs and cost-effectiveness from the patient’s perspective. RESULTS: A1 group reached the steady state first, five cycles (10 months), A2 and B group were behind a period. The steady state cost of A1 group involved 79,396.90 yuan, more than A2 group by 19,379.54 yuan, and 85,806.71 yuan for B group, compared with the extension days with no obvious significant. At a steady state, the cost-effectiveness ratio of A1 Group’s health state was 86.3 thousand yuan, A2 group 82.1 thousand yuan and B group 96.4 thousand yuan, Thus, economic results of A group (A1 and A2) were better than B group, A2 group slightly better than A1 group, but given the time to reach a steady state, the A1 group had more advantages of economic effects. CONCLUSIONS: Based on the results of this model-based economic analysis, A1 strategy was the most cost-effective option in comparisons with A2 and B strategies for acute ischemic stroke in China. PCV2 SUCCESSFUL IMPLEMENTATION OF A P&T APPROVED THERAPEUTIC INTERCHANGE PROGRAM OF ANGIOTENSIN II RECEPTOR BLOCKERS IN A MEDICAL CENTER IN TAIWAN Lee YY1, Hsiao P2, Yen YH1, Lin YM3, Chen HY1 1 Taipei Medical University-Wan Fang Hospital, Taipei, Taiwan; 2Santa Clara Valley Health & Hospital System, San Jose, CA, USA; 3Taipei Medical University- Shuang Ho Hospital, Jhonghe City, Taiwan OBJECTIVES: Therapeutic interchange is not a well-accepted clinical practice in the medical society in Asia. Clinic blood pressure readings, patients’ tolerance, and costeffectiveness were used as measures to evaluate the impact of a therapeutic interchange program implemented at a medical center in Taiwan. METHODS: Taipei Medical University-Wan Fang Medical Center initiated a therapeutic interchange program involving angiotensin II receptor blockers (ARBs). Data were retrospectively collected for 444 hypertensive outpatients who were converted from other ARBs to candesartan. Evaluation of therapeutic efficacy, adverse effects associated with therapy, and drug costs was conducted before and after the program implementation. RESULTS: Patients

A517 who were converted to candesartan experienced no statistically significant differences in blood pressure, and the average number of antihypertensive agents used per patient remained unchanged. A direct cost savings of US$62,237 was estimated for the 444 patients studied, which amounted to a savings of more than US$730,000 annually for all outpatients using ARBs. Only 3.15% of patients developed ADRs potentially related to the candesartan and none required hospitalization. CONCLUSIONS: Based on the results of this retrospective chart review, the present therapeutic interchange program was successfully developed and implemented. This is the first study to establish the positive impact of a well-run ARBs therapeutic interchange program in Taiwan. PCV3 MIXED TREATMENT COMPARISON OF PROPHYLAXIS REGIMENS FOR THE PREVENTION OF VENOUS THROMBOEMBOLISM IN TOTAL HIP REPLACEMENT, HIP FRACTURE, TOTAL KNEE REPLACEMENT AND GENERAL SURGERY Farris M1, LeReun C2, Lee J3, Tilden D1 1 THEMA Consulting Pty Ltd, Pyrmont, NSW, Australia; 2Independent Biostatistician, Carrigaline County, Cork, Ireland; 3Sanofi-Aventis Australia, Sydney, NSW, Australia OBJECTIVES: To compare the efficacy and safety of several different anticoagulant therapies in high-risk surgical patients using a mixed treatment comparison (MTC) approach. METHODS: A manual literature search for randomized double blind controlled trials (RDBCTs) was undertaken that compared any two of the following agents: enoxaparin, rivaroxaban, unfractionated heparin (UFH) or placebo and reported rates of venous thromboembolism (VTE) in patients undergoing total hip replacement, hip fracture surgery, total knee replacement and general surgery. Bayesian methods of MTC meta-analysis were applied to synthesize data from 31 RDBCTs. Safety data (defined as major hemorrhage) were extracted when available. Patients received prophylaxis for up to 15 days, or extended therapy up to 35 days postsurgery. RESULTS: All treatments were compared to placebo using the Odds Ratio as a measure of efficacy. When compared to placebo, the incidence of VTE was reduced by ≥50% with prophylactic therapy post-hip surgery; patients undergoing extended prophylaxis with enoxaparin or rivaroxaban received the greatest benefit, with a reduction in the risk of VTE of 87% or 95% respectively. Prophylaxis also reduced the incidence of VTE after knee surgery by ≥80%; patients treated with enoxaparin or rivaroxaban were 87% or 92% less likely to experience VTE, respectively, when compared to placebo. Prophylaxis after general surgery significantly reduced the risk of VTE over placebo by ≥76%. The risk of major hemorrhage was highest with UFH, followed by rivaroxaban and enoxaparin. CONCLUSIONS: The MTC approach resulted in a comprehensive examination of the effect of thromboprophylaxis on the rate of VTE and major hemorrhage in high-risk surgery. Upon examining the body of clinical data, the evidence is clear: thromboprophylaxis is highly effective at preventing VTE in patients undergoing high-risk surgery. When compared with rivaroxaban and UFH, prophylaxis with enoxaparin resulted in similar and lower incidence of VTE respectively, with less major hemorrhage. PCV4 EFFECT OF PRETREATMENT WITH ANGIOTENSIN-CONVERTING ENZYME INHIBITORS ON IN-HOSPITAL MORTALITY AFTER ACUTE ISCHEMIC STROKE Al-Jabi SW1, Hassan Y2, Abd Aziz N2, Looi I3, Zyoud SH1 1 Universiti Sains Malaysia (USM), Penang, Malaysia; 2Universiti Sains Malaysia, Pinang, Malaysia; 3 Pulau Pinang Hospital, Pinang, Malaysia OBJECTIVES: Pretreatment with angiotensin-converting enzyme inhibitors (ACEI) has shown promising results in decreasing the incidence and the severity of ischemic stroke in populations at risk, and in improving ischemic stroke outcomes. The aims of this study were to investigate the impact of ACEI use before ischemic stroke onset on in-hospital mortality and to identify the independent predictors of in-hospital mortality among ischemic stroke patients. METHODS: It is a retrospective cohort study of all acute ischemic stroke patients admitted to a 1200-bed hospital located in northern Malaysia from June 1, 2008 to November 30, 2008. Data were collected from medical records and included demographic information, diagnostic information, risk factors, previous ACEI use and vital discharge status. SPSS version 15 was used for data analysis. RESULTS: Overall, 327 acute ischemic stroke patients were studied, of which 119 (36.4%) had documented previous ACEI use. During the study period, 52 (15.9%) of the acute ischemic stroke patients died in hospital. In-hospital mortality was significantly lower among patients who were on ACEI before the attack (P = 0.002). The independent predictors for in-hospital mortality among ischemic stroke patients were the presence of diabetes mellitus (P = 0.023), renal impairment (P < 0.001) and heart failure (P = 0.007). Moreover, prior use of ACEI was an independent predictor for survival after ischemic stroke attack (P < 0.001). CONCLUSIONS: This study provides evidence that the prophylactic administration of ACEI is a potential life-saving strategy. Furthermore, knowledge of in-hospital mortality predictors is necessary to improve survival rate after acute ischemic stroke attack.

A518 PCV5 ASSESSMENT OF LIPID ABNORMALITIES: PREVALENCE AND ATTAINMENT OF LIPID GOALS/NORMAL LEVELS AMONG ADULTS IN THAILAND Bash LD1, Benjanuwattra T2, Chaithiraphan V3, Khovidhunkit W4, Nimitpong H5, Ongpiphadhanakul B5, Silaruks S6, Sritara P5, Ambegaonkar BM1 1 Merck & Co., Inc., Whitehouse Station, NJ, USA; 2Maharaj Nakorn Chiang Mai Hospital, Chiang Mai, Thailand; 3Siriraj Hospital, Bangkok, Thailand; 4Chulalongkorn University, Bangkok, Thailand; 5Ramathibodi Hospital, Bangkok, Thailand; 6Khon Kaen University, Khon Kaen, Thailand BACKGROUND: Despite effective lowering of low density lipoprotein cholesterol (LDL-C) with statin for prevention of cardiovascular events, risk remains high in those with mixed dyslipidemia. OBJECTIVES: To examine the prevalence of dyslipidemia before and after lipid modifying therapy (LMT) use in Thailand. METHODS: We recruited 807 patients (mean age 59 and 45.2% male) who, between January 2001 and June 2007, aged ≥35 years, initiated LMT, had ≥1 lipid abnormality and follow-up >12 months after initiating LMT, with a complete lipid panel (LDL-C, high density lipoprotein cholesterol [HDL-C] and triglycerides[TG]) before and 12 months after therapy. Patients with coronary heart disease (CHD), type 2 diabetes or a 10-year CHD risk >20% were classified as high (cardiovascular) risk patients. Threshold levels for LDL-C, HDL-C and triglycerides were specified as per NCEP ATPIII Guidelines. RESULTS: At baseline, 65%, 34%, and 43% exhibited elevated LDL-C, elevated TG, and low HDL-C, respectively while 35% had elevated LDL-C alongside low HDL-C and/or elevated TG. Among high-risk patients (n = 430), 68%, 35%, and 44% exhibited elevated LDL-C, elevated TG and low HDL-C, respectively; while 38% had elevated LDL-C alongside low HDL-C and/or elevated TG. Among the sample, 77% were on statins, 7% on fibrates, and 13% on both. After 12 months, 21%, 32%, and 39% of patients still had elevated LDL-C, elevated TG and low HDL-C, while 18% had elevated LDL-C alongside low HDL-C and/or elevated TG. High-risk patients had similar degrees of persistent dyslipidemia with 27%, 25%, 44%, and 21% experiencing elevated LDL-C, elevated TG, low HDL-C, and elevated LDL-C alongside low HDL-C and/or elevated TG, respectively. CONCLUSIONS: This cohort had improved LDL-C levels after therapy, with no improvement of TG and HDL-C levels among the overall, and moderate and negligible among the high-risk population, respectively. These patients may benefit from other types of LMT. PCV8 LOST YEARS OF LIFE DUE TO DEATH FROM PULMONARY EMBOLISM IN HOSPITALIZED PATIENTS VERIFIED BY AUTOPSY FROM 1983 TILL 2005 Lukac M1, Gaspar L2, Makovnik M2, Stvrtina S3 1 Slovak Medical University, Bratislava, Slovak Republic; 2University Hospital Bratislava, Bratislava, Slovak Republic; 3Faculty of Medicine, Comenius University, Bratislava, Slovak Republic OBJECTIVES: To assess average amount of life-years lost due to death from pulmonary embolism (PE) confirmed by autopsy in patients hospitalized at internal medicine department. METHODS: In our prospective observation we analyzed all fatal cases at Second Department of Internal Medicine at University Hospital Bratislava, where cause of death was verified by autopsy during the period from 1983 till 2005. Age, sex, clinical diagnosis, and pulmonary embolus finding from autopsy were recorded for each case. Sex adjusted life expectancy for each calendar year (1983–2005) was used from official statistical sources available for Slovakia and lost years of life were calculated accordingly. RESULTS: There were 1375 deaths and in 70% (n = 963) autopsy was performed during the observational period of 23 years. Fatal PE was found in 12% of cases (n = 118), from which 64 cases were not clinically diagnosed prior to autopsy. Average age for group with confirmed PE (n = 118) was 76.5 years (interval 38–96) with 54 (46%) males and 64 (54%) females. As calculated from life expectancy data, there were 939 life-years lost due to fatal PE verified by autopsy. In average 42.7 years of life (95% CI 32.8–52.6) were lost each year. CONCLUSIONS: Results of our study shows that PE continues to be important cause of mortality in hospitalized patients, accounting for 12% of deaths at internal department. In average 42.7 years of life were lost each year in one medical department due to PE. Despite the advances in diagnostic methods and thromboprophylaxis of VTE during the past decades, lost of life-years caused by fatal PE are still significant among medically ill patients in the period from 1983 till 2005.

CARDIOVASCULAR DISORDERS – Cost Studies PCV9 POTENTIAL COST SAVINGS OF SMOKING CESSATION WITH VARENICLINE AMONG INSURED CHINESE POPULATION Wang D, Dong P Pfizer China, Beijing, China OBJECTIVES: This study is to estimate potential cost savings by reducing the incidence of two major smoking-related illnesses—chronic obstructive pulmonary disease (COPD) and coronary heart disease (CHD)—through successful smoking cessation among insured Chinese population. METHODS: A predictive economic model was constructed upon a hypothetical cohort of insured patients to assess cost savings of an oral smoking cessation drug (varenicline) in terms of reduced COPD/CHD incidence over 15 years (at 5-year intervals). The size of the cohort was determined by

4th Asia-Pacific Abstracts COPD/CHD prevalence, COPD/CHD population attributable to smoking and percentage of patients insured drawn from public information and literature, and data of smokers’ willingness to use oral smoking cessation drugs obtained from internal market survey. Medical treatment costs of varenicline were estimated according to current pricing scheme. Cost-savings at different intervals were estimated using parameters of successful tobacco abstinence rate calculated from varenicline clinical trials, risk reduction of COPD/CHD drawn from literature, and annual management costs of COPD/CHD patients obtained from provincial health insurance database. Sensitivity analysis on tobacco abstinence rate in varenicline users was also performed. Overall cost-savings were then estimated combining medical treatment costs of varenicline and associated cost-savings together. Cost-savings due to reduced disease risk are calculated at each 5-year interval. RESULTS: A total of 19,583 subjects are included in our cohort. Costs of completed varenicline cycles are estimated to be RMB 42 million. At the first year, medical costs for varenicline outweigh the savings by around RMB 32 million. However, increased cost savings are observed at Year 5, 10, and 15, ranging from RMB 25–27 million, RMB 120–126 million, RMB 242–253 million, respectively. CONCLUSIONS: Successful smoking cessation through varenicline treatment could lower the burden of smoking-related illness in China, and its associated cost savings are substantial to the society in the long run. PCV10 COST SAVINGS OF SINGLE PILL AMLODIPINE/ATORVASTATIN THERAPY IN HYPERTENSION AND DYSLIPIDEMIA PATIENTS IN CHINA Dong P, Wang D Pfizer China, Beijing, China OBJECTIVES: This study is to estimate health expenditures in hypertension and dyslipidemia (HTN/DYS) patients with cardiovascular risks over a period of 10 years in Chinese health-care settings. Costs are compared among patients using single pill amlodipine/atorvastatin (SPAA) therapy, amlodipine therapy only, and no intervention. METHODS: Three hypothetical cohorts of HTN/DYS patients were constructed: 1) SPAA group; 2) amlodipine only group; and 3) no intervention group. The size of eligible population was determined by national prevalence data on patients with HTN/DYS and those not receiving statin therapies, and on percentage of patients currently insured derived from total publication. For each group, data on risk of having cardiovascular illness over 10 years, average costs for cardiovascular illness management per patient, and average costs for HTN/DYS treatment per patient were identified from existing literature. Overall health expenditures were then calculated by costs of having cardiovascular illness within 10 years, plus costs of treating HTN/DYS over 10 years. Comparisons were made between SPAA versus amlodipine only, and SPAA versus no intervention. RESULTS: Of the 26.89 million eligible patients included in this study, the overall health expenditures over 10 years in patients with SPAA therapy, amlodipine only, and no intervention were estimated to be RMB 37.65 billion, RMB 59.17 billion and RMB 75.30 billion, respecbtively. The SPAA therapy allows payers to save annual cost of 2.15 billion when compared with amlodipine only therapy, and annual cost of 3.77 billion when compared with no intervention at all. CONCLUSIONS: SPAA therapy in HTN/DYS patients is likely to generate more cost savings to the society by reducing the costs of cardiovascular illness, when compared with traditional therapies. PCV11 EFFECTS OF MEDICATION SUPPLY ON HEALTH-CARE COSTS AND RE-HOSPITALIZATIONS IN PATIENTS WITH CHRONIC HEART FAILURE Sruamsiri R1, Dilokthornsakul P1, Nimpitakpong P1, Jeanpeerapong N2, Chaiyakunapruk N1 1 Naresuan University, Muang, Phitsanulok, Thailand, 2Buddhachinaraj Hospital, Muang, Phitsanulok, Thailand OBJECTIVES: Previous evidence has shown that using Angiotensin Converting Enzymes (ACEIs) or Angiotensin Receptor Blockers (ARBs) results in decreased morbidity among patients with chronic heart failure (CHF). However, the benefits earned from medications depend on the amount of medications supplied and used. This study aims to determine the effects of medication supplies on health-care costs and hospitalizations in CHF patients receiving ACEIs or ARBs. METHODS: We retrospectively examined an electronic patient database in a provincial hospital located in the north of Thailand. All patients with an ICD-10 of I-50.0 (CHF) receiving either ACEIs or ARBs from January to December, 2003 were included. Their medication supplies were assessed using Medication Possession Ratio (MPR) method during the study period following the index date. The association between medication supply (appropriate: MPR 0.8–1.20, oversupply: MPR >1.20, undersupply: MPR 8%; b) stratification of outcome by dose of gliptin; and c) addition or discontinuation of OHA from baseline all did not demonstrate statistically significant difference. Majority of patients were not on maximal OHA doses at gliptin initiation, however total daily doses of OHAs was not significantly different at exit versus baseline for both arms. Almost 90% of patients in both groups received multiple OHAs for diabetes control. Change in creatinine clearance was comparable in both arms. Safety endpoints microalbuminuria/creatinine ratio, average % weight change and incidence of pancreatitis were not significantly different between both arms (all P > 0.05). Five sitagliptin patients required hospital admission for severe hypoglycemia vs 0 vildagliptin patients. CONCLUSIONS: We present our initial findings that vildagliptin is non-inferior to sitagliptin in HbA1c-lowering efficacy. Both products are well-tolerated without significant differences in safety endpoints save severe hypoglycemia in sitagliptin arm. PDB2 LONG-TERM HEALTH OUTCOMES OF TREATMENT WITH LIRAGLUTIDE VERSUS GLIMEPRIRIDE IN TYPE 2 DIABETES PATIENTS IN ASIAN SETTING Wu J1, Wu JH2, Chang J3 1 Tianjin University, Tianjin, China; 2306 Hospital of PLA, Beijing, China; 3Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China OBJECTIVES: To evaluate the long-term health outcomes associated with Liraglutide 1.2 and 1.8 mg versus Glimepiride 4 mg all combined with Metformin in Asian patients with type 2 diabetes (T2D). METHODS: A published and validated computer simulation model of diabetes (CORE Diabetes Model) was used to make the projection of long-term health outcomes (30 years). Simulated cohorts and treatment effects were derived from 928 T2D patients in the NCT00614120 trial held in China, South Korea and India. HbA1C was significantly reduced in Liraglutide 1.2 mg, Liraglutide 1.8 mg, and Glimepiride groups (−1.3%, −1.4%, and −1.3% respectively). Liraglutide treatments led to greater reduction in Body Mass Index and systolic blood pressure versus Glimepride. No major hypoglycemia was reported in Liraglutide groups, while the rate of major hypoglycemia for Glimepride was 0.029 per patient-year. The rate of minor hypoglycemia was lower in Liraglutide groups than Glimepiride. An annual discounting rate of 3% was used for health and cost outcomes. One-way sensitivity analysis was performed. RESULTS: The treatments of Liraglutide compared with Glimepiride were projected to reduce the cumulative incidences of diabetes complications and improve long term health outcomes for patients with T2D. For Liraglutide 1.2 mg, the cumulative incidences of background retinopathy, end stage renal disease, ulcer, and congestive heart failure event were reduced 0.20%, 0.086%, 0.020% and 0.53% respectively, discounted life expectancy was increased 0.058 year and quality adjusted life-years (QALY) was increased 0.11 QALY. For Liraglutide 1.8mg, the incidences reduction were 0.61%, 0.12%, 0.34% and 0.63% respectively, discounted life expectancy was improved 0.051 year, and 0.107 QALY. CONCLUSIONS: Liraglutide 1.2 mg and 1.8 mg therapy could delay the onset of diabetes complications and reduced related cumulative incidences over patient lifetimes compared with Glimepiride. It improved the life expectancy and quality adjusted life expectancy in Asian patients with T2D.

4th Asia-Pacific Abstracts PDB3 EFFECTS OF EXTENDED-RELEASE VERSUS IMMEDIATE-RELEASE GLIPIZIDES IN PATIENTS WITH TYPE 2 DIABETES MELLITUS: A SYSTEMATIC REVIEW AND META-ANALYSIS Wang L1, Sun X2, Li Y3 1 Sichuan University, Chengdu, Sichuan, China; 2McMaster University, Hamilton, ON, Canada; 3 West China Hospital, Sichuan University, Chengdu, China OBJECTIVES: To address effects of extended-release versus immediate-release glipizide on glucose control, insulin secretion, and compliance. METHODS: We included parallel randomized trials and cohort studies (only for compliance assessment) comparing extended-release versus immediate-release glipizide for type 2 diabetes. We searched Medline, EMBASE, the Cochrane Library, and Chinese Biomedical database, screened reference lists, and contacted pharmaceutical company. We pooled data using random-effect model and explored heterogeneity by pre-specified hypotheses. RESULTS: Sixteen trials involving 1062 patients and two retrospective cohort studies of 13,452 patients were included. Trials are of inadequate quality. No trials reported patient-important outcomes. The reduction of fasting plasma glucose from the baseline appeared larger in extended-release than immediate-release glipizide (mean difference −0.30 mmol/L, 95% CI −0.57 to −0.03). The reduction was not significant different between two drugs in HbA1c (−0.03%, −0.20% to 0.15%) and 2-hour postprandial plasma glucose (−0.28 mmol/L, −1.12 to 0.55). Extended-release glipizide appeared to reduce insulin secretion from the baseline, whereas immediate-release formulation increased the secretion (fasting insulin: −0.86 vs. 0.28 μIU/ml; 2-hour postprandial insulin: −2.94 vs. 0.24 μIU/ml). Patients administering extended-release glipizide had less hypoglycemia (Peto odds ratio 0.26, 95% CI 0.08 to 0.81) and lower missed dosing (Peto odds ratio 10.24, 95%CI 5.22 to 20.08). The cohort studies showed results in compliance consistent with trials. CONCLUSIONS: The two drugs may have comparable effects on glucose control. Extended-release glipizide might achieve glucose control with decreased insulin secretion, and fewer hypoglycemic episodes. The findings are inconclusive due to inadequate study quality, short follow-up, and unavailability of patient important outcomes.

DIABETES/ENDOCRINE DISORDERS – Cost Studies PDB4 MEDICAL SERVICE COST ASSOCIATED WITH PIOGLITAZONE AND SULFONYLUREA TREATMENT AMONG TYPE 2 DIABETIC PATIENTS ENROLLED IN A US INTEGRATED HEALTH-CARE SYSTEM Sun SX1, Ma L2, Bron M1, Lamothe K2, Boulanger L3 Takeda Pharmaceuticals International, Deerfield, IL, USA; 2Abt Bio-Pharma Solutions, Inc., Lexington, MA, USA; 3United BioSource Corporation, Lexington, MA, USA OBJECTIVES: To assess overall and diabetes-related medical service costs associated with pioglitazone (PIO) and sulfonylureas (SU) treatment among T2DM patients. METHODS: This is a retrospective cohort study based on electronic medical records (January 1, 2004–January 31, 2009) from the Geisinger Clinic in the Northeastern region of the United States. The date of the initial prescription for PIO or SU was denoted as the index date. Patients were required to be aged 18 years or older and prescribed an oral antidiabetic treatment in the 1 year prior to index. Patients with type 1 or gestational diabetes and prior insulin use were excluded, as were those who had prescriptions for the index drug in the 90 days prior. Propensity score 1:1 matching and a second stage of generalized linear regression were employed to assess overall and diabetes-related medical service costs (pharmacy costs were not available in the database) in the 2 years following the index date, adjusting for patient demographics, baseline comorbidities, medication use, and health-care resource utilization. RESULTS: A total of 2758 patients, 1379 each in the PIO and SU cohorts, were analyzed. For both cohorts, mean age was 62 years, 46% were male, and 96% were Caucasian. The two cohorts were similar in terms of current smoking status and diabetes-related comorbidities. The unadjusted total and diabetes-related medical costs were $1258 and $705 higher for SU versus PIO patients. After adjusting for covariates, the overall and diabetes-related medical service costs remained higher for patients receiving SU versus PIO ($8360 vs. $7400 for overall, and $5577 vs. $5238 for diabetes-related costs, P < 0.05 for both comparisons). CONCLUSIONS: Over a 2-year follow-up, patients with T2DM initiated on PIO therapy in an integrated system incurred lower overall and diabetes-related medical service costs than patients initiated on SU. Further studies describing clinical and humanistic aspects of PIO versus SU are warranted. PDB5 TOTAL AND DIABETES-RELATED COSTS ASSOCIATED WITH HYPOGLYCEMIA IN TYPE 2 DIABETES MELLITUS PATIENTS INITIATED ON ORAL ANTIDIABETIC DRUGS FROM A LARGE US MANAGED CARE COHORT Bron M1, Yu A2, Marynchenko M3, Sun SX1, Yang H2, Wu EQ2 1 Takeda Pharmaceuticals International, Deerfield, IL, USA; 2Analysis Group, Inc., Boston, MA, USA; 3Analysis Group, Inc., Montreal, QC, Canada OBJECTIVES: To estimate annual health-care costs associated with hypoglycemia among T2DM patients initiated on oral antidiabetic drugs (OADs) in a large managed care cohort with managed care insurance benefits. METHODS: T2DM patients initiated on OADs were selected from the Ingenix Impact database (1999–2008). Patients aged 18 years or older with at least 1 year of continuous eligibility following the index date (the first OAD prescription fill date) who were diagnosed with moderate to severe

4th Asia-Pacific Abstracts hypoglycemia events (ICD-9-CM 250.8×, 251.0×, 251.1×, 251.2×) were identified. Annual total, medical, pharmacy, and diabetes-related costs for patients with or without the diagnosis of hypoglycemia were compared during the 12-month postindex period. Diabetes-related medical costs included all non-pharmacy costs associated with T2DM diagnosis. Generalized linear regressions with robust standard error accounting for skewed cost distribution were performed, adjusting for demographics, comorbidities, and OADs. RESULTS: A total of 212,061 T2DM patients with at least one OAD treatment were identified. Among them, 4860 (2.29%) had a hypoglycemia diagnosis during the first year following the index date. Patients with hypoglycemia had significantly higher average annual total costs ($18,273 vs. $8908, P < 0.001) and diabetes-related total costs ($8969 vs. $3220, P < 0.001) than those without hypoglycemia. After adjusting for confounding factors, hypoglycemia patients had significantly higher incremental annual total costs and diabetes-related total costs than patients without hypoglycemia ($5031 and $3751, respectively, both P < 0.001). Similar trends were observed for annual medical costs and diabetes-related medical costs ($4967 and $3796, respectively, both P < 0.001). CONCLUSIONS: Hypoglycemia in type 2 diabetes patients initiated with OADs in a large US managed care setting is associated with significantly higher total, medical, and diabetes-related costs compared with those who did not experience hypoglycemia. PDB6 THE DISEASE BURDEN ANALYSIS OF 295 INPATIENTS WITH DIABETES MELLITUS FROM TONGJI HOSPITAL IN CHINA He M1, Ma J2, Wang D1, Chang J3, Yu X1 1 Tongji Medocal College, Huazhong University of Science and Technolgy, Wuhan, Hubei, China; 2Huazhong University of Science and Technology, Wuhan, Hubei, China; 3Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China OBJECTIVES: This study was to examine the prevalence and economic burden of in patients with diabetes and/or related complications from Tingyi hospital in Wuhan China. METHODS: This was a retrospective analysis focused on hospitalization costs and complications for all patients diagnosed with diabetes and administered in Wuhan Tingyi hospital in China from June 2007 to December 2007. All information of patient demographic characters, clinical and costs were collected for the analysis. The descriptive statistics was used to describe patients’ demographic characters, duration of the diseases and treatment costs. RESULTS: There were total of 295 inpatients were included in the study. The mean age of patients was 51.6 ± 16.0 years (range 11–87 years), 53.56% were male. The diabetes durations (±SD) were 5.75 ± 6.04 years. 13.56% were type 1 diabetes, and 86.44% were type 2 diabetes. The mean hospital stay (±SD) were 11.99 days (±0.54), and mean of hospitalization costs (±SD) were USD 876.6 (±39.4) (1US$ = 7.6 RMB). The patients with no complication, micro vascular complications, macro vascular complications or both were 10.20%, 30.27%, 8.16%, and 51.36% respectively. Peripheral neuropathy (69.2%) and hyperlipidemia (64.1%) were of the highest prevalence among all the complications, and diabetic foot ulcer was the expensive complication with the highest treatment costs (USD657). The costs of tests and medications accounted for 27% and 26% of total hospitalization expenses. Anti–diabetes drugs, complication treatments and other drugs were of 37.7%, 52.8%, and 9.4% respectively in the medication costs. CONCLUSIONS: Most of hospitalized patients had diabetes complications. Diabetic complications contributed to a substantial economic burden for the patients and society. It implied the urgent need for health-care policymakers to address this issue by adopting effective measures toward prevention and control of diabetes and its complications. PDB7 ESTIMATED COSTS OF INCLUDING GROWTH HORMONE REPLACEMENT THERAPY INTO THE UNIVERSAL COVERAGE BENEFIT PACKAGE IN THAILAND Prakongsai P1, Pachanee K2 1 International Health Policy Program (IHPP), Nonthaburi Province, Thailand; 2International Health Policy Program (IHPP), Nonthaburi, Thailand BACKGROUND: Growth hormone replacement therapy (GHRT) is an expensive health service excluded from the universal coverage (UC) benefit package of Thailand. Its high costs and exclusion from the UC benefit package prevent poor patients from access to such health care. OBJECTIVES: This study aims to estimate costs of providing GHRT to patients diagnosed with growth hormone deficiency (GHD). It also explores demand for, and supply of, and total costs per unit of height increase from universal access to GHRT. METHODS: Methods include comprehensive literature review, in-depth interviews of childhood endocrinologists from three university hospitals about costs of GHRT and its current practice, and modeling costs for height increase using different products. RESULTS: Research findings indicate the incidence of GHD in Thailand is approximately 1 to 50,000 of people aged less than 15 years. A weekly dosage of GHRT recommended by the Thai Pediatric Endocrine Society range from 0.1 to 0.3 mg per kg depending on patient’s responses. Cessation criteria of GHRT comprise 1) attainment of adult height at 165 and 155 cm in male and female, respectively; 2) bone age above 16 and 14 years in male and female respectively; and 3) the annual increase of height lower than 2.5 cm. On supply side, approximately 40 pediatric endocrinologists are available in public and private settings across the country. Costs of GHRT are varied from factors including patient’s age and gender, type and dose of GH use. Data from a university hospital show costs of GHRT range from $950 to $2580 per every 1 cm of height increase when using a cheaper GH product. The costs will increase from $1460 to $4000 per every 1 cm of height increase when using the original GH. CONCLUSIONS: High cost burden of

A527 GHRT poses a challenging question on whether it should be included in the UC benefit package. PDB8 COST-EFFECTIVENESS OF SWITCHING PATIENTS EITH TYPE 2 DIABETES FROM INSULIN GLARGINE TO INSULIN DETEMIR IN A CHINESE SETTING: A HEALTH ECONOMIC MODEL BASED ON THE PREDICTIVE STUDY Yang L1, Christensen T2, Sun F3, Chang J3 1 Peking University, Beijing, China; 2Novo Nordisk A/S, Virum, Denmark; 3Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China OBJECTIVES: To evaluate the long-term health outcomes and economic consequences of the treatment of Insulin Detemir (IDet) in type 2 diabetes patients switching from insulin Glargine (IGla) in the setting of Chinese urban hospitals. METHODS: A published and validated computer simulation model of diabetes (the CORE Diabetes Model) was used to make the long- term (30 years) projection of health economic outcomes. The patient demographic information and clinical endpoints were derived from the PREDICTIVE subgroup study. The study was a large, multi-center, 6 months observational study assessing the safety and efficacy of IDet in everyday clinical practice. HbA1c was reduced of −0.59% by switching from IGla to IDet. Baseline risk factors and racial characteristic data were obtained from Chinese cohort studies. The market retail prices of medications were calculated to estimate treatment costs. The diabetes management and complications costs were obtained from Chinese published data and adjusted to 2009 values using the Chinese Consumer Price Index. An annual discounting rate of 3% was used for both health and cost outcomes. One-way sensitivities analysis was performed and illustrated that the results were robust. RESULTS: The treatment of IDet converted from IGla was projected to reduce the cumulative incidences of Eye disease, Renal disease, Ulcer, Cardiovascular disease, Cataract, EndStage Renal Disease, Ulcer, Neuropathy, Myocardial Infarction events were reduced 0.293%, 0.151%, 0.794%, 3.293%, 1.225% respectively. Patients’ time alive and free of complication was improved 0.15 year. Patient life expectancy was increased by 0.09 year and 0.36 quality adjusted life-year (QALY) when converting to IDet. The costs of complications were reduced by 4931 CNY (89,628 vs. 94,559), resulting in a total direct medical cost saving of 684 CNY. CONCLUSIONS: The treatment of IDet improved patient health and economic outcomes, and was a cost-saving treatment approach in a Chinese setting. PDB9 COST-EFFECTIVENESS OF CSII IN CHINA Ning G1, Lynch P2 1 Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China; 2Medtronic International, Tolochenaz, Switzerland OBJECTIVES: To estimate the long term clinical and cost outcomes of continuous subcutaneous insulin infusion (CSII) compared with multiple daily injections (MDI) of insulin in adult type 1 diabetes mellitus (T1DM) patients from a third party payer perspective in China. METHODS: A previously validated computer simulation model of diabetes (CORE Diabetes Model) was used to estimate the long-term clinical and cost consequences of CSII to MDI in the Chinese consumer-use setting and to calculate the incremental cost-effectiveness ratio (ICER). Clinical and cost data as well as treatment effects and patient characteristics were obtained from previously published studies. Costs and clinical projections were made over patient lifetimes from a thirdparty payer perspective and discounted at 3% annually. The primary input variable was change in HbA1c and was assumed to be an improvement of −1.2% for CSII compared with MDI. RESULTS: CSII was associated with improvements of 1.037 in quality adjusted life-years (QALYs) gained and 0.963 life-years gained per patient versus MDI. CSII was associated with an incremental cost-effectiveness ratio (ICER) of CNY 214,988 per QALY gained for CSII versus MDI. Improved glycemic control from CSII led to a lower incidence of lifetime diabetes complications; with the most significant reductions in proliferative diabetic retinopathy (PDR; a 29% reduction), end stage renal disease (ESRD; a 22% reduction) and peripheral vascular disease (PVD; a 15% reduction). CSII was associated with increased direct lifetime medical costs of Chinese Yuan (CNY) 222,994 per patient compared to MDI. CONCLUSIONS: For type 1 diabetes patients in China, CSII (consumer-use) would be considered cost-effective assuming a willingness-to-pay threshold of CNY 235,000 per QALY gained. PDB10 LOWER EXTREMITY AMPUTATION PREVENTION (LEAP) IN SINGAPORE: ECONOMIC ANALYSIS OF RESULTS Feng J1, Wong D2, Gordois A3 1 Ministry of Health, Singapore; 2Tan Tock Seng Hospital, Singapore; 3Access Economics, Sydney, Australia OBJECTIVES: The LEAP is a multidisciplinary management program (including angioplasty, bypass surgery and medical treatment) designed to enhance limb salvage for patients with diabetes and critical limb ischemia (CLI). This study aims to determine the cost-effectiveness of the LEAP strategy compared with the pre-LEAP strategy (standard practice) for treating patients with diabetes and CLI. METHODS: Clinical data relating to 277 patients in the LEAP were collected from 2001 until 2005. Comparative data were retrospectively collected for 144 patients with diabetes and CLI referred to the hospital in 2000. A lack of follow-up data in the pre-LEAP group restricted this study to the inpatient period. Significance of differences in therapeutic

A528 procedures and clinical outcomes between the two groups were analyzed using chi square and one-way t-tests. Costs included therapeutic management, hospitalization, and amputation. Outcomes included the numbers of amputations and deaths, and length of stay. The analysis was performed from the Singapore hospital perspective, using 2008 Singapore dollars. RESULTS: The LEAP group had a lower amputation rate (29% vs. 76%, P < 0.001), lower death rate (0.4% vs. 13.2%, P < 0.001) and fewer in-hospital days (17.8 days vs. 23.16 days, P = 0.048). Furthermore, implementation of the LEAP strategy generated cost savings of 2566 Singapore dollars per patient compared with the pre-LEAP strategy. The results were sensitive to amputation rates, cost per amputation, and cost per bed day. However, the LEAP is cost saving across all plausible variations in parameters (95% confidence interval limits for the reduction in amputation rates, 25% variation in costs). CONCLUSIONS: Use of non-randomized study data and a retrospective comparator require the results of this study to be interpreted with caution. However, the results suggest the LEAP strategy dominates standard practice for the management of patients with diabetes and CLI being both cost saving and more effective. PDB11 CLINICAL AND COST-EFFECTIVENESS OF BIPHASIC INSULIN ASPART 30 ONCE OR TWICE DAILY VERSUS INSULIN GLARGINE IN TYPE 2 DIABETES MELLITUS: A MODELING ANALYSIS IN THE CHINESE SETTING Palmer JL1, Beaudet A1, White J2, Plun-Favreau J2, Smith-Palmer J1 1 IMS Health, Allschwil, Basel-Landschaft, Switzerland; 2Novo Nordisk, Zurich, Switzerland OBJECTIVES: The OnceMix and INITIATE studies have indicated that biphasic insulin aspart 30 (BIAsp 30) is more effective than insulin glargine (IGlarg) in terms of reductions in glycohemoglobin in type 2 diabetes patients requiring insulin initiation. Our aim was to estimate the cost-effectiveness of BIAsp 30 versus IGlarg in the Chinese setting. METHODS: The previously published IMS Core Diabetes Model was used. The nephropathy, retinopathy and stroke sub-models were modified to incorporate available Chinese clinical data. Diabetes complication costs were derived from hospital surveys in Beijing and Chengdu. Simulated cohorts and insulin treatment effects were based on the OnceMix study for once-daily BIAsp 30 versus IGlarg and on the INITIATE study for twice-daily BIAsp 30 versus IGlarg. Life expectancy and direct medical costs were calculated. Projections were made over 30-year time horizons, with costs and life-years discounted at 3% annually. Extensive sensitivity analyses were performed, including adjustments to cardiovascular risk for Chinese ethnicity. RESULTS: Once-daily BIAsp 30 increased life expectancy by 0.04 years (12.37 vs. 12.33 years) and reduced direct medical costs by Chinese Yuan (CNY) 59,710 per patient (CNY 229,911 vs. CNY 289,621 per patient) compared to IGlarg in the OnceMix-based analysis. Twice-daily BIAsp 30 increased life expectancy by 0.08 years (12.99 vs. 12.91 years) and reduced direct medical costs by CNY 107,349 per patient (CNY 303,142 vs. CNY 410,491 per patient) compared to IGlarg in the INITIATEbased analysis. Improvements in life expectancy were driven by reduced incidences of most diabetes-related complications. Cost savings were attributable to lower insulin costs for BIAsp 30 compared to IGlarg in China. Lowering cardiovascular risk for Chinese ethnicity reduced the modeled clinical improvements for BIAsp 30 but increased treatment-related cost savings. CONCLUSIONS: BIAsp 30, either once- or twice-daily, improves life expectancy and is cost-saving compared to IGlarg in the Chinese setting. PDB12 COST-EFFECTIVENESS ANALYSIS OF GLIMEPIRIDE OR PIOGLITAZONE IN COMBINATION WITH METFORMIN IN TYPE-II DIABETIC PATIENTS IN KOREA Lim K Novartis Korea, Seoul, South Korea OBJECTIVES: Our objective was to compare the cost-effectiveness of glimepiride versus pioglitazone in combination with metformin in patients with type 2 diabetes. METHODS: The cost-effectiveness analysis (CEA) was performed by using Markov Cohort Process Mode with type 2 diabetic patients. Increases in life-years gained (LYG) and quality-adjusted life-years (QALYs) were measured as the main indicators of effectiveness. Direct medical cost includes drug cost, monitoring cost (medicalsupplies expense, total medicine-preparation cost, medical treatment cost, examination cost), and nursing cost (charged). Direct nonmedical cost includes the money spent on transportation by patients and families and indirect cost includes the time cost by patients and families. RESULTS: The Cost-Effectiveness Analysis (CEA) results showed that the combination therapy of metformin and glimepiride incurs total cost of KRW 5,652,814 to gain 8.065 LYG and 6.524 QALY. The combination therapy of metformin and pioglitazone costs KRW 8,613,622 to gain respective 8.836 LYG and 7.159 QALY. Thus the incremental cost-effectiveness ratios (ICERs) between two combination therapies were estimated to be KRW 3,837,632/LYG and KRW 4,658,636/QALY. CONCLUSIONS: In terms of ICER values, KRW 3,837,632/LYG and KRW 4,658,636/QALY, there is no dominant alternative therefore the costeffective alternative between two alternatives will be able to be selected depending on threshold in willingness to pay or ICER of South Korea.

4th Asia-Pacific Abstracts PDB13 THE IMPACT OF BEING AT-RISK FOR DIABETES IN CHINA AND THE UNITED STATES DiBonaventura MD1, Liu GG2, Wagner JS1, Stankus A3 1 Kantar Health, New York, NY, USA; 2Peking University, Beijing, China; 3Kantar Health, Princeton, NJ, USA OBJECTIVES: Because of the rising prevalence of diabetes, the objective of this study was to better understand the population of patients who are at high risk for developing diabetes, but not yet diagnosed. Specifically, this study examined the health outcomes between those at-risk for developing diabetes and controls. METHODS: This study utilized data from two large, cross-sectional, Internet-based survey databases, supplemented in China with centralized locations: the 2009 US and 2009 China National Health and Wellness Surveys (NHWS). Applying a previously-developed algorithm (Bang et al, 2009), patients in each country who were not diagnosed with diabetes were classified as having or not having a high risk for developing diabetes. High-risk patients were compared with controls on health-related quality of life (physical component summary (PCS) scores of the SF-12v2) and the number of emergency room (ER) visits, controlling for demographics (country, age, gender, ethnicity, income, education) and patient characteristics (BMI and Charlson comorbidity index). RESULTS: Of those not diagnosed with diabetes, 2.46% of patients (n = 310) in China and 16.64% of patients (n = 10778) in the United States were classified as high risk for diabetes. After controlling for demographics and patient characteristics, those at high risk for diabetes reported significantly lower levels of PCS (Adjusted Mean [Madj] = 45.5 vs. 48.8, P < 0.0001). Although high-risk patients reported significantly more ER visits than controls in China (M = 0.44 vs. 0.40) they reported significantly less than controls in the United States (M = 0.12 vs. 0.17). These effects were significantly different between countries (χ2[1] = 5.25, P = 0.022). CONCLUSIONS: Although substantially more US patients were at high risk for developing diabetes relative to Chinese patients, the health outcome differences between those at high-risk and controls were more dramatic in China. Chinese patients at high-risk for diabetes reported significantly worse physical quality of life and more ER visits than controls.

DIABETES/ENDOCRINE DISORDERS – Patient-Reported Outcomes Studies PDB14 USE OF 8-ITEM MORISKY MEDICATION ADHERENCE SCALE FOR THE ASSESSMENT OF MEDICATION ADHERENCE IN TYPE 2 DIABETES MELLITUS Al-Qazaz HK1, Sulaiman SA1, Hassali MA1, Shafie AA1, Sundram S2, Morisky DE3 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Hospital Balik Pulau, Balik Pulau, P.Penang, Malaysia; 3UCLA School of Public Health, Los Angeles, CA, USA OBJECTIVES: To assess the patient’s adherence to diabetic medications using 8-item Morisky medication adherence scale (MMAS) and to evaluate the association between patients’ adherence and the diabetic control outcome. METHODS: A convenience sample of 223 type 2 diabetic outpatients was identified between May and September, 2009 from the Penang General Hospital, Penang, Malaysia. Patients who had taken oral antidiabetic medications with or without insulin were eligible for the research. Eight-item MMAS was used to assess medication adherence. Medical records were reviewed for Hemoglobin A1C (HbA1C) levels and other relevant clinical data. RESULTS: Based on the inclusion criteria, only 175 were included in the final analysis. Employing the recommended method of scoring, the mean ± SD of MMAS scores was 6.13 ± 1.72 in which 38% were low, 44% were medium and only 17% were in high adherence group. Significant differences in educational levels, HbA1c levels, number of medications per day, MMAS scores among the three groups (P < 0.05). No significant differences have been found in age, sex, BMI, race, employment and diabetic duration among the three groups of adherence (P ≥ 0.05). MMAS scores correlates significantly with HbA1c (−0.431 P < 0.01). CONCLUSIONS: This study has shown that the short, easily administered 8-item MMAS is a useful tool for measuring medication adherence in diabetic patients. This study found that patients with lower score of MMAS had a higher HbA1c. The present study suggested that MMAS can be used for identifying type 2 diabetes patients with poor adherence to their medications. PDB15 HEALTH STATUS, HEALTH-RELATED QUALITY OF LIFE AND TREATMENT SATISFACTION AMONG PATIENTS WITH DIABETES IN THAILAND Pongmesa T1, Wee HL1, Li SC2 1 Department of Pharmacy, National University of Singapore, Singapore; 2University of Newcastle, Callaghan, NSW, Australia OBJECTIVES: To assess health status, health-related quality of life (HRQoL), treatment satisfaction (TS) and patient characteristics associated with these outcomes among Thai patients with diabetes. METHODS: Randomly selected patients with Type 2 diabetes aged ≥18 years were recruited from Samut Sakhon Hospital in Thailand from December 2008 to February 2009 in this cross-sectional study. The Thai versions of the EuroQol 5-Dimensions (EQ-5D) and the Audit of Diabetes-Dependent Quality of Life (ADDQoL-19) Questionnaire were used to assess health status and HRQoL, respectively, while the Diabetes Treatment Satisfaction Questionnaire status version (DTSQs) was used to assess satisfaction with diabetes treatment. RESULTS: A total of 201 patients were recruited with the majority being female, aged ≥45 years, of primary education and required only oral antidiabetes therapy. Their mean ± SD

4th Asia-Pacific Abstracts duration of known diabetes was 6.80 ± 6.67 years. A substantial negative impact of diabetes on the participants’ HRQoL was observed as assessed by the ADDQoL (mean ± SD average weighted impact (AWI) score: −3.76 ± 2.03 from the possible range −9 to +3), although the level of their TS was rather high as assessed by the DTSQs (mean ± SD TS scale total: 28.60 ± 5.53 out of 36). The participants’ EQ-5D scores (mean ± SD EQ-utility score: 0.74 ± 0.20; EQ-VAS score: 65.22 ± 17.54) were only slightly lower than the norm values for the general population (0.80 ± 0.26 and 79.74 ± 18.23 for the two scores, respectively). Higher income was significantly associated with better health status while lower education and unemployment status were significantly associated with poorer HRQoL. Female gender and older age were associated with higher TS. CONCLUSIONS: Diabetes had major impact on the participants’ HRQoL but little impact on health status despite a high level of TS. Patient characteristics associated with these outcomes were also different. PDB16 DEVELOPMENT OF A THAI VERSION OF THE AUDIT OF DIABETESDEPENDENT QUALITY OF LIFE (ADDQOL-19) QUESTIONNAIRE: LINGUISTIC AND PSYCHOMETRIC EVALUATION Pongmesa T1, Li SC2, Wee HL1 1 Department of Pharmacy, National University of Singapore, Singapore; 2University of Newcastle, Callaghan, NSW, Australia OBJECTIVES: To develop a Thai version of the Audit of Diabetes-Dependent Quality of Life (ADDQoL-19) and to evaluate its psychometric properties among Thai patients with diabetes. METHODS: This study was conducted by firstly cross-culturally adapting the original English ADDQoL-19 into Thai, followed by administering the Thai version together with the Thai EQ-5D and sociodemographic questionnaire to patients (aged >18 years) with diabetes recruited from a tertiary hospital in central Thailand by simple random sampling. The usefulness of a ‘not applicable’ (N/A) option and the necessary of weighting the impact scores, factor structure and internal consistency reliability of the Thai ADDQoL-19 were assessed. Data were subjected to confirmatory factor analysis (CFA) using structural equation modeling (SEM) and exploratory factor analysis (EFA) using principal component analysis (PCA). The known-groups validity of the Thai ADDQoL was assessed by four a priori hypotheses. RESULTS: The Thai ADDQoL-19 was shown to be conceptually equivalent to its original version and culturally relevant to Thai patients with diabetes. Its psychometric properties were evaluated among 201 diabetic patients (71.1% females; 81.6% primary educated; mean age 54.25 [SD = 9.8] years). Its unique characteristics, i.e., weighting and individualization were proved to be necessary for a more accurate assessment of quality of life. The Thai ADDQoL-19 also demonstrated a good internal consistency reliability, with the Cronbach’s alpha of 0.90. However, the fit indices derived from a CFA failed to achieve a good fit (i.e., >0.90). EFA with forced-one factor solution showed factor loadings above 0.40 for all items, except for sexual relationship. The known-groups validity was also not clear as only two hypotheses were fully fulfilled. CONCLUSIONS: The Thai ADDQoL-19 has shown to be culturally appropriate and highly reliable but requires further investigations to provide more evidence on its factor structure and known-groups validity before its utilization in routine clinical settings. PDB17 EVALUATING THE CHOICE OF PEN VERSUS SYRINGE: IMPLICATIONS FOR LONG-TERM DIABETES MANAGEMENT Brod M1, Thomsen TL2, Christensen T2 1 The Brod Group, Mill Valley, CA, USA; 2Novo Nordisk A/S, Virum, Denmark OBJECTIVES: The two most commonly used insulin delivery devices, pens and syringes, have comparable safety profiles and efficacy under similar use conditions. Therefore, the decision regarding the preferred device must be based on other considerations such as treatment satisfaction, treatment adherence, health-related quality of life (HRQoL), health-care expenditure and utilization. METHODS: A literature search was conducted of published articles in peer reviewed journals listed in the US National Library of Medicine’s PubMed (1948–present), DIALOG including PsychInfo (1887– present), EMCARE (nursing lit., 1995–present), BIOSIS (1993–present), EMBASE (Excerpta Medica, 1974–present), SciSearch (1990–present) and MEDLINE. Relevant manuscripts were classified into subcategories based on the major focus of the article reviewed and then subcategories were collapsed into major categories: treatment satisfaction, treatment adherence, HRQoL, and health-care expenditure and utilization. RESULTS: A total of 135 publications were reviewed for inclusion of which 73 manuscripts and nine conference abstracts/poster were deemed relevant and reviewed in depth. Seventeen reported findings from clinical trials and 13 were review articles. There is substantial evidence demonstrating the superior benefits of pens over syringes. Treatment satisfaction and patient preference for pens over syringes was widely reported and patients reported that pens provided them with increased confidence in dose, ability to self-inject, superior convenience and overall ease of use. Pen users reported better treatment adherence and fewer missed insulin doses which may account for the reported better HbA1c control than with syringe. Health-care expenditure and utilization were higher in the short term for pens; however, these expenditures were minor in comparison with long-term savings that result from improved decreases in hospitalizations, incidence of side effects, physician visits and emergency department visits. CONCLUSIONS: Long- and short-term diabetes management should benefit from the choice of pen as the insulin delivery method based on PRO outcomes and long term cost effectiveness.

A529 PDB18 THE PREVALENCE AND IMPACT OF BEING AT-RISK FOR DIABETES ACROSS US, JAPAN, AND EUROPE DiBonaventura MD1, Iwamoto K2, Nakahara N3, Fukuda T4, Nagae T5, Wagner JS1, Stankus A6 1 Kantar Health, New York, NY, USA; 2Eli Lilly Japan K.K., Chuo-ku, Kobe, Japan; 3Eli Lilly Japan K.K., Kobe, Japan; 4Tokyo University, Tokyo, Japan; 5York Pharma KK, Tokyo, Japan; 6Kantar Health, Princeton, NJ, USA OBJECTIVES: As diabetes rates rise, it is important to understand the profile of those who are at-risk for developing diabetes but have not yet been diagnosed. The aim of this study was to assess the burden for these high-risk patients. METHODS: The current study used US 2009, EU 2008, and Japan 2008 datasets from the National Health and Wellness Survey (NHWS). Implementing a previously-developed algorithm, all patients not diagnosed with diabetes were classified as having or not having a high risk for developing diabetes. High-risk patients were compared with controls on health-related quality of life (physical component summary (PCS) scores of the SF-12v2) and the number of emergency room (ER) visits controlling for demographics (country, age, gender, ethnicity, income, education) and patient characteristics (BMI and Charlson comorbidity index). RESULTS: Of patients not diagnosed with diabetes, 16.64% (n = 10778) in the United States, 14.33% (n = 7481) in the EU, and 4.05% (n = 750) in Japan were classified as being at high risk for diabetes. After controlling for demographics and patient characteristics, those at high risk reported significantly lower levels of PCS across all three countries (Adjusted Mean [Madj] = 46.2 vs. 49.6, P < 0.0001) than controls. The effect was stronger in the EU (M-adj = 45.5 vs. M-adj = 49.2) and weaker in Japan (M-adj = 47.9 vs. M-adj = 49.8) relative to the United States (M-adj = 46.6 vs. M-adj = 49.7). Those at high risk reported significantly fewer ER visits (Madj = 0.12 vs. M-adj = 0.16, P < 0.0001) than controls, however, this effect was reversed in Japan in that those at high risk reported significantly more ER visits than controls (M-adj = 0.21 vs. M-adj = 0.12, P < 0.0001). CONCLUSIONS: A substantial number of patients are at a high risk for developing diabetes, particularly in the United States and EU. The physical quality of life was significantly lower, particularly in the EU and the United States, and in Japan this group reported significantly more ER visits than controls. PDB19 HEALTH-RELATED QUALITY OF LIFE AND GLYCEMIC CONTROL IN PATIENTS WITH TYPE 2 DIABETES MELLITUS Shim YT1, Ko Y1, Lee JYC2, Tang WE3, Toh MPHS3 1 National University of Singapore, Singapore; 2National University of Singapore, Singapore; 3 National Healthcare Group Polyclinics, Singapore OBJECTIVES: Diabetes Mellitus (DM) can reduce patients’ physical well-being and compromise other aspects of quality of life. This study aims to investigate the healthrelated quality of life (HRQoL) of DM patients under primary care, and the associations of HRQoL with demographic factors and HbA1c control. Previous studies that examined these associations yielded mixed results and none was done in Singapore. METHODS: A cross-sectional, interviewer-administered survey was conducted in either English or Chinese, depending on the patient’s language of preference, between September and December 2009 in a convenience sample of 301 DM patients who were ≥21 years old, on hypoglycemic agent(s), with at least a year of DM history, and followed up at two outpatient clinics. Survey instrument included demographic questions and one generic and one DM-specific HRQoL measures, i.e., EQ-5D and the Audit of Diabetes-Dependent Quality of Life (ADDQoL), respectively. Patients’ most recent HbA1c values were extracted from medical records in the clinics. RESULTS: The average EQ-5D score was 0.89 ± 0.16 and ADDQoL average weighted impact (AWI) score was −2.5 ± 2.0. Pain was the most common complaint among the five EQ-5D domains (by 28% of the respondents). Lack of freedom to eat was the ADDQoL domain with the greatest negative impact (AWI = −3.8, on a scale of −9 to 3). A significant association was observed between ADDQoL AWI and both duration of diabetes (P = 0.005) and use of insulin (P = 0.014). A negative correlation was found between HRQoL and HbA1c values in both HRQoL measures (both r = −0.2, P < 0.01). CONCLUSIONS: DM patients in primary care settings frequently reported problems with pain and restrictions in diet. Poorer HRQoL was found to be associated with higher HbA1c, which indicates poorer glycemic control. The chronicity of DM and use of insulin may have negative impact on patients’ HRQoL. PDB20 OLDER WOMEN WITH DIABETES: CARDIOVASCULAR MEDICATION USE AND QUALITY OF LIFE Hikmayani NH1, Robertson J2, Dolja-Gore X3 1 Sebelas Maret University, Solo, Central Java, Indonesia; 2The University of Newcastle, Waratah, NSW, Australia; 3Research Centre for Gender, Health and Ageing, Newcastle, NSW, Australia OBJECTIVES: To examine the association between patterns of cardiovascular medication use and health-related quality of life (HRQoL) in elderly Australian women with diabetes. METHODS: A cross-sectional study was conducted using older cohort completing Survey 4 in the Australian Longitudinal Study on Women’s Health (ALSWH) who self-reported to have been diagnosed with diabetes and/or use medications indicative of diabetes. Use of cardiovascular medications was identified from self-reported medications belonging to the Anatomical Therapeutic Chemical (ATC) code B01AC, C03, C07, C08, C09, C02 and C10. Usage patterns were classified as no use, using any hypertensives, lipid lowering agents or antiplatelet drugs, using any two classes, and using all the three classes. The SF-36 was used to measure HRQoL, focusing on the

A530 physical functioning (PF), general health (GH), vitality (VT) and mental health (MH) subscales. Multiple linear regressions were used to assess the association between usage patterns, HRQoL scores and other sociodemographic, health behavior and health service utilization characteristics. RESULTS: Of 7158 women retained at Survey 4, 885 were identified as having diabetes with overall prevalence of cardiovascular medication use of 89.9%. After adjusting for other covariates, being on three classes of cardiovascular medications was significantly associated with increased scores on PF (coefficient 16.134, 95% CI 6.940–25.327), GH (10.058, 95% CI 2.649–17.468) and MH (12.896, 95% CI 6.562–15.882) subscales. Being on any two classes was associated with increased scores on PF and GH (coefficient 14.744, 95% CI 5.988–23.501 and 8.334, 95% CI 1.200–15.467, respectively), whereas using single cardiovascular drug was only significantly associated with increased scores on PF (coefficient 12.346, 95% CI 3.943– 20.750). CONCLUSIONS: Intensive use of cardiovasular medications to support diabetes management was shown to be subjectively beneficial. Concern about perceived deterioration on HRQoL attributable to pill burden might be captured if a diabetesspecific instrument was employed. PDB21 IMPACT ASSESSMENT OF CBIA-DM STRATEGY ON DIABETIC PATIENTS’ QUALITY OF LIFE Hartayu TS1, Suryawati S2, Ibrahim MIM3 1 University Sains Malaysia, Penang, Malaysia; 2Gadjah Mada University, Yogyakarta, Indonesia; 3 Universiti Sains Malaysia, Penang, Malaysia OBJECTIVES: CBIA is a small group discussion which emphasizes on the active role of participants in looking for information. CBIA has been proven effective to improve tuberculosis patient’s adherence to treatment program, pharmacy assistant’s skills in hypertension drug information service, and skills in early detection of Brest cancer. Therefore, CBIA was developed adapting the original CBIA, enriched with key messages in healthy life-style for diabetic patients as CBIA-DM. This study was aimed to assess the impact of CBIA-DM on diabetic patients’ quality of life. METHODS: Time series, pre and post quasi-experimental design. Three groups were involved in the study: CBIA-DM, DM-club and normal-care group. Data were collected in preintervention, immediately, one, three and 6 months post intervention using WHOQOL-bref questionnaires Bahasa version. Categorizing scores in good (>mean + SD), fair (between mean + SD and mean − SD), and poor ( 8%) for each diabetes-related vascular disease was calculated using Kaplan–Meier analysis. Hazards Ratio (HR) was adjusted for age, gender, ethnicity, BMI, duration of diabetes, baseline HbA1c, LDL-cholesterol and insulin use using Multivariate Cox analysis. RESULTS: There were 7739 T2DM patients, with median age of 68 years and baseline HbA1c of 6.4%. Over a quarter (28.9%) of patients reached “poor” control at the end of 4 years, with a median time of 575 days (95% CI: 550–600). Multivariate Cox regression analysis suggested that likelihood of “poor” control was increased with each additional diabetes-related vascular disease (AdjHR 1.14, 95% CI: 1.07–1.23) and higher for eye complications (AdjHR 1.47, 95%CI: 1.17–1.86) and nephropathy (AdjHR 1.43, 95% CI:1.06–1.93). CONCLUSIONS: Despite starting with “good” glycemic control, 3 out of 10 T2DM patients progressed to “poor” glycemic control at the end of 4 years of study. Patients with more diabetes-related vascular diseases were at higher risk of progressing to “poor” control, particularly those with microvascular complications. PDB25 DIABETES-RELATED COMPLICATIONS AND GLYCEMIC CONTROL IN TYPE 2 DIABETIC MELLITUS PATIENTS AT SPECIALIST OUTPATIENT CLINICS IN SINGAPORE Lim BK1, Toh M1, Jong M2, Chionh SB3, Sum CF4 1 National Healthcare Group, Singapore; 2Tan Tock Seng Hospital, Singapore; 3National University Hospital, Singapore; 4Alexandra Hospital, Singapore OBJECTIVES: Type 2 diabetes mellitus (T2DM) patients often have co-existing hypertension and dyslipidemia and some develop retinopathy, peripheral vascular disease (PVD), stroke, nephropathy and coronary heart disease (CAD). This study investigates the prevalence of these conditions which are associated with T2DM and the association with optimal glycemic control (HbA1c up to 7%). METHODS: This is a crosssectional study. We included all T2DM patients who were treated at the specialist outpatient clinics (SOCs) in three public-sector hospitals in Singapore during January 2009 and were on follow-up at the same clinic for at least 12 months. All data were extracted from the National Healthcare Group (NHG) Diabetes Registry (CDMS), including the medical diagnosis and HbA1c result at the last visit. We compared the distribution of glycemic control for each diabetes-related condition and performed logistic regression for “optimal” glycemic control adjusting for age, gender and ethnicity. RESULTS: Of the 3420 T2DM patients, 2919 (85%) with HbA1c results were analyzed. Mean age was 63.2 years, with 53% female and 66% Chinese, high prevalence of dyslipidemia (92.7%) and hypertension (74.5%). About 1249 (42.8%) patients had “optimal” glycemic control. CAD (28.7%) was the most common condition, followed by PVD (22.5%), retinopathy (21.8%), nephropathy (14.4%) and stroke (10.7%). The Malay group had the highest proportion of hypertension

A531

4th Asia-Pacific Abstracts (78.8%), retinopathy (27.9%) and nephropathy (22.2%); Indian were highest in PVD (26.8%), and CAD (34.4%); Chinese highest in stroke (12.1%). After adjusting for age, gender and ethnicity, the odds of having “optimal” glycemic control was higher for T2DM patients with nephropathy (Adj OR:1.50; 95% CI [1.19–1.89]) and stroke (Adj OR:1.51; 95% CI [1.17–1.96]) compared to PVD (Adj OR: 0.61; 95% CI [0.51–0.74]) and retinopathy (Adj OR: 0.65; 95% CI [0.53–0.79]). CONCLUSIONS: There were ethnic differences in the prevalence of diabetes-related complications. T2DM patients with stroke or nephropathy were more likely to achieve “optimal” HbA1c level than those with PVD or eye complications. PDB27 A REVIEW OF HEALTH TECHNOLOGY ASSESSMENTS OF DIABETES IN ASIA AND OCEANIA Adalsteinsson JE1, Hemels M2, Jensen R3, Toumi M4 1 Novo Nordisk A/S, Copenhagen, Denmark; 2Novo Nordisk, Søborg, Denmark; 3Novo Nordisk, Bagsvaerd, Denmark; 4University of Lyon, Lyon, France OBJECTIVES: Health technology Assessment (HTA) has a long history in Australia and NZ and is rapidly evolving in Asian countries. The aim of this research was to identify and review English written reports on diabetes among HTA agencies in Asia and Oceania. METHODS: ISPORs directory (http://www.ispor.org) listing worldwide HTA organizations were searched for agencies in Asia and Oceania with a valid link to a webpage were included. On each agencies webpage a search was done using the search term “diabetes” and English written assessments were included for further review. RESULTS: The following HTA agencies were searched: PBAC, MSAC ASERNIP-S, AHTA, CCE, CHERE, HTA, NIPH, CRECON, MAPI, PHIC, PKEBM, ICTACH, HSAC, NZHTA, HIRA, CDE and HITAP. Of those, only 14 relevant reports were found from Australia, NZ and Thailand. Two were found from PBAC, five from AHTA, one from HSAC, two from NZHTA, and four from HITAP. The two reports from PBAC were describing how to improve National management of diabetes. The reports from AHTA were horizon scanning reports covering glucose monitoring, screening of gestational diabetes and detecting retinopathy. HSAC had a report on insulin pump whereas as NZHTA had one on glucose monitoring and one about clinical features detecting insulin resistance. The reports found on HITAP were in the form of published articles assessing cost effectiveness of two different antihypertensive therapies in diabetes, one assessing cost effectiveness of diabetes management and finally one assessing diabetes self management. Reports from HSAC, NZHTA and HITAP contained economic modeling with QALY as outcome. CONCLUSIONS: Despite of broad search criteria only 14 English written reports were found on diabetes. A headline in English on each assessment would facilitate information sharing and make it easier to evaluate the need for translational support. PDB28 THE EFFECT OF METFORMIN BRAND SHIFT BY HOSPITAL ON ANTIDIABETICS UTILIZATION Chen HY1, Chang HR1, Lang HC2 1 Mackay Memorial Hospital, Taipei, Taiwan; 2National Yang-Ming University, Taipei, Taiwan OBJECTIVES: Due to the effect of National Health Insurance drug price cut policy, the studied hospital has changed the manufactures of metformin four times in 5 years since July, 2004. The purpose of this study is to investigate the effects of the change in hospital drug brands on patients’ utilization and the factors that influence patients’ persistence on the use of metformin. METHODS: Data were sourced from a medical center in Taipei city during 2004–2008. We collected all the oral antidiabetic drugs used during the study period and classified them into five categories: Biguanid, Sulfonylures, α-glucosidase inhibitor, Thiazolidinediones, and Meglitinide. The quantity of all the antidiabetics used was transformed by using WHO/ATC into DDDs. In this study, Metformin was treated as the experimental group and all other antidiabetics drugs were the control groups. RESULTS: The results showed that the quantity of metformin prescribed was increasing during the study period in spite of the brand change. However, other categories of oral antidiabetes were also increasing due to the increase in number of patients. The percentage of all types of drug stayed relatively steadily, in which metformin itself occupied 21–23%, all Sulfonylureas drugs occupied 59–60%, α-glucosidase inhibitor occupied 3–4%, Thiazolidinediones occupied 10–11%, and Meglitinide occupied 4–5% during the study period. In addition, the results of logistic regressions showed that female patients with hypertension or hyperlipemia between the ages of 40~80 who live in Taipei city performed better persistence of seeking medical service. CONCLUSIONS: The results shows changing the brand of metformin did not decrease drug consumption and patient numbers. By analyzing the acceptance of generic drugs of patients and doctors, this study provides important information for hospital administrators and health administrators. PDB29 STATUS AND TREATMENT OF TYPE 2 DIABETES IN CHINA 1 1 2 Vlachopioti Z , Smith HT , Colclough H 1 Eli Lilly & Co, Windlesham, Surrey, UK, 2Adelphi, Macclesfield, Cheshire, UK OBJECTIVES: To explore current disease status and approach to care for patients with type 2 diabetes (T2DM) in China. METHODS: The study was performed using data from the Adelphi T2DM Disease Specific Programme (DSP©), a cross-sectional study of consulting patients providing insights into “real-world” behaviors and attitudes in clinical practice in 2008. The DSP collected data on 1648 patients with T2DM receiving at least one oral anti-diabetic with or without insulin and from 200 physicians (140 specialists and 60 internal medicine) across general hospitals in nine Chinese cities through physician interviews, patient record forms and patient-

completed questionnaires. RESULTS: Of 1648 patients (50% male, mean [SD] age: 57.74 [12.56] years, BMI: 24.04 [3.37] kg/m2, duration of diabetes: 4.52 [5.06] years), 74% had at least one comorbidity, 39% were overweight (24 ≤ BMI < 28 kg/m2) and 9% were obese (BMI ≥ 28 kg/m2). Sixty-one percent had an HbA1c ≥ 7% and only 35% of patients were aware of an HbA1c target. 39.20% of patients were treated with OAD monotherapy (metformin [Met]: 12.26%, Sulphonyureas [SU]: 12.56%, alpha-glucosinase inhibitor [AGI]: 8.37%, prandial glucose regulators [PGR]: 4.67%, thiazolidinediones [TZD]: 2.31%) and 29.67% were on dual OAD therapy (commonly Met + SU: 11.71%, Met + AGI: 4.37%, SU + AGI: 3.82%, Met + PGR: 4.25%, Met + TZD: 1.52%), while 28.09% of patients were on OAD + insulin therapy. Lack of gastrointestinal, liver function side effects and hypoglycemia risk, QD dosage, and beneficial effects on blood pressure were the five most cited unmet needs in non-insulin diabetes treatment highlighted by physicians. Mean (SD) patient-reported QoL (EQ-5D VAS) was 0.86 (0.18) and mean (SD) diabetes treatment satisfaction was 25.97(5.14)/36 where 0 is very dissatisfied and 36 very satisfied. CONCLUSIONS: The study population of Chinese T2DM patients is relatively young and recently diagnosed. More than half were poorly controlled and the need for further therapeutic actions to improve glycemic control as well as better doctor/patient communication regarding treatment goals is clear. PDB30 GLYCEMIC CONTROL IS WORSE AND COMPLICATIONS MORE PREVALENT IN TYPE 2 DIABETES PATIENTS WITH A HIGHER BMI IN CHINA Smith HT1, Vlachopioti Z1, Colclough H2 1 Eli Lilly & Co, Windlesham, Surrey, UK; 2Adelphi, Macclesfield, Cheshire, UK OBJECTIVES: Examine BMI and achievement of diabetes treatment targets in patients with Type 2 Diabetes (T2DM). METHODS: The study used the 2008 Adelphi T2DM Disease Specific Programme© a cross-sectional study of consulting patients providing insights into ‘real-world’ behaviors and attitudes in clinical practice. Data was collected on 1648 patients with T2DM receiving at least one oral anti-diabetic with/ without insulin and 140 specialists and 60 internal medicine physicians distributed across general hospitals in nine Chinese cities through physician interviews, patient record forms and self-completed questionnaires. RESULTS: Across the 1648 patients in the study; mean (SD) age was 57.7 (12.6) years, 46% were ≤2 years from diagnosis and mean (SD) BMI was 24.0 (3.4) kg/m2. Nine percent of patients were obese (BMI ≥ 28 kg/m2), 39% overweight (24 ≤ BMI < 28 kg/m2) and 51% normal/underweight (BMI < 24 kg/m2). Prevalence of comorbidities in this population was high; 49% reporting hypertension and 42% dyslipidemia (25% patients reported both). Glycemic control appeared worse in patients with a higher BMI. Twenty-two percent of patients with a BMI ≥ 28 had an HbA1c ≥ 9% and 39% had an HbA1c ≥ 8% compared to 10% of patients with a BMI between 19–27.9 having an HbA1c ≥ 9% and 22% having an HbA1c ≥ 8%. Only 35% of patients were aware of having an HbA1c target set by their physician. Comorbid conditions were more prevalent in overweight and obese patients. Among normal weight patients 31% reported no comorbid conditions compared to 22% of overweight patients and 17% of obese patients. Prevalence of hypertension, dyslipidemia and other macrovascular conditions were 44%, 36%, and 18% respectively among normal weight patients; 53%, 48%, and 19% among overweight patients and 65%, 55%, and 24% among obese patients. CONCLUSIONS: The results indicate that obese patients report higher HbA1c and higher rates of complications compared to the normal weight patients and strategies to improve target better outcomes could include managing obesity.

GASTROINTESTINAL DISORDERS – Clinical Outcomes Studies PGI2 A SYSTEMATIC REVIEW AND META-ANALYSIS OF LITERATURE ASSESSING CLINICAL EFFICACY OF TREATMENTS FOR PATIENTS WITH HBEAG-POSITIVE CHRONIC HEPATITIS B Tantai N1, Chaikledkaew U2, Werayingyong P3, Teerawattananon Y3 1 Social Administrative Pharmacy Division, Department of Pharmacy, Faculty of Pharmacy, Mahidol University, Bangkok, Thailand; 2Division of Social and Administrative Pharmacy, Bangkok, Thailand; 3Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: An objective of this review was to assess clinical efficacy in terms of “HBeAg seroconversion” of treatment options for HBeAg-positive chronic hepatitis B (CHB). METHODS: A systematic review of randomized controlled trials (RCTs) of treatments for patients with HBeAg-positive CHB was performed through the Medline and Cochrane databases. The clinical studies were included only if they assessed the efficacy among the following treatment options namely 1) lamivudine; 2) entecavir; 3) interferon; and 4) interferon plus lamivudine. Indirect or mixed-treatment comparison meta-analysis with random effect model was employed to combine results of several studies. The meta-analysis was carried out using the WinBUGS14 software. Odds ratio (OR) and its 95% confidence interval (CI) were presented. Heterogeneity test was applied for testing the variation of study outcomes between studies. RESULTS: There were 115 abstracts reviewed with eight relevant RCTs included in the analysis. None of eight RCTs included all four treatment options. Five studies compared HBeAg seroconversion between entecavir and lamivudine. There were three different studies comparing interferon with interferon plus lamivudine, interferon plus lamivudine with lamivudine, and interferon and interferon plus lamivudine with lamivudine. Indirect

A532 comparison was applied to compare entecavir with interferon and the combination of interferon and lamivudine. When compared with lamivudine, the combination of interferon and lamivudine yielded the best efficacy which was about two times more likely to increase HBeAg seroconversion rate (OR = 2.38, 95% CI = 1.21–4.06) than entecavir (OR = 0.98, 95% CI = 0.56–1.44) and interferon (OR = 1.17, 95% CI = 0.44–2.24). In addition, when compared with either interferon or entecavir, interferon plus lamivudine was about two to three times more likely to enhance HBeAg seroconversion rate with the OR of 2.48 (95% CI = 1.05–4.92) or 2.71 (95% CI = 1.13–5.33), respectively. CONCLUSIONS: There was a significant increase in HBeAg seroconversion rate in patients with HBeAg positive CHB receiving the combination of interferon and lamivudine compared with lamivudine, entecavir and interferon. PGI3 EFFECTIVENESS AND COST ANALYSIS OF PARENTERNAL REGIMEN IN CRITICAL ILLNESS PATIENTS OF POSTOPERATIVE: THREECOMPARTMENT BAG SYSTEM AND TRANDITIONAL SEPARATE BOTTLE SYSTEM Hsieh HJ, Chan AL Chi-Mei Medical Center, Tainan, Taiwan OBJECTIVES: To evaluate clinical outcome and direct medical cost using Kabiven and traditional peripheral partenteral nutrition for postoperative patients in ICU. METHODS: This is a restrospective study. Medical records of postoperative patients in ICU were reviewed by clinical pharmacists from July 2008 to July 2009. The retrieved patients were divided into two groups, kabiven group (n = 49) and separate bottle group (SB) (n = 50). Patients, characteristics were evaluated and compared between two groups. The measured outcomes were the clinical effectiveness and direct medical costs. One way ANOVA were used for analysis. RESULTS: There were no statistically significance difference in patients, characteristics between two groups. Survival rate was higher in kabiven group than in SB group (40% vs. 31%, respectively). The length of stay in hospital in kabiven group was longer than in SB group (47.51 ± 38.63 vs. 31.86 ± 21.99 days, respectively; P < 0.015). Total direct medical costs of survivors in kabiven group was more expensive than in SB group ($11,976.9 vs. $9574.9, respectively). CONCLUSIONS: Although the direct medical cost of kabiven group were higher than SB group, use of kabiven is likely to improve mortality rate of postoperative patients in ICU.

GASTROINTESTINAL DISORDERS – Cost Studies PGI4 BUDGET IMPACT ANALYSIS OF ORAL ANTIVIRAL AGENTS FOR THE TREATMENT OF CHRONIC HPATITIS B IN SOUTH KOREA Na JH, Lee EK Sookmyung Women’s University, Yongsan-gu, Seoul, South Korea OBJECTIVES: Heptatitis B is prevalent in South Korea and chronic hepatitis B (CHB) infection is an important public health issue due to its potential to evolve to cirrhosis, hepatocellular carcinoma. This study estimated the direct medical cost of CHB-related disease states in South Korea and compared the cost of South Korea with that of USA, Australia and China. It also aimed to analyze the impact of three therapeutic alternatives for CHB by Budget Impact Analysis (BIA). METHODS: Dynamic budget impact analysis was conducted based on a Markov model for 5 years. Three treatment scenarios were selected as follows: first-line treatment of lamivudine, second-line combination treatment of lamivudine and adefovir on the development of drug resistance, first-line treatment of lamivudine, second-line treatment of entecavir 1.0 mg on the development of drug resistance, first-line treatment of entecavir 0.5 mg, second-line treatment of adefovir on the development of drug resistance, no treatment available. RESULTS: The BIA results of scenario A, B, C and no treatment were 75, 74.7, 85.9, and 48.1 billion Korean Won (KRW), respectively. The results were relatively insensitive to the TP and sensitive to the number of treated patients based on sensitivity analyses. The costs of annual direct medical costs in South Korea were 23.2%~65.8%, 16.2%~59.1%, and 75.8%~381.7% of the annual direct medical costs in the United States, Australia, and China. CONCLUSIONS: Scenario C (first-line treatment of entecavir 0.5 mg, second-line treatment of adefovir) was found to be 10.9~11.2 billion KRW more expensive than scenario A and B from payer’s perspective. In South Korea, the direct medical costs of CHB-related diseases are cheaper than the United States and Australia. It suggests that such factors as the difference of GNP (Gross National Product), health-care system and others contribute to the difference of the direct medical costs. PGI5 EVALUATION OF OCTREOTIDE COST AFTER PHARMACEUTICAL CARE IMPLEMENTATION AT SURGICAL WARD Sawatpanit A Department of Pharmacy, Roi-et Hospital, Muang, Thailand, Thailand OBJECTIVES: Our objective was to compare cost saving before and after of pharmaceutical care implementation and Octreotide use evaluation at surgical ward. METHODS: This study was conducted during October 1, 2008 to April 30, 2009. The research instruments were octreotide use criteria with approval by surgical staffs. Pharmacist reviewed and evaluated the medical chart in order to identify medication related problems by focusing on medication regimens. Medication related problems were classified into three categories: inappropriate route, inappropriate duration and inappropri-

4th Asia-Pacific Abstracts ate indication of administration. Cost saving was based on drug costs. The pre- and post-pharmaceutical care implementation and octreotide use evaluation results were compared and analyzed with descriptive retrospective statistics and paired t–test at 95 % significant level. RESULTS: There were 302 patients received Octreotide. One hundred ninety-three patients (63.9%) were prescribed Octreotide appropriately. Medication related problems were found in 109 patients (36.1%). The most frequent problems were inappropriate duration (44.95%), inappropriate indication (34.86%) and inappropriate route of administration (20.18%). The cost of Octreotide use for inappropriate duration, inappropriate indication and inappropriate route of administration were significantly decreased after pharmaceutical care implementation 3,062,160 versus 1,946,400 baht, 657,000 versus no money, and 919,800 versus 226,800 baht, respectively (P < 0.005). The total cost of Octreotide use was 9,529,950 baht reducing to 7,015,008 baht after pharmaceutical care implementation with statistically significant (P < 0.005). CONCLUSIONS: The implementation of pharmaceutical care and evaluation of Octreotide use resulted in significant cost saving. The drug utilization program, consisting of deriving quality criteria for prescribing, structured order form and good cooperation between physicians and pharmacists with strong support of the therapeutics committee, was an effective strategic approach to promote rational drug use and develop Octreotide use guidelines.

GASTROINTESTINAL DISORDERS – Patient-Reported Outcomes Studies PGI7 THE EFFECT OF ANTIVIRAL THERAPY ON QUALITY OF LIFE IN CHRONIC HEPATITIS PATIENTS: A SYSTEMATIC REVIEW Pansang S, Maphanta S Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: Chronic viral hepatitis B and C are treated primarily with interferon (IFN) and/or nucleotide analogs e.g., ribavirin (RBV) which produced sustained viral response in more than 50% of treated patients. However, the combination may also cause numerous side effects that could reduce patients’ quality of life (QoL). To determine the effect of antiviral therapy on QoL of chronic hepatitis patients. METHODS: Key words were “Quality of life” AND “chronic hepatitis” AND “antiviral therapy,” “lamivudine,” “ribavirin,” “adefovir,” “entacavir,” “telbivudine” “tenofovir” and were searched from PubMed and EMBASE database. Study selection criteria were original articles in which patients received antivirals for chronic hepatitis B or C, and assessed QoL. The comparators were no treatment, placebo or at least one other antiviral. Studies that involved co-HIV infection were excluded. One independent researcher reviewed titles and abstracts to determine relevance. Study design, dose, duration, baseline values, and QoL scores were extracted. RESULTS: Six studies met all the criteria. Four QoL instruments were used in those studies. SF 36 and HQLQ questionnaires were the most common. Patients who received peginterferonμ2a plus placebo reported better QoL than peginterferonμ-2a plus ribavirin during the treatment (week 2–48) in two RCTs. However, the results were less consistent and less prominent during the week of 72 and 96. The effect seemed to be non-dose related. One RCT showed that interferon μ plus ribavirin produced better QoL than untreated when measured with EQ 5D questionnaire, but it did not reach statistical significant. In similarly, peginterferone μ-2a plus ribavirin did not showed a significant better Fatigue severity scale at week 72 when compared with peginterferonμ-2a plus placebo. CONCLUSIONS: Antiviral therapy (interferon plus ribavirin) reduced QoL only during the treatment. Results seemed to be consistent across all four types of QoL questionnaires.

GASTROINTESTINAL DISORDERS – Health Care Use & Policy Studies PGI8 ANTIBIOTIC PRESCRIBING PRACTICES OF PRIMARY CARE PRESCRIBERS FOR DIARRHEA NI NEW DELHI, INDIA Kotwani A1, Roy Chaudhury R2, Holloway K3 1 Vallabhbhai Patel Chest Institute, Univ of Delhi, Delhi, India; 2Indraprastha Apollo Hospitals, New Delhi, India; 3World Health Organisation, Geneva, Switzerland OBJECTIVES: This study was conducted to obtain information on current prescribing rates of antibiotics in diarrhea, a condition where misuse of antibiotics is common. In the absence of community-based databases on antibiotic use in developing countries recently a methodology was established for surveillance of antibiotic use at New Delhi by conducting ‘Exit Interviews’ of the patients. METHODS: Antibiotic use data was collected from public and private sector facilities from four residential localities around a tertiary care hospital where the antibiotic resistance work was being conducted. All the 10 public sector facilities (eight primary and two secondary health cares) situated in the study area under Delhi government were enrolled. For private sector, 20 willing and cooperative general practitioners and specialists practicing in the chosen areas were selected. Patients after consultation with prescriber were asked if they had diarrhea but without blood. Patients with diarrhea were enrolled for exit interview and his/her prescription was monitored. Antibiotic use data was collected per month over 1 year (December 2007–November 2008). The percentage of patients receiving antibiotic and pattern of consumption for various antibiotics was analyzed. RESULTS: At public facilities 43% (171/398) and at private facilities 69% (76/110) of patients with diarrhea were prescribed at least one antibiotic. Main antibiotic class that was prescribed in public and private sector was fluoroquinolones (89% and 94%);

4th Asia-Pacific Abstracts the second group in public sector was cephalosporins (4%) and in private sector were penicillins (3%). In private sector pediatricians prescribed antibiotics to 52% (17/33) of children with diarrhea and fluoroquinolone group was prescribed to all. At public facilities, main members from fluoroquinolones were norfloxacin, followed by ofloxacin and ciprofloxacin. At private clinics, it was ofloxacin followed by ciprofloxacin. Pediatricians mainly prescribed ofloxacin, followed by norfloxacin. CONCLUSIONS: This study clearly shows over-prescription and irrational use of antibiotics for treatment of diarrhea that warrants interventional strategies.

HEALTH CARE USE & POLICY STUDIES – Consumer Role in Health Care PHP1 A STUDY EVALUATING PATTERN OF NON-PRESCRIPTION PURCHASE BY CONSUMERS FROM COMMUNITY PHARMACIES IN MALAYSIA Ahmad Hassali MA, Shafie AA, Mohamad Yahaya AH Universiti Sains Malaysia, Penang, Penang, Malaysia OBJECTIVES: To analyze the pattern of non-prescription medicine purchase by consumer from community pharmacies in Malaysia. METHODS: A cross-sectional survey comprised a sample of 1799 community pharmacy consumers was conducted nationwide. A pharmacy “exit survey” was developed and administered to pharmacy consumers of randomly selected community pharmacies in order to collect information on the purchased nonprescription medicine(s) including its costs. In this study, the nonprescription medicine definition was adopted from the Malaysian Poison Act 1952. Data were analyzed using Kruskal–Wallis, Mann–Whitney, chi-square and Spearman correlation test in SPSS v15. RESULTS: A total of 2175 nonprescription medicines were purchased by consumers interviewed in 2 weeks study period. The total cost estimated for all item purchased was RM41,000 (USD13,000). About 39.6% of the purchased items are listed under scheduled poison, 45.5% were unscheduled poison and 12.5% are those listed as traditional and complementary medicine. Medicine for alimentary tract and metabolism, musculo-skeletal system and respiratory system as categorized by Anatomical Therapeutic Coding were among the highest purchased medicine. Factors such as gender especially females, area of origin especially those from urban area, ethnicity especially chinese consumers and those earning high income level shows to have a significant influence in the spending for non-prescription medicine purchasing. This study also showed purchasing for non-prescription medicine significantly increased as aging. Consumers spent significantly more on noncontrolled medicine such as vitamins and herbal preparations compared to other categories of medicines (χ2 = 185.07, P < 0.001). CONCLUSIONS: The evaluation on pattern of nonprescription medicine purchasing in Malaysia reveals that consumers in Malaysia are able to spend money for buying medicines to treat minor ailments and practice of self-medication. The socio-demographic factors that associated with non-prescription medicine purchase will serve as useful information for policymakers and also the pharmaceutical industry for future development in rational medicine use education among consumers in the country.

HEALTH CARE USE & POLICY STUDIES – Disease Management PHP3 RARE DISEASES, ORPHAN DRUGS, AND THE LEGISLATION IN CHINA 1 1 2 Zhang YJ , Guo JJ , Wang JB 1 University of Cincinnati, Cincinnati, OH, USA; 2People’s Liberation Army 309 Hospital, Beijing, China OBJECTIVES: Public awareness of rare diseases and access to orphan drugs is increasing in China. The purpose of this study was to review the epidemics of rare diseases, the access to orphan drug treatments, and its related legislation in China. METHODS: A systematic literature review was performed based on published articles, government Websites, and some Internet search engines. The rare diseases, available orphan drug treatments, and related legislations in China were reviewed. Some comparisons related to these topics were discussed between China and developed countries like the United States. RESULTS: With conservative estimation, there are at least 10 million Chinese people living with rare diseases. The frequently mentioned rare diseases in China include osteogenesis imperfecta, neuromuscular diseases, Fabry disease, Gaucher disease, phenylketonurias, hemophilia A and B, lymphangioleiomyomatosis, albinism, and acromegaly. Patients with rare diseases in China generally lack the access to appropriate health care especially the orphan drug therapies. While we observed the significant impact of Orphan Drug Act on new drug developments and rare disease treatments in developed countries, there is little new orphan drug designated or developed in China. There are very few imported orphan drugs in the Chinese market. A grouping number of Chinese rare-disease organizations such as the China Albinism Association and the China-Dolls Care and Support Association are working with government on a new legislation about health care assesses and insurance policy for rare diseases. CONCLUSIONS: Public and governmental concerns about rare diseases have been boosted in China. Their available treatments and legislation in China are lagging far behind the United States, the European Union, Australia, Singapore, Japan, and South Korea. An effective public health insurance and public policy are needed for rare disease treatments and orphan drug developments.

A533 HEALTH CARE USE & POLICY STUDIES – Drug/Device/Diagnostic Use & Policy PHP4 EFFECTIVENESS OF TWO POLICIES TO REDUCE DIPHENOXYLATE CONSUMPTION IN IRAN Jaberi Doost M1, Abdollahiasl A1, Safavi Homami N2, Farshchi A1, Ghiasi G1, Anabi M1 1 Tehran University of Medical Sciences, Tehran, Iran; 2Ministry of Health, Tehran, Iran OBJECTIVES: Iran had one of the highest uses of Diphenoxylate in the world in 2008. About 1 billion tablets of Diphenoxylate 2.5 mg have been consumed during that year. In the last months of 2008 the ministry of health tried some policies to reduce the consumption. One was rationing in production the other was doubling the price of 10 medicines which likely were abused including Diphenoxylate. This study tries to show the effectiveness of these two interventions by evaluating monthly consumption of Diphenoxylate during recent 5 years ended to March 2010. METHODS: Data were gathered from the distributors and wholesalers of this medicine from whole the country. These data were crosschecked with importation data in the ministry of health. We did a time trend analysis on tabulated data. RESULTS: There are some variations in monthly use of Diphenoxylate but the trend shows a significant decrease after the rationing in production. In 2009 the average consumption has reached to less than 650 millions of tablet, four Defined Daily Dose per 1000 Inhabitants per day which shows 36% decrease in comparison with the previous year. In second half of 2009 and first 2 months of 2010 after the second intervention, there is no significant change in the trend. CONCLUSIONS: Although the high consumption of opiates in Iran is a multi factorial phenomena but this study shows; the first intervention has been able to control the abuse of Diphenoxylate due to the reduction of market supply and decrease in Diphenoxylate’s unofficial promotion. Considerable result wasn’t seen in the second approach. The user price of Diphenoxylate tablet in Iran is too cheap and doubling the price doesn’t work at all. A significant tax mark-up and increasing the price based on label use may affect on affordability of Diphenoxylate’s abuse. PHP5 THE ESSENTIAL MEDICINE SYSTEM IN CHINA: STRATEGIES AND CHALLENGES Yang L Peking University, Beijing, China OBJECTIVES: In the 1980s, China launched market-oriented reforms. Public hospitals were encouraged to make their own incomes with the aim of mobilizing medical workers and improving hospital efficiency. Less government funding resulted in deficits for public health institutions, which forced hospitals to generate their own revenue by aggressively selling drugs, especially expensive drugs. There are challenges in poor drug access, drug procurement, rising drug costs and financial risk. In March 2009, China finally unveiled its health care reform plan. The Chinese government announced it will institute an essential medicine system within 3 years to drive down prescription costs and quell public complaints of limited accessibility of medicines. This study is to describe the strategies to build the essential medicine system at the provincial level in China and challenges associated with promoting it. METHODS: Key actors, their power, views and interactions will be assessed using appropriate qualitative methods as these are best suited to understanding complexity, richness of material and motivations of actors, and using a policy analysis to know the key governance issues and underlying relationships affecting its success or failure. RESULTS: The system includes a list of essential medicines that would be produced and distributed under government control and supervision, and it should be used at all public health facilities at grassroots levels from 2009. The central government will set reference prices, based on which, provincial governments set the purchase prices of the drugs in their jurisdiction. Public medical and health facilities at the grassroots levels should sell the drugs at the purchase prices. These generally involve the transfer for a fixed fee from the local MOH to the health service center/post in exchange for the removal of retail mark up, which couldn’t offset the profit of selling drugs. CONCLUSIONS: The provider incentive is very crucial for building an essential medicine system, which is dependent upon the reimbursement system. PHP6 RECOMMENDATIONS FOR A BIOLOGICS-SPECIFIC PRICING SYSTEM IN CHINA Yao W, Shao R, Chen Y, Chang F Research Center of Pharmaceutical Industry Development, China Pharmaceutical University, Nanjing, Jiangsu, China OBJECTIVES: The pricing system for chemical drugs is currently applied to biologics in China as no specific pricing system for biologics exists yet in China. This practice does not fully reflect the characteristics of biologics and may potentially prevent the launch of biologics in China and discourage the development of biologics. The objective of this study is to explore adequate pricing mechanisms for biologics. METHODS: Both primary research (field research, expert interviews, surveys) and secondary research were conducted to determine the key determinants of the price of biologics and understand the complexity of R&D, manufacturing and distribution required for biologics. Statistical analyses were performed to assess the relationship of key determinants and the price. Government and manufacturer behavior were modeled using game theory to demonstrate prerequisites of implementing ‘differential pricing’. RESULTS: A different pricing system is required for biologics given its unique characteristics. Certain characteristics (including special clinical value, target-specific R&D

A534 efforts, complex manufacturing process and higher distribution costs) were found to be the key influencing factors when determining biologics’ prices. Since these characteristics vary significantly depending on the type of biologics, the important ratios used in the pricing formula like “SG&A/sales ratio,” “yield rate,” “profit rate (innovationbased),” “distribution mark-ups,” “VAT rate” and “differentiation mark-up (qualitybased)” have to be determined on a case by case basis. Using blood products as an example, three differentiation factors (clinical efficacy, gold standard finish product assessment and overall manufacturing process evaluation) are key to achieve differential pricing. CONCLUSIONS: “Differential pricing” should be included in the biologics pricing system. Ratios in the existing pricing formula like “yield rate,” “distribution mark-up,” “profit rate” and “differentiation mark-up” should be adjusted to reflect the differences between biologics and chemicals. Differentiating criteria should be carefully chosen to reflect the value of biologics thus to develop the biopharmaceutical industry. PHP7 PROSPECTIVE ENCOUNTER STUDY ON PATIENT CARE INDICATORS IN HEALTH-CARE FACILITIES Hettihewa LM, Amarasinghe I, Subasinghe S University of Ruhuna, Galle, south, Sri Lanka OBJECTIVES: Our main objective was to investigate the degree of adherence of WHO recommended patient care indicators in health-care facilities operatin in Galle Sri Lanka. METHODS: A total of 422 patients attending to the Outpatient Department in selected Hospitals in Galle district in Southern Province. Average Dispensing Time (ADT), Percentage of Drugs Actually Dispensed (PDAD), Percentage of Drugs Adequately Labeled (PDAL) and Patients Knowledge on Correct Dosage (PKCD) were compared in these selected Teaching, General and District Hospitals in Galle. RESULTS: ADT in DH (1.16 min) and GH (1.07 min) were high when compared to ADT in TH (0.81 min). PDAD was 100% in DH, 97.79% in GH and lowest in TH (94.64%). PDAL was highest in 22.66% in TH, 17.57% in GH and lowest in DH (1.57%). PKCD was 100% in GH and lowest in DH (0%) and only 50 % in TH in Galle District. We noted that there was a significant difference in ADT, PKCD and average number of drugs labeled in all three categories (P < 0.05). But there was no significant difference in average no of drugs prescribed and dispensed in three categories of these hospitals. We noted that dispensers spend only short dispensing time and there was a tendency for dispensing errors. We found that PDAL was very low in all Hospitals but PDAD was significantly high. Even though the ADT was high in DH, PKCD was 0% due to unorganized dispensing practices. We also noted 100% PKCD only in GH due to practice of well prepared correct labeling system. CONCLUSIONS: Therefore we suggest to implementing proper dispensing techniques by introducing well prepared printed drug labeling system and commencement of recorded announcements to improve patient’s knowledge on drug administration. PHP8 DRUG USE PATTERN IN OVER 60-YEAR-OLD POPULATION OF TEHRAN IN 2007 IRAN Soleymani F Tehran University of Medical Science, Tehran, Iran OBJECTIVES: Elderly are prone to many acute and chronic age related diseases which needed multiple drug usage. Therefore medicine use evaluation pattern in this group is very important and a significant part of public health system policy METHODS: In a cross sectional household survey, 20 clusters were selected from the post office sampling frame in Tehran. Each cluster consists of 40 households who had at least 1 person over the 60 years old. Questionnaire of the study was completed for all of the members over 60 and information about their medicine consumption was gathered. Data was analyzed by the prescription analyzer and SPSS soft wares. RESULTS: A total of 1054 men and women over 60 were enrolled to the study; 49.4% were male and 50.6% were female; 69% of the study population used at least one medicine per day regularly and 7.6% used at least one medicine none daily regularly; 46.2% used to consume medicines as self-medication. Mean items per prescription was 3.57 for patients who used medicines regularly everyday. In this group of patients, 58% used more than two medicines per day (poly pharmacy). Cardiovascular and central nervous system medicines were the most frequent medicines. Aspirin 80 mg, Atenolol and Calcium-D were the most frequent medicines used regularly but not daily. Acetaminophen 325 mg and acetaminophen codeine were the most frequent self medications. CONCLUSIONS: Our study showed that mean drug use in elderly is nearly same the mean of total population. This group of society is very vulnerable because of aging processes and multiple comorbidities. They must take many different drugs of variable groups which have interaction with each other. For minimizing these effects, physicians must be very precise in taking drug history of these patients before the drugs prescription. PHP9 PUBLIC PREFERENCES TOWARD ORAL SOLID DOSAGE FORMS 1 2 1 Ibrahim IR , Izham M , Al-Haddad MS 1 Universiti Sains Malaysia, Gelugor, P.Penang, Malaysia; 2Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia OBJECTIVES: To investigate and cluster the Universiti Sains Malaysia (USM) market in terms of the preferred OSDF and the form that seems easy in swallowing; to determine how consumers manage their choice regarding the physical characteristics like size, shape, color, and taste. Gender and ethnic groups were the possible factors that

4th Asia-Pacific Abstracts associated with public preferences of OSDF and their different physical characteristics. METHODS: A self-administered questionnaire were distributed in a cross-sectional design to 200 individuals at the main campus of Universti Sains Malaysia and 100 individuals outside the campus. A diagram of different oral solid dosage forms (capsule, tablet, caplet, and soft gelatin) was provided during the answering session. Data from 300 individuals were presented by frequencies and percentages for categorical variables and were analyzed by chi-square test. RESULTS: Capsule was the most preferred OSDF and the easiest form in swallowing. Males prefer OSDF to be in blue color while female Prefer pink. Most respondents prefer OSDF to be without taste, small in size, and round in shape. The preferred physical characteristics was size followed by taste, shape, and color respectively. A significant difference (P < 0.05) was found between gender and ethnic groups toward the preferred OSDF. CONCLUSIONS: Based on the result of this study, public have their preferences toward a particular OSDF and a particular physical characteristics. Size is the most important physical characteristics for the public to manage their choice regarding OSDF, followed by taste, shape, and color.

HEALTH CARE USE & POLICY STUDIES – Equity and Access PHP10 USING ASSET INDEX FOR MEASURING HOUSEHOLD LIVING STANDARDS AND MONITORING HEALTH EQUITY IN THAILAND Prakongsai P, Tisayatikom K International Health Policy Program (IHPP), Nonthaburi Province, Thailand OBJECTIVES: To explore potential use of the asset indexes for measuring household living standards, and monitoring equity in the distribution of government subsidy for health among different socio-economic groups of Thais. METHODS: The principal components analysis (PCA) is used to compute the weight or “factor score” of each household asset in the nationally representative household socio-economic surveys (SES) 2000, 2004, and 2006. Households categorized into different socio-economic groups between using money metric measures (income & expenditure) and the asset indexes were compared. Differences in the distribution of government subsidy for health in 2006 between using household income and the asset indexes were also explored. RESULTS: Households in higher asset quintiles frequently owned assets with a high factor score and low availability in most households, while households in lower quintiles did not. The assets with the top-three factor scores in 2000 and 2004 were washing machine, telephone, and refrigerator, while those in 2002 were washing machine, telephone, and video recorder. In all 3 years, households in the first asset quintile (the poorest) had neither air conditioner nor computer, while 48% of households in the richest quintile owned air conditioners in 2000, and increasing to 49% in 2002 and 61% in 2004. The correlations between household quintiles classified by the asset index and by household income/expenditure ranged from 0.47 to 0.54. The distribution of government subsidy on health in 2006 between using household income and asset indexes is similar. CONCLUSIONS: The asset index can be used as an alternative tool to classify socio-economic groups of Thai households and monitoring health equity. Revisit of questions on assets in the household surveys can improve the ability of the asset index to reflect household living standards. PHP12 THE PUBLIC’S PREFERENCE ON THE PRIORITIES IN HEALTH CARE 1 2 3 2 Bae EY , Lim MK , Choi SE , Lee TJ 1 Sangji University, Wonju, Kangwon province, South Korea; 2Seoul National University, Seoul, South Korea; 3College of Pharmacy, Seoul National University, Seoul, South Korea OBJECTIVES: This study investigates the public opinion on the issues of distributive justice in health care. From the literature, we can ascertain several principles outlining priority setting decisions—health maximization, fair distribution, and equity. This study aims to identify the principles the public considers important, and the trade-offs between different values in resource allocation practices. METHODS: Two Focus groups were created and discussed a variety of issues on resource allocation in health care. Participants in the groups were introduced to the objective of the project and problems at hand. To facilitate the group discussion, a simple ranking task and a series of pair-wise choice practices were implemented. In the choice scenario, the severity of diseases was represented as both the quality of life and life-years remained. All participants were expected to choose one scenario, and explain the underlying reason for their choice. Group discussions were recorded and later analyzed by researchers. RESULTS: Severity of disease was the most important criteria of priority setting for participants. The majority supports the idea that the most disadvantaged should have the highest priority, even though their health gain is less than others. Rare diseases were given high priority because of their severity, not their rareness. Through the focus group interview, we could find a strong public support for the equal opportunity principle. Next to severity of disease, socio-economic status was found to be an important consideration in resource allocation decision. CONCLUSIONS: The findings from the group discussions indicate that the severity of disease and the socioeconomic status of the patient is the most important priority setting criteria in the national health insurance setting. To further these results, a discrete choice experiment with ordinary people will be conducted and discussed.

A535

4th Asia-Pacific Abstracts PHP13 FACTORS ASSOCIATED WITH THE PROVISION OF LAPAROSCOPIC SURGERY IN THAILAND: RESULTS FROM THE NATIONWIDE INPATIENT DATABASE Vongkom W1, Tosanguan K2, Chaikledkaew U1, Teerawattananon Y2 1 Division of Social and Administrative Pharmacy, Bangkok, Thailand, 2Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: Laparoscopic surgery (LS) requires advanced and expensive surgical instruments but offers better quality of life and shorten hospitalization compared to conventional surgery. This study aimed to evaluate the utilization of LS between patients under Civil Servant Medical Benefit scheme (CSMBS) and those under Universal Coverage scheme (UC) and to determine the factors associated with the provision of LS in most common diseases. METHODS: A hospital data of patients undergoing LS were obtained from the Central Office for Healthcare Information. The database contained 686,553 admissions with principle diagnoses related to LS from January 2005 to December 2007. Descriptive analyses and binary logistic regression models were used to analyze the data. RESULTS: The total of 24,175 hospitalizations (3.52%) was operated with LS. The proportion of CSMBS patients undergoing LS (7.8%) was higher than that of UC patients (2.68%). It was found that diseases of gallbladder and cholecystitis, diseases of gynecology and acute appendicitis were the most diseases performing LS in Thailand. The provision of LS was significantly associated with age, sex, principal diagnosis, admission year, type of hospitals and type of health insurance coverage. Patients with CSMBS were about two or three times more likely to undergo LS compared to UC. Type of health insurance was the most significant factor associated with the use of LS. CONCLUSIONS: There is an unequal access to LS among patients owing to both medical and nonmedical indications. Health insurance coverage plays a significant role in LS provision in Thailand. PHP14 PHARMACEUTICAL PROCUREMENT AND SUPPLY CHAIN IN PUBLIC AND PRIVATE SECTOR IN DELHI, INDIA Kotwani A Vallabhbhai Patel Chest Institute, University of Delhi, Delhi, India OBJECTIVES: In India, medicines are dispensed free in the public sector facilities. However, due to low availability of medicines in the public facilities, 80% of healthcare expenditure is out-of-pocket. The survey was conducted to investigate the procurement system and distribution of medicines in the public sector and supply chain in private sector in National Capital Territory of Delhi (NCT, Delhi). METHODS: The majority of public health care in Delhi is provided by central government and the government of NCT Delhi; there are two additional public sector providers, Municipal Corporation of Delhi and New Delhi Municipal Corporation of Delhi. Procurement system and distribution in each of the four public sectors was studied. For private sector data was collected from reliable wholesalers and retailers by liaising with their association to know the details of mark ups in the supply chain. RESULTS: Each public sector entity has its own procurement list and procurement system. Procurement for central government hospitals and dispensaries is handled separately by a government agency that charges fees. In case of stock outs or non-availability of rate control of medicines tertiary care hospitals and central government dispensaries can do local purchase. Medicines bought as local purchases are usually more costly and rarely undergo the same quality assurance checks. In the private sector medicine distribution is characterized by a high number of generic equivalents and high levels of competition at each level of the system. Manufacturer can only sell to a licensed wholesaler; wholesaler can only sell to licensed retailers. Trade schemes run between manufacturer, wholesaler and retailer. CONCLUSIONS: Several strengths and weaknesses in the public procurement systems were observed. Huge amount of local purchases were reported. Patients do not get benefit from trade schemes; manufacturers and retailers are the beneficiary. Policy and regulatory interventions are needed to improve access and affordability of medicines.

HEALTH CARE USE & POLICY STUDIES – Formulary Development PHP15 RISK-SHARING SCHEMES WORLDWIDE: A LANDSCAPE ANALYSIS OF HEALTH OUTCOMES-BASED REIMBURSEMENT AGREEMENTS Coulton L1, Annemans L2, Javier J3, Brown R3, Keskinaslan A4 1 United BioSource Corporation—Europe, London, UK; 2Ghent University—Brussels University, Ghent, Belgium; 3United BioSource Corporation, Bethesda, MD, USA; 4Novartis Pharma AG, Basel, Switzerland OBJECTIVES: To assess and analyze the number, type, and extent of risk-sharing agreements worldwide based on published literature. METHODS: A structured literature review using predefined search criteria was conducted to identify references to, or descriptions of, health outcome-based risk-sharing agreements within peer-reviewed and trade publications between the years of 2000–2010. The identified publications were categorized by strength of evidence (i.e., systematic or non-systematic), and then aggregated by type of agreement, technology, and companies involved within the agreement. Analysis was completed to demonstrate commonalities among identified agreements as well as their unique aspects. RESULTS: Five database and publication sources were reviewed using 17 predefined search terms. The literature review suggests that many risk-sharing agreements are not published and those that are vary widely in design, scope, and intent. The search resulted in 61 abstracts which identified eight

individual published risk-sharing schemes. While all identified agreements link improvements in health outcomes with reimbursement, definitions of what constitutes improved health outcomes, as well as the type of evidence required to prove that improvement, varied dramatically. The published risk-sharing schemes were from the UK (n = 3), United States (n = 3), France (n = 1), and Sweden (n = 1). There is more publically available information on agreements outside of the United States, but it is unclear the extent to which this is due to greater transparency in reimbursement versus a reflection of more risk-sharing agreements. CONCLUSIONS: Health outcomesbased risk-sharing agreements offer the potential for both benefit and frustration to manufacturers and payers alike. The ability to review progress within this field to-date and attempt to offer trends toward best practices will be key to the long-term viability of these novel reimbursement efforts. Despite the heterogeneity of agreement types, methods, and foci, successful utilization of these agreements has been achieved and could potentially offer a guide for replication in future use.

HEALTH CARE USE & POLICY STUDIES – Health Care Costs & Management PHP16 A COMPREHENSIVE STUDY OF GENERIC DRUG ENTRY IN THE UNITED STATES: 1991–2008 Kelton CM1, Guo JJ2, Safi A1, Yu Y1 1 University of Cincinnati College of Business, Cincinnati, OH, USA; 2University of Cincinnati, Cincinnati, OH, USA OBJECTIVES: It is commonly believed that after the patent expires for a branded pharmaceutical, the average price for the generic compound falls following generic entry into the market. The objectives of this study were to 1) determine and explain the trend in drug price post-entry, and 2) predict the number of generic-company entrants, one of the most likely factors influencing price; and to accomplish 1) and 2) in a more comprehensive manner than previously in the literature. METHODS: Quarterly transaction-price data were constructed using the national summary file of Medicaid outpatient drug utilization maintained by the Centers for Medicare and Medicaid Services. Data from 1991–2008 were extracted for 65 drugs that experienced initial generic entry between 1992 and 2004. Generic relative price (GRP) was constructed as reimbursement per unit for a specific firm and quarter divided by average reimbursement per unit over the year before entry, not accounting for pharmaceutical manufacturer rebates. Least-squares regression models were estimated on the panel data to explain GRP, average GRP across firms (AGRP), and number of entrants. RESULTS: The number of firms had a statistically significant (P < 0.0001), nonlinear negative effect on GRP and AGRP. High demand, as indicated by high post-entry expenditures, had a statistically significant (P < 0.0001) positive effect on both GRP and AGRP. Statistically significant (P < 0.0001) predictors for number of entrants included pre-entry market size, number of quarters since entry, and administration form (oral, injectable, or topical) of the drug. Evidence suggests as well that rebranding following generic entry commands a price premium. CONCLUSIONS: This study generally supports the common wisdom, from prior econometric studies, that drug prices fall following generic entry. However, other factors, such as small market size or rebranding, can mitigate or even reverse the post-entry price drop, reducing the ability of payers to benefit from significant cost savings. PHP17 MEASURING MEDICATION RECONCILIATION’S IMPACT ON THE FINANCIAL COST Huang YC, Chang HT, Lin YM, Liu HP, Leu WJ, Chien HY Taipei Medical University—Shuang Ho Hospital, Jhonghe City, Taiwan OBJECTIVES: Medication reconciliation is the process to review patients’ complete medication regiments at all the time of transitions of care and compare patients’ existing and previous medication with the regiments being considered for the new setting of care. The objective of this study is to implement medication reconciliation to reduce drug-therapy problems and errors, and to evaluate the financial impact of these approaches in Taiwan. METHODS: The criteria for cases recruiting in this study were patients staying in the academic medical center for more than 3 days. Nevertheless, the patients with cancer or admitted to the ICU were excluded. Pharmacists reviewed the patients’ admission charts and compared the patients’ medication regiments at present and past. Once the inconsistence medications were identified, pharmacists would discuss with the multispecialty team consisting of physicians, pharmacists and nurses to ensure if therapy should be adjusted. The determination of financial impact is expressed by cost avoidance per year in Taiwan. RESULTS: The data from the study suggests approximately 11.07% of drug-related problems were identified and prevented by pharmacists through medication reconciliation. This can be translated into approximately 350,697 episodes per year in Taiwan (based on the assumption for total hospitalization of about 3,168,000 persons per year). The estimated cost avoidance will be as high as NT 1941 million per year (based on the assumption for cost of prolonged length of hospital stay is about NT 5000) through achieving the medication reconciliation at admission. CONCLUSIONS: This study demonstrated that the pharmacists’ interventions were able to prevent potential adverse drug events, reconcile discrepancies among medications, and reduce medical expenditure as well. Effective assurance in medication safety is also achieved through the collaboration of multiple disciplines in patient care.

A536

4th Asia-Pacific Abstracts

PHP18 LIFETIME DISTRIBUTION OF MEDICAL COST IN KOREA Jung YH, Ko S Korea Institute for Health and Social Affairs, Seoul, South Korea OBJECTIVES: It has been well recognized that health-care cost is strongly dependent on age. The study aims at estimating the magnitude and distribution of lifetime medical costs. METHODS: We employ a period life table model using single year’s medical expenditures from Korea National Health Insurance Corporation, population data and life table from Korean Statistical Agency. To estimate lifetime medical costs for male and female, 100,000 hypothetical cohorts for each sex whose mortality rate coming from life table is constructed and assigned to medical costs at each age and sex. This procedure permits us to estimate lifetime medical costs at each age and for each sex. The method used here is to conceptually convert cross-sectional costs data into a longitudinal pattern of costs, which generates profiles of medical cost from birth to death. RESULTS: The average member of the birth cohort will spend 74,150 thousands Won (98,883 PPP US$) for man and 87,868 thousands Won (117,177 PPP US$) for woman in 2007 over the course of his or her life. Total lifetime medical cost is 19 percent higher for females than males due to mainly life expectancy difference. For the average life table member, half of all lifetime cost occurs after about age 65. We also estimate the distribution of lifetime medical cost by phases of life course. For a male (female), 12.7 (9.2) percent of a cohort’s expenditures occurs from birth to age 19; 8.4 (10.0) percent accrues during ages 20~39; 30.2 (28.4) percent is expended during ages 40~64; 42.7 (44.0%) percent is realized during ages 65~84; 5.9(8.4) percent accrues over age 85. CONCLUSIONS: The findings of this study are similar to the patterns of medical costs of other countries’ experiences provided by other literatures—high during infancy, low during childhood and young adulthood, then rising during middle-age and rapidly growing during the senior years. PHP19 THE COST OF ILLNESS IN KOREA (2007) Jung YH Korea Institute for Health and Social Affairs, Seoul, South Korea OBJECTIVES: Illness is associated with significant economic burden on both individual and society. This study aims at measuring the cost of illness in Korea in 2007. METHODS: We estimate both direct and indirect costs of illness using a prevalencebased approach. Direct cost includes medical expenditures (inpatient, outpatient), caregiver’s cost and traffic costs. As indirect costs, productivity loss representing lost workdays due to illness and lost earnings due to premature death is estimated based on human capital theory. The results are reported according to age, sex, diseases and each cost items in USA Purchasing Power Parity (PPP) Dollars at 3% discount rate. The major data sources are National Health Insurance Statistical Yearbook, Annual Report on the Cause of Death Statistics, and Survey Report on Wage Structure. We also use other information such as the Korean Statistical Information System. We provide the results according to age, sex, and diseases at each cost items. RESULTS: The cost of illness of Korea in 2007 is 56,633,552 million Won (75,523 million PPP US$) based on 3% discount rate, which represents approximately 6.28% of GDP. Of this total, 47.8% (36,065 million PPP US$) is devoted to medical costs, followed by productivity loss (46.4%; 35,053 million PPP US$), caregiver’s costs (3.9%; 2,962 million PPP US$), and transportation cost (1.9%; 1,444 million PPP US$). The cost of illness is higher for males (45,912 million PPP US$) than females (29,611 million PPP US$), which results from the difference of productivity loss cost. However, direct cost of females is higher than that of males. We find that the cost of aged 40~49 accounts for the highest proportion (23.5%) followed by aged 50~59 (22.5%). The four major diseases incurring highest costs are as follows; neoplasms (18.76%), circulatory system (13.92%), digestive system (13.55%), and respiratory system (10.53%). CONCLUSIONS: This study could provide basic information for establishing prioritizations of health policies to reduce economic burden of illness and enhance quality of life. PHP20 TRENDS IN DRUG EXPENDITURES IN SONGKHLA HOSPITAL Kultavayporn S, Saowapark P Songkhla Hospital, Songkhla, Thailand OBJECTIVES: To assess situations and trends in overall drug expenditures and to understand factors influencing drug expenditures in Songkhla hospital. METHODS: It was across-sectional descriptive design research. Drug expenditures of fiscal year 2005–2007 at Songkhla hospital were studied. Overall drug purchase expenditures as well as their proportions including in the National drug list 2007 were figured out. The expenses of drug subgroups and drug items were ranked and sorted from highest to lowest expenditures, compared with the previous 3 years. The study also presented the cost impact of the use of new drugs available in the fiscal 3 years. The cost-savings of all generic drugs instead of original drugs were estimated. RESULTS: Total drug expenditure grew by 46.17% from 2005 to 2006 and grew by 43.03 % from 2006 to 2007. And analysis of 17 drug groups and all drug items in terms of expenditure indicated that 1) the expenditure of top three groups in the fiscal year 2004 and 2006 was the same group (cardiovascular, infections, endocrine system), and the expenditure of top three groups (cardiovascular, infections, central nervous system) about 116.48 million baht, in the fiscal year 2007; and 2) the expenditure of top 30 drug items (from about 612,651,677 items) was about 38.4, 69.18, 103.64 million baht, accounted for 36.52 %, 44.63%, and 41.42% of over all drug expense in the fiscal year 2005–2007. CONCLUSIONS: The study indicated that forecasting and manag-

ing of rising drug expenditure remained a challenge for hospital administrators. The dynamic reimbursement environment would further complicated drug budgeting, and policy-makers of the hospital must consider factors affecting drug expenditures as well. PHP21 WHAT ARE THE TOP MOST COSTLY DISEASES FOR USA? THE ALIGNMENT OF BURDEN OF ILLNESS WITH PREVENTION AND SCREENING EXPENDITURES Kockaya G1, Wertheimer A2 1 Istanbul University Istanbul Medical Faculty, Istanbul, Turkey; 2Temple University School of Pharmacy, Philadelphia, PA, USA OBJECTIVES: It was the intention of the authors to generate a list of the top diseases responsible for the greatest financial expense in the United States. This listing would then inform policymakers as to the highest priority target conditions. With such information available, funding for the NIH could be accomplished based upon factual criteria rather than political clout or uninformed consensus. METHODS: The first step was a literature review to explore articles and reports which were published about the cost of illness (COI) up to December 2009. Therefore the source of the data used in this investigation was obtained from this retrospective search approach. RESULTS: With reference to total cost for disease, first was HCVD with US$475.3 billion, followed by alcohol abuse and substance at US$300.6 billion, digestive diseases at US$259.6 billion, cancer at US$239.5 billion and mental disease at US$216.6 billion. CONCLUSIONS: As has been seen, the total societal costs of the diseases do not synch with the degree of attention paid to these various disease states. Several of the diseases have very powerful and vocal support organizations that encourage attention and clinical research support. On the other hand we hear very little about efforts against allergic rhinitis or infectious and parasitic diseases. PHP22 UNIT COST OF BURMESE TRANSLATOR ACTIVITIES IN PHARMACY DEPARTMENT, SAMUTSAKHON HOSPITAL Angkoonsit D Samutsakhon Hospital, Mahachai, Thailand OBJECTIVES: Burmese laborers are major Non-Thai patients in Samutsakhon Hospital. Culture and language barriers between Samutsakhon Pharmacist and Burmese patients may lead to many drug related problems that effect to patients safety. For this reasons, Burmese transalators were worked like pharmacists assistants since May 2008. This activity may improve quality of service as increased cost. This research is aim to determine labor cost of Burmese translators at pharmacy department, Samutsakhon Hospital. METHODS: This research is descriptive study in health-care provider prospective to determine labor cost of Burmese translators between October 1, 2009 to December 31, 2009. Unit cost of Burmese translator was calculated from labor cost multiply with estimated time in each translator activities by observation. RESULTS: The results of this study found that labor cost of Burmese translator is 0.98 bath/min, and unit cost of each activity included translating drug information for general patients costs 2.45 bath/case,while, dispensing vitamins for Burmese pregnancies which costs about 0.98 bath/case. Although giving drug information for TB and AIDs patients is 4.9 bath/case or translating the evaluating adverse drug reaction is 9.8 bath/case. The most expensive cost is teaching to use special device for asthmatic patients about 14.7 bath/case. There are other assignments, for example, preparing Burmese drug leaflets or prepack medicines. Every activities are under responsibed by Thai pharmacists. CONCLUSIONS: From this study is quite similar to labor cost of Nurse assistants at Ledsin Hospital, Bangkok in 2005 (0.81 bath/min). Also, this information can be applied for other hospitals that in the similar situation. This study can help decision-maker to spend more budget to improve quality of pharmaceutical care with humanity. PHP23 HEALTH DECENTRALIZATION, GOVERNMENT HEALTH EXPENDITURE AND HEALTH OUTCOME IN CHINA Chen Q, Li L Peking University, Beijing, China OBJECTIVES: Since the early 1980s, decentralization has reemerged as a valued political and economic goal in most developing countries. China’s health-care system is also in the process of decentralization, local governments take more responsibilities of health sector, accompanied by the diminution of public role in health financing especially after the fiscal reform in 1994. This research looks at the health outcome (both the level and gradient) during the period 1991–2007 to identify the effects of these policy changes on health outcomes in China. METHODS: we borrow the micro Grossman health production model using province-level pooled cross-section and time-series data during the period 1991–2007. We first investigates the “overall” and “partial” effect of income on health, and then focus on the policy effects. Life expectancy, infant mortality, mortality, and maternal mortality are examined as health variables (level, gradient, and convergence). Medical care, public health, socioeconomic, environmental factors as well as institutional variables are considered in the production function. RESULTS: The results are coincident with theoretical expectation i.e., economic growth alone does not promote health in typical developing countries, but some channels like its impact on public health do, and also the system does matter. The decentralization of government responsibilities and reduction of public role are significant for the fall of mortality and infant mortality and both significantly affect the sigma and beta convergence of health

4th Asia-Pacific Abstracts variables. CONCLUSIONS: Economic development is important support for health development, but economic growth will not naturally improve health especially for the government investment on health. Rational incentives for the government is essential for health development. PHP26 VALIDATION OF ELECTRONIC DATABASE IN COMMUNITY HOSPITALS IN PATIENTS WITH ATRIAL FIBRILLATION Chotchaisuwatana S1, Jedsadayanmata A2, Chaiyakunapruk N2, Jampachaisri K1 1 Naresuan University, Muang, Phitsanulok, Thailand; 2Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: The electronic databases (ED) has been increasing used for clinical research since it reflects a routine care with large and heterogeneous samples. However, few studies report on its validity for such purposes, especially the ED from community hospitals. This study aims to assess the validity of EDs based on data from patients with atrial fibrillation (AF) receiving care from community hospitals in Phitsanulok. METHODS: The validity of ED was determined using out-patient medical records (OPMRs) as a gold standard. A total of 200 AF patients were randomly sampled from a pool of patients with ICD-10 of AF (I48) from two community hospitals during August 2007–July 2008. For each patient, data of a randomly selected visit from the ED was matched to data of the same visit from OPMRs, abstracted by a standardized data collection form. The ED was cross-validated with OPMRs based on patient’s comorbidities (risk factors for stroke) and bleeding events. All data were tabulated in a 2 × 2 format to calculate sensitivity, specificity and the Cohen’s Kappa adjusted for chance agreement. RESULTS: Out of 200 AF patients, 176 were documented as having diagnosis of AF in OPMRs (88%). The ED data on risk factors of stroke showed moderate to high sensitivity (range 66.67–100%) and high specificity (range 99.36–100%). These results suggest that the risk factors of stroke were coded accurately. Based on these data, the agreement between two databases was considered good to very good (calculated kappa range 0.7940–0.9680). The specificity based on major bleeding was 100%; however, sensitivity and the Cohen’s Kappa could not be determined because the major bleeding diagnosis was found neither in the EMRs nor the OPMRs. CONCLUSIONS: The electronic database of AF patients from community hospitals was valid and in good agreement with the OPMRs. PHP27 USING COLLABORATIVE PROGRAM BETWEEN PHARMACISTS AND NURSES YO IMPROVE SPONTANEOUS ADVERSE DRUG REACTION REPORTING SYSTEM IN THAILAND Prakongsai N1, Pongchaidecha M2 1 Prapokklao Regional Hospital, Ministry of Public Health, Chantaburi province, Thailand; 2 Silpakorn University, Nakorn Pathom Province, Thailand OBJECTIVES: To evaluate the collaborative program between pharmacists and nurses in improving the spontaneous reporting system (SRS) of ADR in Prapokklao hospital, and to explore the impact of the collaborative program on nurses’ knowledge about ADRs, capacity in detecting and reporting ADRs, and the level of satisfaction. METHODS: Several methods including experimental approach which categorized nurses into experimental and control groups, then compared knowledge before and after a training program and performance in detecting and reporting ADRs between both groups. Eight out of 52 nurses in the experimental group were intensively trained with knowledge about ADRs, and how to report ADRs in hospitalized patients. The rest (44 nurses) in the experimental group obtained knowledge from researchers and the training nurses. Nurses in both experimental and control groups were assessed about knowledge before and after the start of the program, performance in monitoring and reporting ADRs compared to pharmacists, and satisfaction using a self-administered questionnaire. RESULTS: Eight nurses trained by the program and 44 nurses in the experimental group had a significant higher score of knowledge than before training. Differences in the pre and post-test scores between experimental and control groups were statistically significant at 0.05. Intensive ADR monitoring indicated that experimental wards had 3572 patients with 371 ADR problems, while control wards had 3467 patients with 266 problems. Nurses in the experimental groups could correctly report 174 ADR problems (46.9%), while 29 problems (10.9%) were reported by their counterparts. The difference in the capability to reporting ADR was at the significance level of 0.05. The level of probability and the difference in the proportion of ADR reported by the experimental group was fourfold to the control group. CONCLUSIONS: This collaborative program between pharmacists and nurses significantly improved performance and knowledge of nurse in SRS reporting system in Prapokklao hospital. PHP28 IMPACT OF SAFETY ISSUES’ BASED DECISIONS OF CONSULTATIVE COUNCIL ON THE POLISH NATIONAL HEALTH FUND’S BUDGET Farkowski MM, Baran J, Matusewicz W Agency for Health Technology Assessment in Poland, Warsaw, Poland OBJECTIVES: Nowadays growing number of new health technologies entering market together with more aggressive therapies administered by physicians lead to increase in number and cost of care of related adverse effects (AE). We were interested in the impact of costs of AE related to medicines not recommended for public funding by the Consultative Council (CC) of Polish Heath Technology Assessment Agency (AHTAPol) on the ground of safety issues on the Polish National Health Fund’s budget. METHODS: Among decisions of CC published until the end of 2009, we distinguished those where safety issues were significant arguments for decline, and

A537 analyzed submissioned budget impact analyses (BIA) of those medicines in order to find the incremental costs of AE related to the treatment. Then, we compared those costs to the total incremental costs of public funding reported by manufacturers. RESULTS: Among 148 CC’s decisions analyzed, 70 were negative and in 22 safety issues were significant arguments against the positive recommendation (31% of all negative decisions). Out of 22 analyzed BIA’s only in two cases (10%) incremental costs of AE were explicitly stated, both medicines being used in oncology. In one case (4.5%), a reason for absence of those costs was stated. Stated incremental cost of AE of those two medicines was about 0.25 to 0.35 million USD in years 1 to 3 of public funding, which corresponds to about 2.5% of total incremental cost of those medicines and only 0.2% of total incremental costs of all 22 submissions (about 110 to 135 million USD). CONCLUSIONS: In submissions rejected by CC where one of the main concerns were safety issues, manufacturers failed to explicitly show the impact of costs of adverse effects of their drugs on Polish National Health Fund’s budget. PHP29 ASSESSMENT OF THE ORPHAN DRUG PRICING AND REIMBURSEMENT LANDSCAPE IN THE ASIA PACIFIC REGION Mukku SR1, Pang F2 1 Double Helix Consulting, London, UK; 2Shire Human Genetic Therapies, Basingstoke, UK OBJECTIVES: In recent years, there has been a steady increase in products for chronic rare diseases. The aim of this research is to assess the current pricing and reimbursement framework for orphan and ultra-orphan drugs - identifying emerging trends, defining the key components of health technology assessments, observing government and payer policies, and reviewing previous access decisions in relation to the AsiaPacific region. METHODS: Interviews were conducted with a variety of stakeholders (payers, academics, KOLs) to identify the key determinants influencing the pricing and reimbursement and access strategies of orphan drugs in AP countries including Australia, China, Japan, India, Singapore, South Korea. These were supported by in-depth secondary analyses from government websites, the WHO, reports and OECD statistics. RESULTS: Decisions reviewed for a selection of drugs (e.g., Naglazyme, Glivec, Tracleer) across a number of therapeutic areas in the context of indicators including population size, price level, %GDP on health care (Singapore 3.4%, India 4.9%, China 4.5%, Australia 9.1%), number of specialists etc, inferred that there is significant variability in uptake. Orphan definitions vary by country from 4/10,000 in Japan to 1.2/10,000 in Australia. Pricing, payment and regulation mechanisms also differ across countries including financial incentives, protocol assistance, fast-track procedures and market exclusivity (Years: US = 7, EU = 10, Japan = 10, Australia = none). Nearly 100 orphan drugs have been approved over the past 12 years in Japan. In predominantly self-pay markets such as China and India, many orphan diseases remain untreated. Policy determinants include economic troubles impacting healthcare budgets, stakeholder attempts to rationalize investment, population demographic shifts, affordability, desire for consistency across and within markets and the requirement to contain rising health-care costs while respecting the philosophy of the healthcare system (egalitarian vs.ulitarian). CONCLUSIONS: Health system and payer attitudes toward health provision are evolving over time, impacting the type of approach taken to address orphan and ultra-orphan product access in the Asia-Pacific region. PHP30 MARKET ACCESS, AFFORDABILITY AND PRICE COMPONENTS OF MEDICINES IN THE UNITED ARAB EMIRATES (UAE) Ansary A1, Abuelkhair M2, Mukku S3 1 University of Cambridge, Cambridge, UK; 2Health Authority-Abu Dhabi, Abu Dhabi, United Arab Emirates; 3Double Helix Consulting, London, UK OBJECTIVES: The research was aimed at understanding the complexities of pricing and reimbursement practices in UAE with analysis focussing on price mark-ups, affordability, stakeholder analysis and strategic recommendations for pharmaceutical companies to launch products in this market. METHODS: Interviews were conducted with Payers and KOLs to understand patient flow, health-care issues and drug pricing in UAE. In addition in-depth secondary analysis was conducted. RESULTS: UAE with a population of 4.2 million has seen more than 700% growth from 1975–2006, contributed mostly by expatriate work force. This has significantly impacted on the management of health care in recent years. Medicines priced in the UAE are generally high because of the lack of pharmaceuticals manufacturing base and subsequent reliance on imports. While individuals spend $52 in the gulf countries and $20 in other Arab countries on medicines, an individual in the UAE spends about $80 a year. There is a markup of up to 20% and 24% by the wholesalers and retail pharmacies, respectively; which is relatively high in comparison to many other countries resulting in expensive drugs that become unaffordable. A lack of competition legislation and anticompetitive practices means that pharmacies in the UAE are required to buy medicines solely through few agents who hold exclusive rights to importing and distributing products. Also there are differences among the UAE regions; e.g., HAAD is introducing additional controls through a differential co-payments and reference pricing when insufficient competition is observed. As part of market access policy development, the UAE MoH has started applying basic pharmaco-economic methodologies that considers the active ingredient’s strength, dosage, and therapeutic effect. CONCLUSIONS: The pricing and reimbursement policy in its current form is not-sustainable. New pricing models need to be considered in order to ensure a healthy population with good access to affordable medicines and health care in the UAE.

A538 PHP31 WILLINGNESS TO PAY PER QUALITY ADJUSTED LIFE-YEAR: IS ONE THRESHOLD APPLICABLE FOR ALL DECISION-MAKING? Zhao FL1, Yue M2, Yang H3, Wang T3, Wu JH2, Li SC1 1 University of Newcastle, Callaghan, NSW, Australia; 2306 Hospital of PLA, Beijing, China; 3 The First People’s Hospital of Yunnan Province, Kunming, Yunnan, China OBJECTIVES: To estimate the Willingness to pay (WTP) per quality-adjusted life-year (QALY) ratio with the stated preference data and compare the results obtained between chronic prostatits (CP) patients and general population. METHODS: CP patients were recruited from two tertiary referral hospitals and the general populations were randomly approached in China at the beginning of 2009. WTP per QALY was calculated with a formula combining the subjects’ own health-related utility and the WTP value. Two widely used preference-based health-related quality of life instruments, EQ-5D and SF-6D, were used to elicit utility for participants’ own health. The monthly WTP values for moving from participants’ current health to a perfect health described by “11111” status of EQ-5D were elicited using closed-ended iterative bidding contingent valuation method. RESULTS: A total of 268 CP patients and 364 participants from general population completed the questionnaire. We obtained four WTP/QALY ratios ranging from $4700 to $7400, which were lower than the proposed thresholds and published researches eliciting the preference for avoiding the risk of death. In addition, the WTP/QALY ratios from the general population were significantly lower than those from the CP patients and different determinants were associated with the within group variation identified by multiple linear regression. CONCLUSIONS: Preference elicitation methods are acceptable and feasible in the socio-cultural context of an Asian environment and the calculation of WTP/QALY produced meaningful answers. The lower WTP/QALY elicited than published values and higher value from CP patients compared with the general population highlight the necessity of considering disease specific QALY in estimating WTP/QALY. Our results inferred that one threshold might not be enough to serve all decision-making under different situations. Further studies using the same methods to confirm whether the WTP/QALY value would be dissimilar among diseases with different impact on QoL would be needed. PHP32 PROBABILISTIC SENSITIVITY ANALYSIS—A NECESSARY EXTRA? 1 2 1 Kim H , Gurrin L , Liew D 1 The University of Melbourne, Fitzroy, Victoria, Australia, 2University of Melbourne, Carlton, Victoria, Australia OBJECTIVES: Probabilistic sensitivity analysis (PSA) is a useful tool to assess parameter uncertainty, but being among the more technically advanced methods in cost effectiveness modeling, it is sometimes underutilized. However, following the incorporation of PSA by NICE, the English HTA agency, into their guidelines in 2005, there was a call for the routine use of PSA in economic modeling. This study investigates whether these two developments have had an effect on cost-effectiveness modeling practice, and also reviews current requirements for PSA in reimbursement guidelines globally. METHODS: The following three journals in which cost effectiveness analyses are most often published were included in the study: Medical Decision Making, Pharmacoeconomics and Value in Health. All papers published in these three journals in 2004 and 2009 respectively were assessed. In addition, pharmacoeconomic guidelines from 31 countries were compared for the requirement of PSA in reimbursement submissions. RESULTS: In the three journals from 2004 to 2009, the overall number of articles presenting cost-effectiveness modeling increased from 41 (2004) to 55 (2009). In 2009, 69% of these articles presented PSA, compared to only 32% in 2004. Of a total of 31 national pharmacoeconomic guidelines, 12 mention PSA. However, only six of these require this form of analysis to be included in reimbursement submissions. Many countries with well established requirements for economic analysis in reimbursement submissions, such as Australia, do not require PSA. CONCLUSIONS: The usage of PSA to investigating parameter uncertainty is now common and increasing. The trend is also starting to show in the reimbursement agencies’ guidelines. PHP33 CHINESE PAYERS’ VIEW OF PHARMACEUTICAL VALUE ATTRIBUTES: WHICH EVIDENCE DRIVES ACCESS? Shankland BDT Double Helix Consulting, London, UK OBJECTIVES: The evidence needs of payers in mainland China at the national, regional and local hospital level are poorly understood. As such, analysis is required to capture payer preferences regarding evidence submissions from pharmaceutical companies and identify emerging trends in value evidence needs. METHODS: A qualitative telephone survey of 30 key payers in Tier 3A hospitals, provincial funding bodies and national decision-making groups was conducted. Respondents were asked to score the impact of the following value attributes in their routine assessment of health-care products for formulary listing and/or reimbursement: 1) Innovation; 2) Health-related quality of life; 3) Unmet need; 4) Cost effectiveness; 5) Price; 6) Budget impact; 7) Safety; 8) Efficacy in head-to-head trials; and 9) Efficacy in selected populations. Scores derived were weighted according to a market impact assessment in order to derive maximal scores for each evidence type. RESULTS: Hospital payers reported that budget impact was the least significant evidence type (Score 5.6), due to the “selfpay” model of patient access in China. Among hospital and provincial payers, price was the second least significant attribute (Score 6.3), due to the separation of pricing

4th Asia-Pacific Abstracts functions and assessment of clinical evidence in China. However, efficacy in head-tohead studies (Score 7.5) and safety (Score 8.3) were the most desired attributes. CONCLUSIONS: Awareness of cost-effectiveness is at an early stage in major hospitals in China; the key attributes continue to be safety and efficacy. However, there is a trend toward requirement of health-related quality of life data at the national payer level, particularly regarding oncology medicines. Further research is needed in this area, in order to gain a more detailed understanding from the pharmaceutical company perspective.

HEALTH CARE USE & POLICY STUDIES – Health Care Research & Education PHP34 THE EVALUATION OF HEALTH-CARE SYSTEMS OF CHINA, HONG KONG, VIETNAM, THAILAND, MALAYSIA, SINGAPORE AND AUSTRALIA Chan HYH, Steadman KJ, Hollingworth S, Nissen LM University of Queensland, Brisbane, Australia OBJECTIVES: To examine the health-care systems of China, Hong Kong, Vietnam, Thailand, Malaysia, Singapore and Australia using a pre-determined set of indicators. These were used to assess each system in terms of accessibility, effectiveness, efficiency, quality and responsiveness. METHODS: This study was a qualitative study. A set of indicators and sub-indicators was developed based on indicators used by the World Health Organisation, United Nations, the Organisation for Economic Co-operation Development (Health Care Quality Indicator Project), and from health frameworks of Australia, New Zealand, Canada, USA and UK. The indicators were used to assess health-care systems from an operational perspective. A literature search and interviews with relevant academics and government officials were conducted to address each indicator. This provided a more comprehensive view of the functioning of each healthcare system and how the system itself is regulated and provides health-care services to its people. RESULTS: Each of the surveyed countries have certain measures or programs in place to address health-care accessibility. The majority of the countries have clear guidelines to improve effectiveness and efficiency. However, most developing Asian countries lack clear programs to assess the quality and responsiveness of their health-care systems. CONCLUSIONS: Different countries have different health considerations and priorities in terms of politics, finance and resources where healthcare systems are concerned. Each has its respective strengths and weaknesses. What is appropriate for one country may not be suitable for another. This evaluation provided clarity and insight into the operation of each system and highlighted areas that require further attention. Interviews with local academics, government officials and other health-care stakeholders in each country yielded a more comprehensive and in-depth understanding on the functioning of each health-care system. PHP35 HEALTH TECHNOLOGY ASSESSMENTS: INDICATORS OF DEVELOPMENT IN ASIA Chowdhury CA, Martin de Bustamante M Insight Strategy Advisors, New York City, NY, USA OBJECTIVES: To understand the historical climate through which ten Asian markets arrived at the establishment of an HTA authoritative body, and the extent to which pharmacoeconomic evaluations have evolved in each of these markets. METHODS: Secondary and primary research was conducted in South Korea, Thailand, Philippines, Taiwan, Singapore, Malaysia, Indonesia, India, China and Japan to examine political and economic changes over the last 30 years. The results were compiled and analyzed along with other metrics, such as pricing legislation, health-care coverage, and HTA development to determine a correlative pattern for Asian pharmacoeconomic development. RESULTS: While three of the studied markets are exceptions to the trend, we believe the other seven will follow the a discernible course of events. A change in political climate, desire for universal health care, financial crisis and strong development of a national formulary are all policy objectives and events along the path toward HTA development. HTA establishment in those seven countries is at various phases of development, but for most of the markets creation of an HTA unit and utilization of it is the next step in their pharmaceutical market evolution. CONCLUSIONS: Developing pharmaceutical markets have become reliant on pharmacoeconomic data to determine the true value of innovative pharmaceutical products. The establishment of an HTA body in these Asian countries will continue to develop and will become an important aspect of their pharmaceutical market expansion, therefore the subject matter warrants further research and attention. PHP36 ANALYSIS OF THE TECHNICAL EFFICIENCY IN THREE TEACHING HOSPITALS IN MALAYSIA Moshiri H, Aljunid S UNU-IIGH, Kuala Lumpur, Malaysia OBJECTIVES: The measurement of efficiency is usually the first step in auditing performance of hospitals. Measuring hospital efficiency provides useful information for hospital managers. It constitutes the rational framework for the distribution of human and other resources between and within hospitals. This study focuses on measuring and evaluating technical efficiency in teaching hospitals at department level

A539

4th Asia-Pacific Abstracts in Malaysia. METHODS: A non-parametric method called Data Envelopment Analysis (DEA) with two assumptions : Variable Return to Scale (VRS) and Constant Return to Scale (CRS), was used to calculate and compare the efficiency scores for selected hospitals’ clinical departments between the year 1998 and 2006. DEA input oriented analysis indicates how the inefficient units could adjust their inputs to reach the efficiency frontier. RESULTS: Based on CRS model the mean efficiency scores in Hospital A departments were 76%. One department was around 50% and six departments were between 50% and 90% and three departments were more than 90%. The mean efficiency scores in Hospital B was 92%. In this hospital, two departments were between 75% and 90% and two departments had efficiency score equal 100% during study period. In Hospital C all departments were more than 75% and one department had efficiency score equals 100%. The results based on VRS model showed similar trends. CONCLUSIONS: The mean of efficiency score according different assumptions of Hospital B was higher than two other hospitals. The results showed that few departments are efficient and rests are considered inefficient and need to find optimum mixture of inputs combination. It is suggested that Hospital A and Hospital C should consider to improve their management of the resource inputs in inefficient department in order to enhance their efficiencies.

HEALTH CARE USE & POLICY STUDIES – Health Technology Assessment Programs PHP37 PRIORITIZATION OF HEALTH POLICY AND SYSTEM RESEARCH TOPICS IN THAILAND: MAKING IT SYSTEMATIC, TRANSPARENT AND PARTICIPATORY Pilasant S, Yothasamut J, Tosagnuan K, Werayingyong P, Tantivess S, Teerawattananon Y Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: A growing concern is that health policy and system researches (HPSR) should address the needs of potential users, and subsequently have substantial impacts on policy decisions and professional practices. The purpose of this study is to describe the experience on the annual HPSR topic prioritization conducted by Thailand’s Health Systems Research Institute and its alliances. METHODS: Narrative descriptive and quantitative approaches were employed to illustrate the processes of and results from the HPSR topic prioritization in 2010. RESULTS: The prioritization process was carried out on the basis of systematicness and transparency, with participation by key stakeholders including policymakers, academics, health professionals, civil society, and industries. There was a call for research topic proposals from stakeholders in November and December 2009. A total of 120 topics suggested by 66 organizations were then prioritized by 90 representatives of stakeholder organizations. Multiple criteria introduced in this step involved policy relevance; disease burden; economic impact; social and ethical aspects; variation in practices; possibility of changing practices; and public concerns. It was found that topics related to diseases with high burden, relating to service delivery especially on health promotion and disease prevention, and those submitted by central government agencies were more likely to get high priority than others. In addition, results from self-administrative survey demonstrated that over 92% of stakeholders strongly supported and expressed their interest to participate in the next annual topic prioritization process. CONCLUSIONS: This case study demonstrated that it is feasible to develop clear criteria and transparent process for prioritization of HPSR topics. Lessons learned from this case study can be useful for improving mechanism for selecting HPSR topics in other settings. PHP38 COMPLEXITY INCREASES UNCERTAINTY: THE IMPACT OF PBAC GUIDELINES (VERSION 4) ON PBAC DECISION-MAKING Chollet M, Lindsay P, Gonzalo F Sanofi-Aventis Australia, Macquarie Park, Australia OBJECTIVES: In Australia, the Pharmaceutical Benefits Advisory Committee (PBAC) makes recommendations to the Minister for Health on the reimbursement of pharmaceuticals. The sponsor’s submission is accepted by the PBAC if the drug is determined to be clinically effective and also cost-effective. New PBAC Guidelines on how to prepare a submission (version 4) were introduced in 2008. These new Guidelines sought to reduce the uncertainty for the PBAC in accepting the many inferences made in major submissions. We assessed whether the New Guidelines have indeed reduced the PBAC’s uncertainty in their decision-making. METHODS: Since June 2003 all PBAC recommendations have been made public on the Department of Health & Ageing website. Public Summary Documents (PSD) are available for PBAC considerations relating to the PBS listing of medicines since July 2005 meeting. We reviewed all the PSDs reported during the period of July 2005 to July 2009. For each PSD, we estimated the average number of times that the words “uncertain/uncertainties/uncertainty” appear per PSD page. We compared the results for the period before and after the introduction of version 4 of the PBAC Guidelines. RESULTS: The average number of times that the words “uncertain/uncertainties/uncertainty” appeared per PSD page was significantly higher for the period after the introduction of version 4 of the PBAC Guidelines compared to the period before (0.51 vs. 0.66, P < 0.00001). CONCLUSIONS: The introduction of version 4 of the PBAC Guidelines in 2008 has led to an increase in the complexity and, thus, uncertainty faced by PBAC during their deliberations around reimbursement of pharmaceuticals in Australia. There was a significant 30% increase in the number of times that the word “uncertain/uncertainties/uncer-

tainty” was found per PSD page compared with the period prior to the introduction of the version 4 of the Guidelines (2003–2008).

HEALTH CARE USE & POLICY STUDIES – Prescribing Behavior & Treatment Guidelines PHP39 THE EFFECTS OF DIRECT BILLING SYSTEM IN PATIENTS WITH CIVIL-SERVANT MEDICAL BENEFIT SCHEMES ON PRESCRIBING PATTERNS Dilokthornsakul P, Chaiyakunapruk N, Nimpitakpong P Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: In 2006, the reimbursement system for civil-servant medical benefit schemes beneficiaries in Thailand was changed to direct billing system. It was unknown how this new system affects drug expenditures and the number of drug supply given to beneficiaries. This study aims to assess the effects of direct billing system on prescribing patterns. METHODS: This study was undertaken with retrospective cohort approach. We used the data recorded in databases of a university hospital in northern part of Thailand. We undertook the data on all patients with civil-servant medical benefit schemes who came to out-patients department between October 1, 2005 and March 31, 2007. Mean cost of medication, number of days’ supplies and medication possession ratio (MPR) of five highest costs of oral medication were calculated in 1 year before and after the system was changed. RESULTS: Out of 43,897 visits made at the hospital, 15,632 (36%) were under civil-servant medical benefit schemes. Eighty-eight percent (13,785/15,632) received medications during visit. The total costs of medication increased from 2 million to 4 million baths in 1 year. Glucosamine, atorvastatin, rosiglitazone, clopidogrel and diacerin were highest used in terms of drug cost. Averages of day’s supplies based on these medications increased from 1.29 to 1.48 months per a prescription. Proportion of patients receiving medications more than 3 months, was slightly increased from 1.82% to 2.43%. Three out of five medications had higher MPR after system was changed (Relative risk ranged on 1.19–2.32). Two of these were statistical significant. CONCLUSIONS: The direct billing system affects prescribing patterns as indicated by trend of increased number of day’s supplies and higher medication possession ratio. Further evidence remains needed. Policymakers need to consider all relevant and important consequences associated with the new system prior to making policy decision-making.

INDIVIDUAL’S HEALTH – Clinical Outcomes Studies PIH1 CRITICAL APPRAISAL OF SYSTEMATIC REVIEWS ASSESSING SAFETY OUTCOMES OF SSRIS IN THE PERINATAL PERIOD Merlo GB, Elliot L, Campbell S, Norris S Health Technology Analysts Pty Ltd, Sydney, NSW, Australia OBJECTIVES: A systematic appraisal was conducted of published systematic reviews that assessed the harms associated with selective serotonergic reuptake inhibitors (SSRIs) in the perinatal period, both for the mother and infant. METHODS: A systematic method of literature searching and selection was employed for this review. Searches were conducted in EMBASE, Medline and the Cochrane Database of Systematic Reviews. Studies were eligible if they evaluated pregnancy or infant-related safety outcomes for SSRI use in pregnant or lactating women. RESULTS: The literature search identified seventeen systematic reviews and three subsequently published prospective cohort studies. None of the systematic reviews assessing serotonergic antidepressants as a group found an association with congenital malformations. An association between paroxetine exposure and infant cardiovascular malformations has been reported in the literature; however, more recent evidence from a large systematic review shows no relationship between paroxetine exposure and congenital cardiac malformations. Neonatal symptoms (such as withdrawal symptoms, lower Apgar score, and diminished response to pain stimulus) have been reported in 20–30% of infants with third trimester SSRI exposure. All of the reviews reported the symptoms as mild and self-limiting. Several SRs found a significant association between SSRI use in pregnancy and premature delivery, low birthweight, and admission to special care nurseries. There is conflicting evidence regarding the long-term neurodevelopmental risks of serotonergic antidepressants. Although the levels of SSRIs in breast milk are relatively low, the evidence for the safety of antidepressant exposure via breastfeeding is limited. CONCLUSIONS: SSRI exposure during pregnancy is associated with mostly minor and temporary adverse outcomes for the newborn. The risk of these outcomes needs to be balanced with the risk of adverse outcomes resulting from SSRI withdrawal for the mother. PIH2 IMPACT OF HPV VACCINATION ON CERVICAL CANCER IN ASIA: RESULTS OF A STATIC MODEL Demarteau N1, Van Kriekinge G2, Castellsagué X3 1 GlaxoSmithKline Biologicals, Wavre, Belgium; 2GSK Bio, Wavre, Belgium; 3Institut Català d’Oncologia, L’Hospitalet de Llobregat, Barcelona, Spain OBJECTIVES: Estimate the potential clinical effect of HPV vaccination with a bivalent HPV-16/18 vaccine in Asian countries including the effect of cross-protection against non-vaccine oncogenic HPV types. METHODS: A static population model estimates

A540 the expected annual number of cervical cancer cases (CC) and deaths prevented by HPV vaccination of 12-year-old girls in Asian countries (WHO classes) at steady state. Input data are, for each country, the incident CC cases and deaths (GLOBOCAN 2002), the distribution of HPV types in CC (http://www.who.int/hpvcentre) and the clinical trial vaccine efficacy against CIN2+ related to HPV-16/18 and HPV-31,45,33,52,58,35, 39,51,56,59 combined (cross-protection). Lifetime vaccine protection is assumed. The effect of vaccination coverage and the use of alternative estimates for countries without sufficient data are explored. RESULTS: Of 47 countries, sufficient country-specific input data are available for eight (China, Japan, Iran, Indonesia, Republic of Korea, Thailand and Philippines). The model predicts that, with 100% vaccine coverage, the HPV-16/18 CC and death reduction ranges from 57% (Japan: −4421 cases; −2032 deaths) to 74% (Thailand: −4627 cases; −1942 deaths) while cross-protection related CC and death reduction ranges from 9% (Iran; −102 cases, −53 deaths) to 27% (Japan; −2073 cases; −953 deaths). The overall reduction ranged from 78% (Iran: −873 cases, −454 deaths) to 89% (Indonesia: −13,375 cases; −6724 deaths; Thailand −5534 cases; −2323 deaths). With 70% vaccine coverage the CC cases prevented ranges from 611 to 80,052 and CC deaths prevented from 318 to 44,921 for Iran and India respectively. Eastern, southeastern, southern Asia and Asia continent distributions are available and can be used as a proxy for countries without sufficient HPV data. CONCLUSIONS: Modeling predicts HPV vaccination with a bivalent HPV-16/18 vaccine could result in substantial reductions in CC cases and deaths in Asian countries. Cross-protection could play an important role in this reduction.

INDIVIDUAL’S HEALTH – Cost Studies PIH3 A COSTING STUDY COMPARING MIDWIFERY GROUP PRACTICE WITH USUAL CARE IN AN AUSTRALIAN METROPOLITAN HOSPITAL Turkstra E, Toohill J, Gamble J, Scuffham PA Griffith University, Meadowbrook, Queensland, Australia OBJECTIVES: To undertake an economic evaluation of the costs and outcomes of midwifery group practice care compared with usual care in an Australian metropolitan hospital. METHODS: In a cohort study, pregnant women at low risk of complications could select to receive midwifery group practice care or usual care early in their pregnancy and were recruited when they were 35 weeks pregnant. Midwifery group practice provides care by the same two to three midwives and labor in a birth center. Usual care consists of women generally attending a GP or midwives antenatal clinic for antenatal care, followed by labor in the hospital. Costing data was collected from week 36 of pregnancy until 6 weeks postpartum. Costing for antenatal, labor, baby, and postnatal care were collected using the hospital accounting system. Women kept a diary with the number of antenatal and postnatal visits. Costing data on GP visits were calculated using the diaries and government reimbursement costs. RESULTS: The study included 102 women, with 52 women receiving midwifery group practice care and 50 women receiving usual care. Midwifery group practice care was associated with fewer antenatal visits, lower rate of induction and pharmacological pain relief, shorter stay in hospital and more postnatal visits. There were no statistical differences in clinical outcomes of the baby. The cost of antenatal care was similar between the groups; labor and baby costs were lower for midwifery group practice, while postnatal costs were higher in the midwifery group practice. Midwifery group practice was associated with a lower total cost per woman compared to usual care (A$4447 vs. A$5772, P = 0.047). CONCLUSIONS: For women at low risk of complications midwifery group practice is a cost-effective option, with better clinical outcomes and lower total costs. PIH4 COST OF PUBLIC HEALTH DELIVERY OF CHILDHOOD IMMUNIZATIONS IN NOVA SCOTIA Michaels CL1, Kumaranayke L2, Scott J3, Sarwal S4, Coombs A5, Holmes E6, Sketris IS4 1 University of London, London, England, UK; 2Nova Scotia Department of Health, Halifax, NS, Canada; 3IWK Health Centre, Halifax, NS, Canada; 4Dalhousie University, Halifax, NS, Canada; 5Immunization and Respiratory Infectious Diseases, Centre for Infectious Disease Prevention and Control, Halifax, NS, Canada; 6Nova Scotia Health Promotion and Protection, Halifax, NS, Canada OBJECTIVES: Childhood immunizations are recognized as one of the most costeffective health interventions. Yet little is know about the actual delivery costs, or variation in costs by service providers and geographic regions. This study undertook an economic analysis of public health delivery of childhood immunizations in Nova Scotia. METHODS: The analysis was performed from the perspective of the government health-care provider for a 1-year period from April 2005 to March 2006. An incremental approach was used to assess the cost of delivering childhood immunizations in addition to existing services. Primary cost data collected included capital and recurrent costs. Total provider economic costs were estimated using a combination of ingredients-based costing and step-down cost allocation methodologies. Sensitivity analysis was used to examine the influence of data uncertainty on cost results. Multivariate econometric analysis was used to estimate cost functions. RESULTS: Data was collected from four District Health Authorities in Nova Scotia, representing the delivery of 2951 immunizations. Average cost per immunization ranged from $51 to $105 when delivered in main public health offices and $45 to $150 when delivered in off-site clinics. The main cost driver was personnel costs. Econometric analysis showed a link between average cost and volume of service delivery. CONCLUSIONS:

4th Asia-Pacific Abstracts The cost of public health delivery of childhood vaccines varies according to the volume of services delivered and the delivery setting. The approach developed can be applied to the introduction of new vaccines such as human papilloma virus. Public health delivery of vaccines can be efficient if programmed on the appropriate scale. PIH5 MEDICAL COST AND UTILIZATION FOR PATIENTS WITH POSTPARTUM HEMORRHAGE IN KOREA Jang EJ, Ryu HG, Nam MH, Cho JH, Lee NR, Bae JM National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea OBJECTIVES: To investigate medical cost and health resource utilization for patients with postpartum hemorrhage (PPH) in Korea in 2008. METHODS: This populationbased study utilized the claims data of the Korea Health Insurance Review Agency from January 2008 to December 2008. Patients with PPH were identified based on ICD-10 code (O72, O721, O722, O723) that occurred within 6 weeks after normal or cesarean delivery. Treatments for PPH were classified into four categories: uterine contraction drugs only, drugs and transfusion, drugs + transfusion + uterine artery embolization (UAE), and drugs + transfusion + hysterectomy. Medical costs for PPH patients included all costs incurred for delivery and treating PPH. Costs and length of stay were analyzed in regard to the mode of delivery, age, and type of hospital where the delivery took place. Data were analyzed using SAS. RESULTS: There were 18,142 (4%) PPH patients out of 452,219 deliveries in Korea in 2008. The medical cost (mean ± SD) in patients who underwent normal delivery (ND) without PPH was 610 ± 153 US$ and 758 ± 859 US$ for patients with PPH. The medical cost in patients who underwent cesarean delivery (CD) without PPH was 973 ± 1403 US$ and 860 ± 511 US$ for patients with PPH. Length of stays (mean ± SD) were 3.7 ± 1.8 days, 3.3 ± 0.7 days, 7.7 ± 8.8 days and 7.1 ± 3.2 days for ND with PPH patients, ND without PPH patients, CD with PPH patients, and CD without PPH patients, respectively. The mean medical cost of patients who were treated with drugs only, drugs and transfusion, drugs + transfusion + UAE, and drugs + transfusion + hysterectomy were US$628, US$1028, US$3850, and US$3501, respectively. CONCLUSIONS: The medical cost was higher in patients with PPH compared to patients without PPH. Although the medical cost of treating PPH with UAE is higher than that of hysterectomy, preserving fertility seems well worth the additional cost. PIH6 BUDGET IMPACT ANALYSIS OF INCLUDING LIVER TRANSPLANTATION IN CHILDREN INTO THE UNIVERSAL COVERAGE BENEFIT PACKAGE OF THAILAND Pachanee K1, Prakongsai P2 International Health Policy Program (IHPP), Nonthaburi, Thailand, 2International Health Policy Program (IHPP), Nonthaburi Province, Thailand BACKGROUND: Liver transplantation in children is an expensive health service excluded from the benefit package of the universal coverage (UC) scheme, which covers 47 million of Thais. Its high costs and exclusion from the UC benefit package prevent poor biliary atresia (BA) patients from access to such health care. OBJECTIVES: This study aims to estimate financial impact of including liver transplantation in children in the UC benefit package. It also explores demand for, and supply of, and financial feasibility in implementing universal access to such expensive health care in Thailand. METHODS: Methods include comprehensive literature review, in-depth interviews of medical specialists in liver transplantation in three university hospitals about incurred costs during and after operations, and modeling budget requirements for expanding universal access to such medical care. RESULTS: Research findings indicate that the incidence of BA patients in Thailand is approximately 1 to 15,000 of live births, approximately 60 new cases of BA patient annually. Only three university hospitals in Thailand can provide liver transplantation to children with the annual maximum surgical capacity of 40 cases. Literature shows the survival rate of BA children receiving liver transplantation ranges from 83% to 94% in the first year, and 82% to 92% in the fifth year. Costs of liver transplantation in the first year are approximately 1 million Baht (US$30,000). Other costs after the first year including immunosuppressive drugs and prophylaxis antibiotics range from US$360–600 per month. The government will spend approximately 40 million Baht in the first year of implementation. The financial burden will increase to 184 and 328 million Baht per year in the year 30th and 70th, respectively. CONCLUSIONS: The budget requirement for this medical care is trivial compared to the total UC budget, but can save lives of BA patients and improve equity in access to health care. PIH7 ECONOMIC EVALUATION OF HUMAN PAPILLOMAVIRUS (HPV) VACCINATIONS IN THE PREVENTION OF CERVICAL CANCER Wan Puteh SE National University of Malaysia, Kuala Lumpur, Malaysia OBJECTIVES: Cervical cancers are the second highest incidence of female cancers in Malaysia, causing high impact on nation’s health cost and patient’s quality of life that can be avoided by better screening and HPV vaccination. METHODS: This is a cross sectional study done from 2006–2009 and respondents were interviewed from six public hospitals. Methods include experts’ panel discussions to estimate treatment costs and respondents’ interviews using costing and SF-36 quality of life (QOL) questionnaires. Three programs options were compared i.e., Pap smear screening; quadrivalent HPV vaccination and combined strategy (screening plus vaccination). RESULTS: Five hundred two cervical cancer patients participated in the study. Mean

4th Asia-Pacific Abstracts age was 53.3 ± 11.21 years, educated till secondary level (39.4%), Malays (44.2%) and married for 27.73 ± 12.12 years. Life expectancy gained from vaccination is 13.04 years and average Quality Adjusted Life-years saved (QALYs) is 24.40 in vaccinated versus 6.29 in unvaccinated women. Cost/QALYs saved for Pap smear at base case is RM 1214.96/QALYs and RM 1100.01/QALYs at increased screening coverage. In HPV vaccination, base case is at RM 35,346.79/QALYs and RM 46,530.08/QALYs when vaccination price is increased. In combined strategy, cost/QALYs at base case is RM 11,289.58/QALYs; RM 7712.74/QALYs at best case and RM 14,590.37/ QALYs at worst case scenario. Incremental cost-effectiveness ratio (ICER) showed that screening at 70% coverage or higher is highly cost effective at RM 946.74 per QALYs saved and this is followed by combined strategy at RM 35,346.67 per QALYs saved. Budget impact analysis indicated that it cost the government RM 180.4 million per year and 2.5% of the national health budget. CONCLUSIONS: Vaccination increase life expectancy with better QOL. Cost effective strategies will include increasing the Pap smear coverage to 70% or higher. Since feasibility and long term screening adherence is doubtful among Malaysian women; vaccination of young women is more cost effective strategy against cervical cancer. PIH8 ESTIMATED HEALTH AND ECONOMIC IMPACT OF QUADRIVALENT HPV TYPES 6, 11, 16, 18 VACCINE IN THAILAND USING A TRANSMISSION DYNAMIC MODEL Singhal PK1, Termrungruanglert W2, Khemapech N2, Havanond P2, Taneepanichskul S2 1 Merck & Co., Inc., West Point, PA, USA; 2Chulalongkorn University, Bangkok, Thailand OBJECTIVES: The quadrivalent (6,11,16,18) HPV vaccine has been approved in Thailand for prevention of cervical cancer, vulvar/vaginal pre-cancers, and genital warts in women age 9 to 26 years. To assess the health and economic impact of the quadrivalent (6,11,16,18) HPV vaccine in Thailand. METHODS: A published mathematical model of the transmission dynamics of HPV infection and disease was adapted for Thailand. Model inputs were used from Thailand or the Asia/Pacific region when available; otherwise, the default values in the original model were used. Maintaining current cervical cancer screening practices in Thailand, we evaluated two strategies: routine vaccination of females by age 12 (S1), and S1 combined with a temporary (5 years) female catch-up program for age 12–24 years (S2). The vaccine coverage rates were 70% for the routine and 50% for the catch-up vaccination programs. The perspective of analysis was health-care system. RESULTS: The most effective strategy was S2. Using this strategy over 100 years in the population of Thailand, the estimated cumulative percent reduction in incident HPV 6/11/16/18related genital warts-female, genital warts-male, cervical intraepithelial neoplasia (CIN) grade 1, CIN 2/3, and cervical cancer cases was 72%, 55%, 61%, 63%, and 48%, respectively. The cost-effectiveness ratios were Thai Baht 145,447 or US$4,453 (weakly dominated), and Thai Baht 131,845 or US$4,036 per quality-adjusted lifeyears (QALY) gained for S1 and S2 compared with no vaccination, respectively. CONCLUSIONS: In Thailand, vaccination of females age 12–24 years with a quadrivalent (6,11,16,18) HPV vaccine can reduce the incidence of cervical cancer, CIN, and genital warts at a cost per QALY ratio within the range typically regarded as cost-effective. PIH9 COST EFFECTIVENESS OF CONTRACEPTIVES IN A PACIFIC REGION COUNTRY: COLOMBIA Gutierrez MV1, Canon O2 1 Fundación Salud y Equidad, Bogotá, Colombia; 2Fundación Salud y Equidad, Universidad Jorge Tadeo Lozano, Bogota, Colombia OBJECTIVES: To estimate the relative cost effectiveness of available contraceptives in a pacific region country (Colombia) from a social perspective. METHODS: A Markov model was constructed to simulate costs for 13 contraceptive methods and for no contraceptive method over a 10-year period, with a 5% discount rate. Failure rates, adverse event rates and resource utilization were derived from literature. Sensitivity analyses were performed on costs and effectiveness rates. RESULTS: The four least expensive methods per person month were: copper-T IUD (USD $0.34), vasectomy (USD $1.28), tubal ligation (USD $1.44) and implant (USD $2.19). The costeffectiveness analysis of methods to prevent pregnancy showed: the Copper T IUD has a ratio of incremental cost-effectiveness (ICER) of USD $5,540 (per month, per pregnancy avoided) when it is contrasted with the vasectomy. The Vasectomy is dominant compared to tubal ligation, and it presents an ICER of USD $7.371 in relation to implant methods. Comparing Copper T with the second most affordable reversible method, an ICER of USD $12,033 was obtained. The order of results did not change with a discount rate of 0% but were sensitive to the time horizon. CONCLUSIONS: Differences in time horizon are influential factors that determine the dominance of contraceptive methods. It is noteworthy that although the cost of a tubal ligation and a vasectomy is the same, the total costs for a vasectomy are lower due to reduced complication-related costs. Given that the monthly premium of the compulsory health insurance in Colombia for a woman of childbearing age (19–44 years) is USD$22, the value of USD$12,033 or $5540 per month to prevent pregnancy through implant or vasectomy is excessive. Therefore, according to the case analysis, the Copper T is a cost-effective option for Colombia, including reversible and nonreversible methods.

A541 PIH10 COST-EFECTIVENESS OF QUADRIVALENT AND BIVALENT HPV VACCINATIONS AGAINST CERVICAL CANCER Wan Puteh SE National University of Malaysia, Kuala Lumpur, Malaysia OBJECTIVES: Cervical cancer is the second highest incidence of female cancers in Malaysia. This can be avoided by Pap smear screening and Human Papillomavirus (HPV) vaccination i.e., the bivalent vaccine (BV) and quadrivalent (QV). Three programs cost effectiveness (CE) options were compared i.e., screening via Pap smear; modeling of HPV vaccination (QV and BV) and combined strategy (screening plus vaccination). Scenario based sensitivity analysis using screening population coverage (40–80%) and costs of vaccine (RM 100–200/dose) were calculated. METHODS: This is a cross sectional study from 2006–2009 and respondents were interviewed from six public Gynecology–Oncology hospitals. Methods include experts’ panel discussions to estimate treatment costs by severity and direct interviews with respondents using costing and quality of life questionnaires. RESULTS: A total of 502 cervical cancer patients participated with mean age at 53.3 ± 11.21 years, Malays (44.2%) and married for 27.73 ± 12.12 years. Cost/QALYs for Pap smear at base case is RM 1,214.96/QALYs and RM 1,100.01/QALYs at increased screening coverage. In QV only, cost/QALYs saved in base case are at RM 15,662/QALYs and RM 24,203/ QALYs when vaccination price is increased. In BV only; cost/QALYs saved in base case is at RM 1359,057/QALYs and RM 2530,017.56/QALYs when vaccination price is increased. In QV combined strategy cost/QALYs at base case is RM 4937/QALYs; RM 3395/QALYs at best case and RM 7992/QALYs at worst case scenario. In BV combined strategy, cost/QALYs at base case is RM 6624/QALYs; RM 4033/QALYs at best case and RM 10,543/QALYs at worst case scenario. Incremental costeffectiveness ratio (ICER) showed that screening at 70% coverage or higher is highly cost effective at RM 946.74 per QALYs saved and this is followed by best case combined strategy with QV at RM 13,000 per QALYs saved. CONCLUSIONS: QV is more cost effective than BV. The QV combined strategy is more CE than any method except Pap smear screening at high population coverage. PIH11 IDENTIFYING COST-EFFECTIVE TREATMENT WITH RALOXIFENE IN POSTMENOPAUSAL WOMEN USING RISK ALGORITHMS FOR FRACTURES AND INVASIVE BREAST CANCER Ivergard M1, Strom O1, Borgstrom F1, Burge RT2, Tosteson A3, Kanis JA4 1 i3 Innovus, Stockholm, Sweden; 2Eli Lilly & Company, Indianapolis, IN, USA; 3Dartmouth Medical School, Hanover, NH, USA; 4WHO Collaborating Centre for Metabolic Bone Diseases, Sheffield, UK OBJECTIVES: Raloxifene (RLX) reduces the risk of vertebral fractures and invasive breast cancer (IBC). The National Osteoporosis Foundation (NOF) has recommended a threshold for treatment initiation of 10-year major fracture risk of 20%, but given RLX’s beneficial efficacy on IBC it may be clinically beneficial and cost-effective to treat women with RLX even if they do not meet this threshold. The aim was to identify how 5-year IBC risk affects the cost-effectiveness of treating younger postmenopausal women who do not meet the 20% absolute fracture risk threshold with RLX compared to no intervention. METHODS: A micro-simulation model populated with data specific to American women was used to quantify the costs and benefits of 5-year treatment with RLX from a societal perspective. Possible events were vertebral fracture, IBC, VTE and death. EQ-5D societal health state values were used to estimate quality-adjusted life-years (QALYs). The population was selected based on 10-year major fracture risk as estimated with FRAX® (5–19.9%) and 5-year IBC risk as estimated with Gail risk model (1–5%). RESULTS: The cost per QALY gained ranged from US$17,100 to 93,000. RLX was progressively cost-effective with increasing fracture risk and IBC risk holding age constant. At lower fracture risk in combination with lower IBC risk or when no preventive RLX effect on IBC was assumed, the cost-effectiveness of RLX decreased markedly and was not cost-effective given a willingness-to-pay of US$50,000. At fracture risk of 15–19.9% RLX was cost-effective also in women at lower IBC risk. CONCLUSIONS: RLX is potentially cost-effective in young postmenopausal women at elevated IBC risk who do not meet the suggested NOF 10-year fracture threshold. This highlights the importance of considering a woman’s full risk profile when deciding which anti-osteoporosis treatment to recommend. PIH12 COST UTILITY ANALYSIS OF HPV VACCINATION WITH BIVALENT AS04 ADJUVANT VACCINE IN SLOVAKIA AND IMPACT OF VACCINE PROTECTION DURATION ON COST-EFFECTIVENESS ENDPOINTS Bielik J1, Foltan V2, Psenkova M3, Hroncova D4, Marusakova E5, Sadovsky O6 1 Faculty of Health, Trencin, Slovak Republic; 2Comenius University, Bratislava, Bratislava, Slovak Republic; 3Pharm-In, spol.s r.o., Bratislava, Slovak Republic; 4Mediforum, Noncommercial Educational Center of GSK, Bratislava, Slovak Republic; 5GlaxoSmithKline Slovakia, Bratislava, Slovak Republic; 6National Oncological Institute, Bratislava, Slovak Republic OBJECTIVES: The study evaluated the cost effectiveness of the bivalent AS04 adjuvant vaccine in Slovakia and the correlation between vaccine protection duration and incremental cost effectiveness of HPV vaccination of 12-year-old girls with bivalent AS04 adjuvant vaccine in Slovakia. METHODS: A 1-year cycle Markov model (20— health state) used age-specific data on dysplasia or cervical cancer. A base case analysis assumed 98% vaccine efficacy against HPV—16/18 and 50% vaccination coverage.

A542 Neither cross protection nor herd immunity were assumed in the model. The model was run in different horizon (10, 12, 14, 16, 18, 19, 20, 25, 30, 35, 40, 50, 60, 70, 80, 87 cycles) reflecting hypothetically an estimated duration of vaccine-induced protection. ICER results in different horizon were compared with the λ1 threshold (EUR18,000 QALY) valid in Slovakia in 2010. Both future costs and utilities were discounted by 5%. Adherence to the three-dosage schedule was 2.6 for the covered cohort. A calibration reflects relevant Slovak statistics. RESULTS: Using real-world incidence data on dysplasia and cervical cancer, real local cost data and assuming 50% vaccination coverage rate, vaccination with bivalent AS04 adjuvant vaccine in Slovakia is a cost-effective strategy. Cost/QALY lies well below the local threshold in the base case analysis. The results of the secondary analysis showed that the vaccination value of the incremental cost-effectiveness in different time horizon decreases exponentially (horizon/EUR/QALY): 10-year horizon/89778 EUR/QALY, 12/55655, 14/37383, 16/27207, 18/20262, 19/18028, 20/16076, 25/9903, 30/6825, 35/5000.40/3873, 50/2643, 60/1938, 70/1613, 80/1514, 87/1510. CONCLUSIONS: Taking into account conservative mathematical prediction of 20 years of sustainable duration of antibody levels and Slovak threshold, the study indicates a very high probability that vaccination of 12-year-old girls with the bivalent AS04 adjuvant vaccine in Slovakia is a cost-effective strategy.

INDIVIDUAL’S HEALTH – Patient-Reported Outcomes Studies PIH13 COMPLIANCE WITH HORMONE THERAPY AMONG MENOPAUSAL WOMEN AFTER THE WHI PUBLICATION IN TAIWAN Chen WC Tang CH, Taipei Medical University, Taipei, Taiwan OBJECTIVES: The publication of results from the Women’s Health Initiative (WHI) has led to a drop in the use of hormone therapy (HT) among postmenopausal women. This study aimed to compare the magnitude of changes in the utilization of HT before and after the WHI report among women with different socioeconomic status in Taiwan. METHODS: Women who were aged 40–69 and regular estrogen users for menopausal symptom in the Taiwan National Health Insurance (NHI) claims database during January to June 2001 were identified as the study subjects (N = 57,358). Number of outpatient visits with estrogen prescription, total dosage of estrogen, and proportion of oral estrogen dosage to total estrogen dosage, were compared on an annual basis, between the pre-WHI period (July 2001–June 2002) and the post-WHI period (January 2003–December 2003), using generalize estimating equation models. Interaction terms of socioeconomic factors and pre- and post-WHI period were entered to examine the differential effects of women’ socioeconomic characteristics on the changes of HT use, after adjusting other patients’ and providers’ characteristics. RESULTS: Significant decreases in number of outpatient visits with estrogen prescriptions, total dosage of estrogen prescription, and the proportion of oral or injection estrogen reduced after WHI (P < 0.0001). However, the extents to which this compliance behavior changes varied by women’ socioeconomic characteristics. Women with higher wage income level and living in areas of higher unbanizational levels had lower number of outpatient visits with estrogen prescriptions, total dosage of estrogen prescription, and proportion of oral or injection estrogen, compared to their counterparts. CONCLUSIONS: There seemed to exist discrepancy in the compliance behavior of HT use among women with different socioeconomic status after the publication of WHI report. Interventions to improve the accessibility of health information among less privileged women with which resulted in differential changes in health behavior among women in different socioeconomic groups. PIH14 VALIDITY OF EQ-5D AND SF-12 AMONG MALAYSIAN 1 1 2 Shafie AA , Hassali MA , Liau SY 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Universiti Sains Malaysia, Penang, Malaysia OBJECTIVES: The objective of this study was to assess the construct validity of the EQ-5D instrument among Malaysian population. METHODS: This was a crosssectional study conducted among Malaysians adults in the northern states of Malaysia. A pre-developed questionnaire consisting of both the EQ-5D and SF-12 items was used for data collection. Concurrent and convergent validity of EQ-5D were assessed by the use of independent t-test and Pearson correlation. Finally, construct validity was assessed using three known relationship. RESULTS: A total of 596 valid questionnaires were received and analyzed. The participants were predominantly female, Malay ethnic origin and married, with a mean age of 35.67 years (SD = 12.73). The mean EQ-5D score was 0.93 (SD = 0.13) and median of 1 (IQR = 0.15). Majority of the participants (73.2%) did not have any problem in all 5 dimensions of the EQ5D questionnaire. The mean PCS-12 and MCS-12 scores were 48.9 (SD = 7.4) and 49.1 (SD = 8.0) respectively. Participants with present disease have lower PCS-12 and MCS-12 scores and reported more problems with all of EQ-5D dimensions as well as lower EQ-5D and EQ-VAS scores (P < 0.05) as compared to those without disease (P < 0.05). Convergent validity was supported by a moderate and positive correlation between EQ-5D and EQ-VAS with MCS-12 and PCS-12 scores as well as the stronger effect size between PCS-12 and the physical dimensions of EQ-5D as well as between MCS-12 with anxiety/depression scores. Construct validity of EQ-5D was confirmed by lower score in anxiety dimension among female and older people (P < 0.05). CONCLUSIONS: This study has demonstrated the construct validity of EQ-5D among Malaysian population.

4th Asia-Pacific Abstracts PIH15 SENSITIVITY OF GENERIC PATIENT-REPORTED OUTCOMES MEASURES AMONG CAREGIVERS OF CHILDREN WITH CRANIOFACIAL MALFORMATION Payakachat N1, Grosse SD2, Tilford JM3 1 College of Pharmacy, University of Arkansas for Medical Sciences, Little Rock, AR, USA; 2 Centers for Disease Control and Prevention, Atlanta, GA, USA; 3College of Public Health, University of Arkansas for Medical Sciences, Little Rock, AR, USA OBJECTIVES: Family effects and caregiver burden should be included in economic evaluations of child health interventions, but the methods to include these effects remain understudied. This research explores the sensitivity of three generic instruments for preference-weighting health states of family caregivers of children with craniofacial malformation (CFM). METHODS: Caregivers of children born with CFM were identified through the Arkansas Birth Defects Monitoring System. A survey instrument was mailed that included the Health Utilities Index (HUI3), the 6-Dimension ShortForm (SF-6D), the Quality of Well-being (QWB-SA), and two health-related quality of life measures for caregivers: the CES-D and the CarerQol. The HUI3, the SF-6D, and the QWB-SA were examined in relation to the CES-D, the CarerQol, and each other. RESULTS: Sixty-five parents of minor children responded for a completion rate of 57%. The mean SF-6D, HUI3, QWB-SA scores were 0.80, 0.84, 0.67. The mean CES-D score was 13.3 with 28.6% of the sample scoring 16 or above—a common threshold for depressive symptoms. The mean of the CarerQol was 2.8 (0-not at all caregiving burden; 8-much caregiving burden). The correlations of the HUI3 and the SF-6D with the CES-D were similar (R2’s of 0.513, 0.511) while the R2 was only 0.230 for the QWB-SA. There was a significant difference in the preference-weighted scores of caregivers scoring 16 and above on both the SF-6D and the HUI3. All three generic instruments showed moderate relationships with the CarerQol. CONCLUSIONS: The HUI3 and SF-6D are more sensitive among the caregiver group than the QWB-SA. They might be appropriate to use in calculating quality-adjusted life-year outcomes for caregivers in economic evaluation of child health interventions. Variation of preference-weighted scores of the generic instruments in caregivers should be further investigated.

PIH16 CONTENT AND PSYCHOMETRIC VALIDITY OF THE EYELASH SATISFACTION QUESTIONNAIRE (ESQ) IN CHINA Crawford B1, Tanaka E2, Zhu Y3, Burgess SM4 1 Mapi Values, Tokyo, Japan; 2Mapi Values, Tokyo, Japan; 3Mapi Values, Boston, MA, USA; 4 Allergan, Inc, Irvine, CA, USA OBJECTIVES: The Eyelash Satisfaction Questionnaire (ESQ) was developed to gather consumer information about their overall satisfaction with eyelash appearance. The ESQ was not available in Chinese and therefore needed to be translated and validated. METHODS: Following the translation of the ESQ using a forward-backward methodology, the ESQ (23 items) was cognitively debriefed and assessed for relevance in 15 Chinese women. It was scored according to the developers’ recommendation. Validation of the ESQ in 150 women consisted of the evaluation of floor/ceiling effects, internal consistency reliability, item-item correlations, multitrait analysis, test–retest, and known-groups methods for comparing subject groups. The 9-item ESQ (ESQ-9) was also evaluated using the same psychometric methods. RESULTS: Chinese women found the 23 items of the ESQ relevant and acceptable. In the ESQ, the Length, Fullness, Overall Satisfaction (LFOS) domain had no difficulties replicating its structure, and only one item each in the Confidence, Attractiveness, and Professionalism (CAP) and Daily Routine (DR) domains was problematic. The reduced ESQ-9 had no floor or ceiling effects, had acceptable internal consistency reliability (α: 0.76–0.86), and good test–retest reliability (ICC: 0.86–0.89). Item convergent and discriminant validity for the ESQ-9 were very good with all items correlating with their own scale greater than 0.84 and with other scales less than 0.47, with all tests demonstrating similar statistics as the original US-English version. Internal consistency was acceptable and only slightly lower in the Chinese version than the US-English version. The item convergent validity demonstrated much stronger correlations for the DR domain than the US-English version. CONCLUSIONS: The ESQ-9 is a valid measure of satisfaction with eyelash appearance among Chinese women. Although the full version was acceptable to women, several items did not fit well into their respective domains. It is therefore recommended that the ESQ-9 be used in studies.

PIH17 GENERAL PUBLIC SATISFACTION TOWARD THE PROVISION OF HEALTH-CARE DELIVERY BY GOVERNMENT HOSPITALS AND CLINICS IN THE STATE OF PENANG Al-Haddad MS1, Hassali A2, Mohamed Ibrahim MI3, Shafie AA4 1 Universiti Sains Malaysia, Gelugor, P. Penang, Malaysia; 2Discipline of Social & Administrative Pharmacy, Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia; 3Universiti Sains Malaysia, Penang, Malaysia; 4Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: 1) To determine the level of satisfaction of general public toward the provision of health-care delivery by government hospitals and clinics in the state of Penang; and 2) to find out factors that may be associated with the general public’s satisfaction toward public health-care delivery in the state of Penang. METHODS: A cross-sectional study using a prevalidated survey was undertaken with a convenience sample of 1000 residents in the State of Penang, Malaysia. At the end of the survey, a total of 771 residents had agreed to participated (response rate of 77.1%). All the

A543

4th Asia-Pacific Abstracts data gathered was analyzed using SPSS version 12.0. Chi-square test was used to test the association between demographic profiles and level of satisfaction. All results with P-value less than 0.05 were considered significant. RESULTS: A high percentage of the respondents (n = 369, 47.8%) were satisfied with the provision of health-care delivery at the government hospitals and clinics. They were satisfied with the competency of doctors and staffs and standard of the facilities and services. However, 63.1% (n = 486) of the respondents had expressed their dissatisfaction toward the waiting time in the government hospital and clinic. The finding also showed that the factors that influenced the level of satisfaction were ethnicity and age. CONCLUSIONS: Most of the people in Penang were satisfied with the health-care delivery at the government hospitals and clinics. However, the waiting time in the hospitals and clinics need to be improved as an overwhelming proportion of the respondents were discontented with it. PIH18 QUALITY OF LIFE IN THAI WOMEN DIAGNOSED WITH GENITAL WARTS, CERVICAL CANCER, AND CERVICAL INTRAEPITHELIAL NEOPLASIA AT KING CHULALONGKORN MEMORIAL HOSPITAL Khemapech N1, Havanond P1, Termrungruanglert W1, Lertmaharit S1, Pongpanich S1, Khorprasert C1, Jirakorbchaipong P2, Kitsiripornchai S2, Taneepanichskul S1 1 Chulalongkorn University, Bangkok, Thailand; 2MSD (Thailand) Ltd., Bangkok, Thailand OBJECTIVES: To determine health related quality-of-life (QoL) of patients diagnosed with genital warts, cervical cancer, and cervical intraepithelial neoplasia (CIN), compare QoL across genital warts, CIN1, CIN2/3 and stages of cervical cancer and to study association between QoL and general characteristics of patients. METHODS: The cross-sectional descriptive study was conducted at the King Chulalongkorn Memorial Hospital using a standard questionnaire. The questionnaire elicited information on general patient characteristics such as age, occupation, and educational attainment, and on QoL using Functional Assessment of Cancer Therapy-General (FACT-G) questionnaire. The study population were patients with a diagnosis of one of the four FIGO stages of cervical cancer, cervical intraepithelial neoplasia low-grade (CIN1) or cervical intraepithelial neoplasia high-grade (CIN2/3) or genital warts as confirmed by physical and pathological examination. RESULTS: A total of 197 patients completed the study questionnaire with a diagnosis of genital warts (n = 25), CIN1 (n = 25), CIN 2/3 (n = 25), IA1 (n = 25), IA2 IB IIA (n = 37), IIB-IVA (n = 32) and IVB (n = 28). The mean age was 47.3 years and majority were housewives with educational attainment of primary school or lower. The adjusted mean scores were as follows: global QoL 78.08 (95% CI = 76.4, 79.8); physical well-being, 81.9 (95% CI = 79.6, 84.3); social well-being, 72.5 (95% CI = 70.1, 74.9); emotional well-being, 78.7 (95% CI = 76.3, 81.2); functional well-being, 79.3 (95% CI = 76.9, 81.7). There were no significant differences between stages of diseases and global QoL or subscales. We found a positive association between age and emotional well being scores (P = 0.001). Lower educational attainment was associated with higher emotional well being scores (P = 0.049). CONCLUSIONS: Our finding shows that, impact of QoL on disease of CIN or genital warts did not differ from cervical cancer. Young HPV related-disease patients and those with higher education may experience emotional difficulties. Social, family or religion support may be considered for these patients. PIH19 ETHNIC/RACIAL DIFFERENCE IN SELF-ASSESSMENT OF HEALTH STATUS IN THE GENERAL US POPULATION Luo N1, Ko Y1, Fu AZ2, Johnson JA3 1 National University of Singapore, Singapore; 2Cleveland Clinic, Cleveland, OH, USA; 3 University of Alberta, Edmonton, AB, Canada OBJECTIVES: In population health surveys, self-reported health status is usually assessed by a single question. In spite of good face validity, such a measure may be susceptible to differential item functioning (or measurement bias). This study aimed to examine the differences in health status measured by the self-assessed health status (SAHS) question and the EQ-5D visual analog scale (EQ-VAS) between Hispanics, blacks, and others (all other ethnicities/races as one group) in the US. METHODS: We used the data collected in the US Valuation of the EQ-5D Health States study where a representative sample (N = 4048) of the non-institutionalized adult population was surveyed in 2002. Hispanics and blacks were oversampled for the survey. In face-to-face interviews, all respondents answered the SAHS question (“overall, how would you rate your health?”) using a 5-point Likert-type scale (excellent/very good/ good/fair/poor) and the EQ-5D which includes a vertical, 0–100 VAS for assessing global health. RESULTS: Excellent/very good health was reported by 64.4%, 65.1%, and 70.8% of Hispanics, blacks, and others, respectively. Mean (standard deviation) VAS scores for the Hispanics, blacks, and others were 84.7 (0.8), 83.5 (1.0), and 83.4 (0.6), respectively. In an ordinary least-square model, Hispanics (regression coefficient: 0.8) and blacks (regression coefficient: 0.2) were not different in VAS scores compared to others after adjusting for socio-demographics, smoking status, and presence of chronic conditions (both P > 0.05). In contrast, Hispanics (odds ratio: 0.51) and blacks (odds ratio: 0.64) were less likely to report excellent/very good health than others in a logit model with adjustment of other factors (both P < 0.001). CONCLUSIONS: Hispanics, blacks and other ethnicities in the US respond similarly to the EQ-VAS but differently to the SAHS question. Caution should be exercised when interpreting selfreported health status of culturally diverse populations as assessed by single-item measures.

PIH20 HEALTH-RELATED QUALITY OF LIFE IN CHINESE CHRONIC PROSTATITIS PATIENTS Zhao FL1, Yue M2, Yang H3, Wang T3, Wu JH2, Li SC1 1 University of Newcastle, Callaghan, NSW, Australia; 2306 Hospital of PLA, Beijing, China; 3 The First People’s Hospital of Yunnan Province, Kunming, Yunnan, China OBJECTIVES: To examine the Health-related quality of life (HRQoL), and factors associated with HRQoL in Chinese CP patients using two generic preference-based HRQoL instruments, EQ-5D (plus EQ-VAS) and SF-6D, with the results compared with general population. METHODS: CP patients were recruited from two tertiary referral hospitals and the general populations were randomly approached in China. After informed consent, subjects were interviewed using EQ-5D, EQ-VAS and SF-6D, and their socio-demographic and medical information were solicited. RESULTS: Compared with the general population (n = 364), CP patients (n = 268) reported significantly worse HRQoL with median score of the EQ-5D utility index (0.73 vs. 0.85), SF-6D utility index (0.76 vs. 0.81) and EQ-VAS (70.0 vs. 85.0). Multiple linear regression analyses showed pain symptom had the strongest predictive power for HRQoL, compared with symptom duration and urinary symptom. Socio-demographic factors and comorbidities did not significantly contribute to poorer HRQoL. CONCLUSIONS: CP patients experienced deteriorated HRQoL with lower health-related utility scores compared to general population and pain severity was the main physical symptom predicting decreased health-related utility. Further studies are needed to provide the reference utility index for the comparison and better characterizing the influence of geographic and cultural factors in variation of health-related utility of CP patients. PIH21 QUALITY OF LIFE IMPACT IN PARENTS CARING FOR SEVERE GASTROENTERITIS IN CHILDREN IN SEVEN COUNTRIES WORLDWIDE Curran D1, Lamotte M2, Van Kriekinge G3 1 Omega Research, Dublin, Ireland; 2IMS Health Consulting, Brussels, Belgium; 3GSK Bio, Wavre, Belgium OBJECTIVES: To assess QoL impact of parents caring for children with severe gastroenteritis (GE) world wide using a questionnaire. Pooled analysis was carried out to investigate the relationship between disease severity and QoL. METHODS: A questionnaire was developed to retrospectively assess the disease severity and QoLimpact of parents with hospitalized children (16 years old) in 10 randomly selected community pharmacies in seven different states in Malaysia by trained data collectors from May 2008 to June 2008. The self-administered questionnaire includes the EuroQoL EQ-5D (Health State Classification), EQ-VAS (thermometer) and demographic questions. Data were analyzed using Kruskal-Wallis and Mann Whitney tests in SPSS v15. RESULTS: A total of 599 customers was included in this study with mean EQ-5D score of 0.90 (SD = 0.16) and VAS score of 0.74 (SD = 0.17). This study found that pain/discomfort (30.36%) and anxiety/depression (13.3%) were the major HRQoL problems among Malaysian pharmacy customers. Elderly had lower mean EQ-5D scores and VAS were lower for older age group and reported more problems in every dimension of EQ-5D Health Classification. This study also found that there was significant differences between EQ-5D score and VAS-score (P < 0.001). CONCLUSIONS: Community pharmacy customers’ HRQoL who are using non-prescription medicine is comparable with the general population. The HRQoL of the customers is affected by their age, gender, ethnicity, education, and household incomes. PIH28 DETERMINANTS OF HEALTH SERVICE UTILIZATION FOLLOWING THE 2004 TSUNAMI IN THAILAND Isaranuwatchai W1, Guerriere DN1, Andrews G2, Coyte PC1 1 University of Toronto, Toronto, ON, Canada, 2McMaster University, Hamilton, ON, Canada OBJECTIVES: On December 26, 2004, a massive earthquake struck Indonesia, triggering a tsunami that affected several countries. This catastrophic event had important implications for health system planning; and thus, underscores the need to study the long-term impact of this disaster. This prospective cohort study aims to identify the determinants of health service utilization in tsunami-affected provinces of Thailand, 1 and 2 years after the disaster. The relationship between being affected by the tsunami and health service utilization were also investigated. METHODS: Participants were randomly selected Thai citizens (aged 14+), living in Phuket, Phang Nga, Krabi, or Ranong. Approximately 1 and 2 years after the tsunami, participants were interviewed in-person on demographic and socio-economic characteristics, disaster impact, health status, and health service utilization, using five questionnaires. Five types of health services were examined: outpatient, inpatient, home care, medications, and informal/ unpaid care. Two-stage sample selection model, more specifically probit model and negative binomial model with Heckman correction, were employed to identify determinants of the probability and intensity of utilization, respectively. RESULTS: There were 2079 participants in the first interview and 1989 in the second interview. Determinants of health service utilization depended on the type of health service being examined, and the assessment time after the disaster. Being affected by the tsunami was significantly associated with the propensity to use medications, 2 years after the disaster, P = 0.0125. CONCLUSIONS: The study findings can assist health providers in identifying population at risk of using health services. Accordingly, regional health officers can request appropriate amount of health resources in order to deliver high quality care tailored to specific population. The findings may also assist policy-makers in the development of long-term disaster recovery plan. Additionally, the study findings may not be specific to a tsunami disaster and may provide insights on post-disaster contexts of other natural disasters.

4th Asia-Pacific Abstracts PIH30 THE DRINKING BEHAVIOR OF RESIDENTS OF A VILLAGE IN AN INDUSTRIAL PARK AREA: THE CURRENT SITUATION AND REASONS BEHIND THIS BEHAVIOR Khampang R, Yothasamut J, Putchong C, Pilasant S, Teeraananchai S, Tantivess S, Teerawattananon Y Health Intervention and Technology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: To identify the current situation regarding the alcohol consumption of factory workers and other residents living in a village located in an industrial park on the outskirts of Bangkok and explore the relationship between exposure to antialcohol campaigns introduced by the Thai Health Promotion Foundation and drinking behavior. METHODS: Participatory and non-participatory observations were employed to understand the lifestyles of the study population, determine risk behaviors and factors, and identify exposure to anti-alcohol campaigns. After the first set of qualitative observational data collected during the first 10 weeks was analyzed, a survey questionnaire was developed. Seven hundred ninety-nine residents were subsequently included in this quantitative survey. Finally, focus groups were organized to verify findings obtained from the observations and survey. RESULTS: Alcohol consumption was determined to be major risk behavior. More than 50% of interviewees drank alcohol, 5% were under 20 years of age, and 29% were hazardous drinkers. Factory workers are more likely to drink than those not working in factories. Onethird of drinkers were also cigarette smokers. Although the Alcohol Control Bill limits the time of selling alcohol (11am–14pm and 6pm–12pm), 36% of those buying alcohol bought it outside the permitted times. Several anti-alcohol campaigns were implemented and most respondents recognized them well; however, no significant correlation between drinking behavior and exposure to the anti-alcohol campaigns was found. This may be explained by a number of factors, e.g., easy access to alcohol and alcohol delivery services to homes, both of which enable easy alcohol consumption. CONCLUSIONS: Factory workers had a higher rate of alcohol consumption than non-factory workers. Although anti-alcohol campaigns targeted changing social attitudes toward alcohol were implemented, drinking behavior remained a major problem. It is recommended that additional public measures e.g., effective law enforcement need to be implemented alongside public campaigns. PIH31 THE EFFECTIVENESS OF SOCIAL MARKETING INTERVENTIONS IN REDUCING ALCOHOL CONSUMPTION AMONG CONSTRUCTION WORKERS AND OTHER RESIDENTS IN A COMMUNITY IN BANGKOK, THAILAND Yothasamut J, Teerawattananon Y, Tantivess S, Putchong C, Pilasant S, Khampang R, Butchon R Health Intervention and Techology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES: To assess the effectiveness of social marketing interventions in reducing alcohol consumption among construction workers (CWs) and other residents in a community in Bangkok, and to identify the opportunity to expose to, ability to understand, and motivation to follow the suggestions provided through the interventions among the study populations. METHODS: A cross-sectional survey was conducted and 417 samples (52% were CWs) were interviewed in February 2010. Two interventions for reducing alcohol consumption, namely the television- and radiobased “stop drinking during Buddhist-lent period” campaign and the introduction of a law imposing restriction on alcohol-selling time, i.e., during 11.00–14.00 and 17.00–24.00, were addressed. RESULTS: Respondents were not different for their ability to access to public media. Although both groups had the same level of exposure to and understanding on the information distributed through the stop drinking campaign, a smaller fraction of CWs stopped drinking during the last Buddhist-lent period than its counterpart. Those who stopped drinking argued they were motivated to do so as the campaign fit well with their religious beliefs. In contrast to the stop-drinking campaign, the information on the legal-based restriction of alcohol selling time was not obtained by most of the CWs as they rarely visited large convenience stores, where such information was publicized. Instead, they usually bought alcohol at groceries in the community where law enforcement was inadequate and beverages were sold all the time. Therefore, the effectiveness of the legal-based intervention was higher when introducing among other residents than the CWs owing to several elements. CONCLUSIONS: This study illustrates the limitations of social marketing campaigns and highlights the importance of social, cultural, and economic factors as determinants of the intervention’s effectiveness. Accessibility and understanding of information disseminated through social marketing campaigns are crucial. Social marketing interventions need to be introduced with other measures, i.e., effective law enforcement. PIH32 EVALUATING THE PERCEPTION OF ACADEMIC PHARMACIST TOWARD THEIR ROLE IN HEALTH-CARE SYSTEM OF PAKISTAN Azhar S1, Hassali MA2, Izham M1 1 Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia; 2Discipline of Social & Administrative Pharmacy, Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia OBJECTIVES: To investigate academic pharmacists’ perception toward their role in Pakistan’s health-care system. METHODS: A quantitative approach was used to gain the understanding of academic pharmacist role. The study took place in all four provinces of Pakistan, from January 2009 until April 2009, using a questionnaire, developed after extensive literature review. The study population consisted of academic pharmacists from public and private universities of Pakistan. A total 131

A545 academic pharmacists were responded from 6 public and 4 private universities. RESULTS: Majority 93.9% of the respondents were satisfied working as academic pharmacist, 68.7% expressed that their interest in research is the reason embarks carrier as an academic pharmacist. Only half (55.7%) of the respondents were agree about pharmacy curriculum standard in Pakistan, and their satisfaction toward curriculum is less, as after the implementation of Pharm D in the country, the curriculum is still focusing more toward theoretical knowledge rather than practical. In this case significant difference was noted with respect to university (P = 0.012). Academic pharmacists were appeared to be not satisfied with respect to the Pharm D program in all the Pakistani pharmacy school. Half of the respondents 53.4% were agree, followed with 32.1% were disagree that Pharm D will help to develop better future practitioners. In this case significant difference was marked with respect to gender (P = 0.044). CONCLUSIONS: Academic pharmacists do have concern about the pharmacy profession in Pakistan. They attempted to identify the key issues facing the profession and its educators. Suggestions have been made to improve the current pharmacy curricula in order that future pharmacists in Pakistan practice effectively with the collaboration of other health-care professional. PIH33 COMMUNITY PHARMACIST PERCEPTION TOWARD THEIR ROLE IN HEALTH-CARE SYSTEM OF PAKISTAN: A QUANTITATIVE ASSESSMENT Azhar S1, Hassali MA2, Izham M1 1 Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia; 2Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: Evaluating the perception of community pharmacists’ regarding their role in Pakistan’s health-care system. METHODS: A quantitative approach was used to gain the understanding of community pharmacist role. Study conducted in three cities of Punjab province Islamabad, Faisalabad and Lahore, using a questionnaire, developed after extensive literature review. One hundred and four community pharmacists were responded. The questions asked, were aimed to evaluate their perception regarding the current role as community pharmacist followed with their experience with other health-care professional. RESULTS: Most of the community pharmacists (94.2%) in Pakistan perceive their role as patient educator. Around (n = 102 98.1%) of the respondents expressed that their role is more focused toward the record keeping of the pharmacy. A big majority of the community pharmacist (n = 100 96.2%) thinks that their role should be more toward provision of health services other than just dispensing medicines. Most of the pharmacist (n = 101, 97.1%) expressed that they follow guidelines given by the authorities to run a pharmacy. As far as collaboration with other health-care profession is concern (n = 66 63.5%) of the respondents were disagree that there is no collaboration at present in the system. In this case significant difference was marked with respect to gender (P = 0.000). CONCLUSIONS: The findings indicate that the pharmacists in Pakistan have concern about their present role as community pharmacist in health-care system. Pharmacists are willing to work as team with other health-care professionals. They are currently struggling to get their desired role directly to be involved in medicine use and patient care, suggestions have been made to improve their present image in the society. PIH34 DRUG COST REDUCTION IN LONG-TERM CARE PATIENTS WITH CHRONIC DISEASE Su HC1, Chen CH2, Lin CA1, Juang SY3, Liang TC1, Chan AL1 1 Chi-Mei Medical Center, Tainan, Taiwan; 2Chi Mei Medical Center, Tainan, Taiwan; 3ChiMei Medical Center, Tainan, Taiwan OBJECTIVES: The aim of this study was to assess the potential benefit of pharmacist home care visit for elderly patients who have at least one or more than one chronic diseases at long term care. METHODS: Using the claim database of Chi-Mei Medical Center from 2007 to 2008, patients aged 65 years and treated with over five medications were recruited. Pharmacists visited their home once a month to give them education on drug administration, drug interaction, duplicated drug use, adverse drug reaction etc and then followed up by telephone to see their condition after education. Pharmacists were created drug profile for each patient and recorded the items of education and discussed with their visiting physicians about their drug regimen in case there were inappropriated drug usage. RESULTS: At first visit, the average number of drugs prescribed to one patient was 5.89. There were 43.07% (28/65) and 27.69% (18/65) patients treated with poly-pharmacy and inappropriate drug use, respectively. The association of polypharmacy and inappropriate medication was significant. (P < 0.001; 95% confidence interval: 82.56). The most common medication-related problems were the use of medication without proper indication (38.46%, 25/85), inappropriate administration route (34.85%, 22/65), poor compliance (55.38%, 36/65). The potential adverse drug reactions and drug interactions appeared in 8 patients and 12 patients, respectively. The outcome of pharmacist home care services include the reduction rate of outpatient visit from 1.38 times to 0.98 times every month; the decrease of items of drugs prescribed in one prescription was from 5.89 to 3.21; The average drug cost was reduced about NT 15 million per year (P < 0.05). About 42% of patients satisfied with pharmacists’ home visit. CONCLUSIONS: The pharmacist home care service provides an assessment on many problems with drug administration not otherwise detected easily. These assessments can lead to potentially useful interventions that can improve medication regimens, compliance and reduce drug costs.

A546 PIH35 DETERMINATION OF COMMUNITY PHARMACY’S DISPENSING COST IN MALAYSIA Shafie AA, Hassali MA Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: The main objective of this research is to determine the dispensing cost in community pharmacy and to evaluate the factors that affect the costs. METHODS: This was a cross sectional study on the community pharmacist in Penang Island. Convenient sampling was used whereby 10 community pharmacies in Penang area were selected. A 17 items self administered questionnaire was developed consisting of question on the type of pharmacy license, pharmacist salary, non pharmacist salary, pharmacy’s area size, dispensing’s area size, overhead and capital resources in each pharmacy. Cost for dispensing prescription were calculated and Kruskal Wallis test was used to compare the dispensing cost with the type of pharmacy license, postcode, weekly trading hour and availability of pharmacist personal office. RESULTS: The study found that dispensing cost in a community pharmacy is MYR 0.39 (SD = 0.08) per minute. A standard prescription with less than 5 medicines usually takes 10 minutes of dispensing time and this would cost about MYR 3.91 (SD = 0.80). The pharmacist’s salary contributes the highest percentage for the pharmacy dispensing cost which was 38.06% followed by rental (24.92%) and non-pharmacist staffs’ salary (23.25%). CONCLUSIONS: This study has estimated that the dispensing cost in a community pharmacy for a standard prescription to cost about MYR 3.91 (US$1.18). The major cost drivers in the pharmacy dispensing cost are pharmacist salary, and rent. This study would provide empirical basis for a proper reimbursement structure for pharmacy dispensing service in Malaysia. PIH36 ASSISTED REPRODUCTIVE TECHNOLOGIES: A HEALTH TECHNOLOGY ASSESSMENT PERSPECTIVE Hiller JE, Elshaug AG, Watt AM, Astute SG University of Adelaide, Adelaide, South Australia, Australia OBJECTIVES: Internationally, public funding for assisted reproductive technologies (ART) faces constant scrutiny; however the arguments employed—both for and against—rarely consider the procedures from a health technology assessment (HTA) perspective. This project sought to evaluate the impact of female age, male age and cycle rank on the safety, effectiveness and cost-effectiveness of ART through systematic reviews. METHODS: Using an a priori protocol, six databases were searched for relevant primary studies and existing HTAs. After meeting the inclusion criteria, studies underwent appropriate quality assessment and standardised data extraction. RESULTS: The review is currently ongoing. The majority of the 70 included studies consider the impact of female age on the success of treatment; there are little data investigating the contribution of male age or cycle rank to treatment outcome. There are substantial flaws in the evidence base that challenge the application of established HTA methods to ART. There is considerable heterogeneity in the evidence base, and the policy- and patient-relevant outcome of live birth is not well reported, with many studies presenting pregnancy (variously defined) as the primary outcome. Many studies fail to account for inherent sources of confounding within their patient populations, thus rendering outcome estimates unreliable. Despite broadening the scope of the review to include national registry data, there remain key questions that cannot be confidently answered with the existing evidence base. The final findings of the review will be discussed, alongside their implications for policy-making and public funding in this area. CONCLUSIONS: Thirty years of clinical practice in ART has created a disparate evidence base that cannot support the application of an evidence-based assessment (HTA) framework. ART exemplifies the contextual nature of HTA and the challenge to researchers and policymakers in generalizing the evidence from which to localize the decision. PIH37 RELATIONSHIP BETWEEN HEALTH-CARE PROFESSIONALS ND POPULATION HEALTH STATUS: DO NUMBERS COUNT? Shafie AA, Hassali MA Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: This study aims to assess the relationship between the quantity of health-care providers, a country income and the health status of population. METHODS: This was a cross sectional study on the data collected from World Health Organization (WHO) and World Bank database. Only data from 78 countries between the years 2003–2008 were selected among the 194 listed countries as they provide the most complete data for analysis. Relationship between health-care professionals, countries’ income and country contributions to neonatal mortality rate, infant mortality rate, adult mortality rate, health life expectancy and life expectancy were assessed using Spearman correlation analysis in SPSS v15. RESULTS: Most of the countries included can be categorized as low income (n = 36), 25 as middle income countries and 11 as high income countries. Health-care professionals were found to be uncorrelated with the infant mortality rate of the population except for moderate correlation between environment and public health workers (r = −0.31; P < 0.05). The number of dentistry personnel was moderately correlated with infant mortality rate, healthy life expectancy and life expectancy at birth of both sexes with P < 0.05. A country income was however found to have a strong correlation (r = −0.52 to −0.63); P < 0.05) with all assessed health status indicator. CONCLUSIONS: The number of health-care professionals may not play a role as significant as a country income for the population overall health.

4th Asia-Pacific Abstracts PIH38 EVALUATING THE HEALTH TECHNOLOGY ASSESSMENT PROGRAM IN THAILAND Cairns J1, Chalkidou K2, Panpiemras J3, Suriyawongpaisal P4, Yothasamut J5 1 London School of Hygiene and Tropical Medicine, London, UK; 2National Institute for Health and Clinical Excellence, London, UK; 3Thailand Development Research Institute, Wangthonglang, Bangkok, Thailand; 4Faculty of Medicine Ramathibodi Hospital, Ratchatewi, Bangkok, Thailand; 5Health Intervention and Techology Assessment Program (HITAP), Nonthaburi, Thailand OBJECTIVES AND METHODS: The review aimed to assess the appropriateness and performance of the newly established HTA agency in Thailand. The Health Intervention and Technology Assessment Program (HITAP), a research institute under Thailand’s Ministry of Public Health, was formally established in January 2007, to be responsible for appraising a wide range of health interventions. During November 2008 to March 2009, HITAP vision and mission, as well as the program output were evaluated by domestic and international experts with logistic assistance from HITAP staff. Document reviews, in-depth interviews with key informants, and focus group discussion were employed as evaluation methods. RESULTS AND CONCLUSIONS: The study found that HITAP has been both effective and efficient in building up HTA capacity and delivering related researches to inform policy decisions in Thailand. A key success of HITAP influencing policy decisions was that HITAP developed four interlink strategies, namely 1) establishing national standards and a body of knowledge for HTA in the country; 2) building up researchers’ competence and organizational capacity; 3) conducting comprehensive HTA studies using the standard methodological guidelines and tools developed within the first strategy; and 4) developing HTA systems and mechanisms which are effective, transparent, and acceptable to stakeholders. It could reasonably be argued that these interlink strategies were promising as a means of fillings the major gaps e.g., lack of research capacity and infrastructures for HTA, lack of knowledge /understanding of HTA and mistrust of methods among potential users, lack of timely and good quality of evidence. Because these similar problems were also identified in other resource-poor settings, experiences and lessons learnt from HITAP development are likely to be useful for other low and middle income countries interested in setting up HTA agencies.

INFECTION – Clinical Outcomes Studies PIN1 SURGICAL SITE INFECTION IN CHINA: A SYSTEMATIC REVIEW OF THE INCIDENCE AND ECONOMIC BURDEN Tan JT1, Coleman K1, Norris S1, Metz L2 1 Health Technology Analysts Pty Ltd, Sydney, NSW, Australia; 2Johnson and Johnson Medical Asia-Pacific, Singapore OBJECTIVES: To conduct a systematic review of literature on the epidemiological and economic burden of surgical site infection (SSI) in China. METHODS: A literature search of the EMBASE and Medline databases was conducted. The search was limited to 1995–2010 to ensure the pertinence of the data. Relevant studies were identified using pre-defined inclusion criteria (i.e., reports the rate, risk factors, cost of SSI; conducted in a hospital setting; not an intervention study). Data on the prevalence, incidence, risk factors and cost of SSI were extracted. Searches to identify epidemiological and economic studies were conducted separately. RESULTS: The literature search identified six publications, comprising three studies conducted in Beijing (N = 95 to 196), one in Shandong (N = 2126), and two in Guangdong (N = 13,798). The incidence of SSI varied with surgical procedure. High incidence rates of SSI were reported for laryngectomy (21.1%), general surgery (18.8%), thoracic surgery (13.9%) and bone surgery (11.2%); while the rates for pancreas surgery (7.1%), neurosurgery (3%) and caesarean section (0.7%) appeared lower. The included studies found that risk factors such as wound classification, BMI and use of antibiotics significantly increased the risk of SSI. SSI was associated with a significant increase in neoplasm recurrence following laryngectomy (35% vs. 5.3%, P < 0.001), and extended postoperative hospital stay by an average of 33.6 days. The additional cost experienced by patients who developed a SSI was estimated at RMB 17,332/patient. CONCLUSIONS: SSI has the potential to represent a substantial burden on the healthcare system and patients in China, mainly attributable to the extended length of stay in hospital and additional cost of treatment required. Interventions aimed at reducing SSI would provide cost-savings to the health-care system and improve its efficiency. PIN2 SURGICAL SITE INFECTION IN INDIA: A SYSTEMATIC REVIEW OF THE INCIDENCE AND ECONOMIC BURDEN Tan JT1, Coleman K1, Norris S1, Mapari J2, Shastri S2, Metz L3 1 Health Technology Analysts Pty Ltd, Sydney, NSW, Australia; 2Johnson and Johnson Medical, Mahim, Mumbai, India; 3Johnson and Johnson Medical Asia-Pacific, Singapore OBJECTIVES: To conduct a systematic review of literature on the epidemiological and economic burden of surgical site infection (SSI) in India. METHODS: A literature search of the EMBASE and Medline databases was conducted. The search was limited to 1995–2010 to ensure the pertinence of the data. Searches to identify epidemiological and economic studies were conducted separately. Relevant studies were identified using pre-defined criteria (i.e., reports the rate, risk factors, cost of SSI; conducted in a hospital setting; not an intervention study). RESULTS: Twelve studies met the inclusion criteria and are presented in this review. The overall incidence of SSI in India is

4th Asia-Pacific Abstracts approximately 10%. The rates of SSI varied depending on the surveillance period, data collection method, and surgical procedure. The incidence of SSI was high for cardiovascular (5–18%) and gastrointestinal (6–16%) surgery. The rates of SSI appeared lower for neurosurgery (0.8–2.5%), although those studies only considered SSI cases before discharge from hospital. Diabetes and obesity were found to increase the risk of SSI by over 70%. Other significant risk factors identified include surgery duration and wound classification. The most common pathogens in SSIs were Staphylococcus aureus, Pseudomonas aeruginosa and Escherichia coli. SSIs were found to extend hospital stay by 5–18 days and increase treatment cost by 4–30%. The estimated cost of hospitalization was significantly higher in patients with SSI compared to patients without SSI (29,000 vs. 16,000 rupees, P < 0.001). CONCLUSIONS: In India, where an estimated 72% of health-care expense is out-of-pocket, the additional cost associated with SSI (treatment, loss of ability to work) represents a significant burden to patients and their families. The increase in hospital stay also lays additional burden to an already resource-constrained health-care system. Interventions aimed at reducing SSI would provide cost-savings and improve the efficiency of the health-care system. PIN3 SURGICAL SITE INFECTION IN AUSTRALIA: A SYSTEMATIC REVIEW OF THE INCIDENCE AND ECONOMIC BURDEN Coleman K1, Tan JT1, Norris S1, Hardy K2, Berg T2, Metz L3 1 Health Technology Analysts Pty Ltd, Sydney, NSW, Australia; 2Johnson and Johnson Medical Australia, Sydney, NSW, Australia; 3Johnson and Johnson Medical Asia-Pacific, Singapore OBJECTIVES: To conduct a systematic review of literature on the epidemiological and economic burden of surgical site infection (SSI) in Australia. METHODS: A literature search of the EMBASE and Medline databases was conducted. The search was limited to 1995–2010 to ensure the pertinence of the data. Searches to identify epidemiological and economic studies were conducted separately. Relevant studies were identified using pre-defined criteria (i.e., reports the rate, risk factors, cost of SSI; conducted in a hospital setting; not an intervention study). RESULTS: Thirty-five studies were included in this review. Differences in study design (surveillance period, data collection method, surgical procedure) made it difficult to synthesise data to derive a single estimate of SSI in Australia. The overall incidence of SSI in Australia is approximately 5–10%. However, the rate of SSI varied across different procedures: higher rates were seen following gastrointestinal (~11%) and cardiovascular (6–13%) surgery, while the rates for orthopedic (4.7–8%) and gynecological surgery (2.3–10%) appeared lower. Risk factors identified include diabetes and obesity, which increased the risk of SSI by 60–180%. The National Nosocomial Infections Surveillance (NNIS) risk index was positively correlated with the risk of developing an SSI. The most common organisms identified in SSIs were Staphylococcus aureus and Pseudomonas aeruginosa. The hospitalization cost attributable to SSI is estimated at approximately AUD$54 million annually. With a large proportion of SSIs occurring after discharge from hospital, the incidence of SSI is likely underestimated. Post-discharge SSI, in turn, increases the burden to community health services. Indirect costs, such as loss of productivity, further add to the economic burden of SSI. CONCLUSIONS: The incidence and cost estimates demonstrate that SSI represents a significant burden to the Australian health-care system. Interventions aimed at reducing SSI would provide cost-savings and improve the efficiency of the health-care system. PIN4 SURGICAL SITE INFECTION IN JAPAN: A SYSTEMATIC REVIEW OF THE INCIDENCE AND ECONOMIC BURDEN Tan JT1, Coleman K1, Norris S1, Maki A2, Metz L3 1 Health Technology Analysts Pty Ltd, Sydney, NSW, Australia; 2Johnson and Johnson Medical K.K., Chiyoda-ku, Tokyo, Japan; 3Johnson and Johnson Medical Asia-Pacific, Singapore OBJECTIVES: To conduct a systematic review of literature on the epidemiological and economic burden of surgical site infection (SSI) in Japan. METHODS: A literature search of the EMBASE and Medline databases was conducted. The search was limited to 1995–2010 to ensure the pertinence of the data. Searches to identify epidemiological and economic studies were conducted separately. Relevant studies were identified using pre-defined criteria (i.e., reports the rate, risk factors, cost of SSI; conducted in a hospital setting; not an intervention study). RESULTS: Thirty-five studies (22 retrospective, 12 prospective and one case-control study) were included. Differences in study design, surgical procedure and surveillance period made it difficult to synthesise data to derive a single estimate of SSI in Japan. The overall incidence of SSI in Japan is approximately 5–10%. The rate of SSI was higher following gastrointestinal surgery (~20%), and lower for cardiovascular (3 months was significantly associated with reduced odds of psychiatric hospitalization (OR = 0.46, 95% CI = 0.37–0.58), nonpsychiatric hospitalization (OR = 0.77, 95% CI = 0.64–0.92), and ER visit (OR = 0.69, 95% CI = 0.60–0.79). Treatment differences in health-care utilization were not statistically significant. These findings had no essential changes with adjustment for demographics, comorbid conditions, and prior health-care utilization and separate analyses for duloxetine and venlafaxine XR patients. CONCLUSIONS: Duloxetine therapy appears to have longer persistence than venlafaxine XR for high-cost patients with MDD. Medication persistence >3 months is associated with reduced hospitalization and ER visit. PMH14 ADHERENCE AND PERSISTENCE WITH DULOXETINE VERSUS ESCITALOPRAM AMONG HIGH HEALTH-CARE RESOURCES UTILIZERS WITH MAJOR DEPRESSIVE DISORDER Tepper P1, Liu X2, Able SL2, Johnstone B2 1 University of Pittsburgh, Pittsburgh, PA, USA; 2Eli Lilly and Company, Indianapolis, IN, USA OBJECTIVES: Adherence and persistence to prescribed medication is important in the management of major depressive disorder (MDD). This study compared adherence and persistence between duloxetine and escitalopram among high health-care utilizers with MDD. METHODS: In a large US commercial managed-care claims database, 10,803 MDD patients (18 to 64 years) who initiated on duloxetine (n = 4542) or escitalopram (n = 6261) during the calendar year 2006 were identified. All of the patients had no recorded prescription of study medications for 6 months prior to initiation and had continuous enrollment for 12 months prior to and post-medication initiation. High health-care utilizers (duloxetine, n = 3113; escitalopram, n = 3157) were defined as patients whose total treatment costs in the prior 1 year were above the median of total costs for all MDD patients in 2005. Adherence was defined as Medication Possession Ratio (MPR) ≥0.8, and persistence was defined as the length of therapy without exceeding a 30-day gap. Statistical analyses included chi-square test, Wilcoxon rank sum test, multivariate logistic and Cox regression. RESULTS: In the post 6 months (6M) and 12 months (12M), adherence rates were significantly higher for duloxetine-treated patients (6M 54.6% and 12M 40.9%) than escitalopram-treated patients (6M 47.4% and 12M 31.7%) (P-values < 0.001). Duloxetinetreated patients also stayed on the medication significantly longer (6M 115.9 days and 12M 171.1 days) than escitalopram-treated patients (6M 108.4 days and 12M 154.6 days) (P-values < 0.001). Results were essentially unchanged with adjustment for demographics, comorbid conditions, prior medications use, and health-care utilization, resulting in duloxetine therapy maintaining its better adherence and longer persistence versus escitalopram. CONCLUSIONS: Duloxetine-treated patients appear to be more adherent and have a longer stay on the medication than escitalopramtreated patients for high resources utilizers. Further research is needed to examine clinical and economic benefits of better adherence and persistence with duloxetine therapy. PMH15 THE VALIDATION OF TWO MEASURES ASSESSING REASONS FOR ANTIPSYCHOTIC DISCONTINUATION AND CONTINUATION FROM PATIENTS’ AND CLINICIANS’ PERSPECTIVES Faries D, Ascher-Svanum H, Nyhuis A, Anderson J, Phillips G Eli Lilly and Company, Indianapolis, IN, USA OBJECTIVES: Treatment discontinuation is an important effectiveness measure for evaluating antipsychotic treatment, yet is typically assessed only on a high level (e.g., “lack of efficacy,” “adverse event,” or “patient decision”). The RAD-I was developed as a structured interview assessing patient’s perspective regarding specific reasons for antipsychotic discontinuation and continuation and the RAD-Q as a questionnaire assessing clinician’s perspective. This study assessed the validity of the RAD measures in a randomized clinical trial. METHODS: The RAD was assessed at baseline (referring to medications prior to the trial) and endpoint (referring to discontinuation/ continuation of study drug) in a 12-week schizophrenia clinical trial. To assess its psychometric properties, we examined content validity, discriminate validity, and agreement between RAD scales using endpoint data (N = 158). RESULTS: Of 29 potential primary reasons for medication discontinuation, the most common was insufficient improvement/worsening of positive symptoms (24%) followed by adverse events (22%). A high level of concurrent validity was observed, as 89% of patients discontinuing medication due to “lack of efficacy” (87%, “adverse events”) per the standard trial form had lack of efficacy (“adverse events”) as a primary or very important reason on the RAD. Patients indicating lack of improvement/worsening of positive symptoms as a primary reason for medication discontinuation on the RAD had significantly less improvement in PANSS positive score than patients not reporting this as a reason. Similar results were observed for the RAD mood items (per MADRS) and functional status (per QLS) as well as for corresponding analyses assessing reasons for continuation. Agreement between the clinician and patient scores was high (range 73% to 100%). CONCLUSIONS: This initial psychometric assessment suggests the RAD is a valid tool for gathering detailed information regarding reasons for antipsy-

A556 chotic discontinuation and continuation from patients’ and clinicians’ perspectives. Further information is needed about the psychometric properties in usual care settings. PMH16 PHARMACIST-RUN METHADONE CLINIC IN A MALAYSIAN PUBLIC HEALTH CENTER: EVALUATING PATIENT SATISFACTION AND QUALITY OF LIFE OUTCOMES Chiew GP1, Shafie AA2, Hassali MA2, Awaisu A2, Cheah WK3 1 Larut Matang & Selama Health Office, Taiping, Perak, Malaysia; 2Universiti Sains Malaysia, Minden, Penang, Malaysia; 3Hospital Taiping, Taiping, Malaysia OBJECTIVES: To assess the satisfaction and health-related quality of life (HRQoL) improvement of patients enrolled in a pharmacist-run Methadone Maintenance Therapy (MMT) program. METHODS: A cohort study design was used to measure satisfaction and to evaluate changes in HRQoL of patients after one month of receiving methadone treatment at Taiping Health Clinic. Respondent’s satisfaction was measured by using an eight-item pre-validated questionnaire. A post-survey reliability analysis of the questionnaire showed a high internal consistency of the items (Cronbach’s α = 0.785). Meanwhile, the HRQoL was measured using a validated EQ-5D and EQ-VAS questionnaire that are administered by face-to-face interview in two phases, after 1 month interval. The data were analyzed by using both descriptive statistics (frequencies and percentages) and inferential statistics (χ2 test, paired t-test and the McNemar χ2 test). RESULTS: All 54 patients in MMT clinic participated, but only 40 (74.1%) completed this study. Average methadone dose in both phases were low (Phase One = 37.4 mg [SD = 22.2], Phase Two = 44.4 mg [SD = 21.3]) that caused majority of respondents wishing to increase their current dose. Respondents were not satisfied with needs to come clinic daily (n = 18, 33.4%) and did not believe that MMT clinic can help in cessation of drug abuse (n = 9, 16.7%). These two items were significantly associated with travelling distance of respondents to clinic (P = 0.001 and P = 0.039, respectively). Only pain/discomfort domain of the EQ-5D showed a significant improvement from the baseline (P = 0.035). EQ-VAS score significantly improved from 64.69 (SD = 16.7) at baseline to 71.43 (SD = 14.9) during the 1-month follow-up (P = 0.008). CONCLUSIONS: MMT program was able to improve patients’ QoL even in short duration of time. Improvement on dissatisfactions toward travelling distance, needs to travel daily to clinic and inadequate dose will help to increase treatment success. PMH17 COMPARISON OF HRQOL BETWEEN PATIENT RECEIVING METHADONE MAINTENANCE THERAPY (MMT) AND REHABILITATION PROGRAM Abdul Ghani N, Shafie AA, Hassali MA Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: This study aims to compare the HRQoL between patients receiving methadone maintenance therapy (MMT) program with those in the rehabilitation program. METHODS: This was a cross-sectional study involving 400 randomly selected patients from two primary and secondary hospitals and one rehabilitation center (PUSPEN) in northern part of Malaysia. Consented patient was interviewed to collect their socio-demography, drug consumption and quality of life information. The quality of life was measured by using EQ-5D and EQ-VAS questionnaires. Their quality of life scores were then compared using independent t-test. RESULTS: Majority of the participants were male with mean age 38.49 (SD = 9.6) in MMT group and 34.77 (SD = 8.6) in rehabilitation group. Mean duration of treatment for MMT group was 17.5 months (SD = 15.74) and rehabilitation group was 7.8 months (SD = 3.52). EQ-5D score was significantly (P = 0.01) higher among MMT participants (mean = 0.783, SD = 0.190) compared to those in rehabilitation program (mean = 0.707, SD = 0.227). Participants receiving MMT treatment also had lesser problems in mobility (10.5% vs. 21%; P < 0.01), normal activity (6.5% vs. 15%; P < 0.05) and anxiety (54.5% vs. 71%; P < 0.001). There was no statistically difference found between the groups in self care activity and pain dimensions. However, EQ-VAS score was significantly lower in MMT group compared to rehabilitation group (65.5 [SD = 17.9] vs. 73.5 [SD = 17.6], P < 0.001). CONCLUSIONS: This study shows that treatment for patient with substance use disorder with MMT program can provide a better quality of life compared to rehabilitation program (PUSPEN). Variation seen between EQ-5D and EQ-VAS scores suggest that patients might perceived their health worst than the general population value. PMH18 THE EFFECT OF DIAGNOSED, SELF-REPORTED, AND AT-RISK DEPRESSION ON HEALTH-RELATED QUALITY OF LIFE AND WORK PRODUCTIVITY IN JAPAN AND EUROPE DiBonaventura MD1, Fukuda T2, Nagae T3, Wagner JS1, Stankus A4 1 Kantar Health, New York, NY, USA; 2Tokyo University, Tokyo, Japan; 3York Pharma KK, Tokyo, Japan; 4Kantar Health, Princeton, NJ, USA OBJECTIVES: The aim of the current study was to establish the burden of depression (diagnosed, self-reported, and at-risk) in both Europe and Japan. METHODS: Data from the 2008 EU and 2008 Japan National Health and Wellness Survey (NHWS) were used. Patients were categorized into four groups: diagnosed depression, selfreported depression, at-risk for depression, and controls. Differences among these groups were examined on quality of life (mental (MCS) and physical component summary (PCS) scores of the SF-12v2), as well as overall work impairment, controlling

4th Asia-Pacific Abstracts for demographics and patient characteristics. RESULTS: In the EU, 5848 patients (10.9%) were diagnosed, 2037 (3.8%) were self-reported, 13,168 (24.6%) were at-risk, and 32,471 (60.7%) were controls. In Japan, 884 (4.4%) were diagnosed, 162 (0.8%) were self-reported, 5681 (28.4%) were at-risk, and 13,273 (66.4%) were controls. After controlling for demographics and patient characteristics, those with diagnosed depression (Adjusted Mean [Madj] = 34.4), self-reported depression (M—adj = 35.5), and at-risk depression (M—adj = 41.0) reported significantly lower levels of MCS scores than controls ((M—adj = 51.10, P’s < 0.0001) across all countries. The gap between controls and self-reported depression (b = 3.18, P < 0.0001) and at-risk depression (b = 0.63, P < 0.001) was significantly greater in Japan than in the EU. Both those with diagnosed (Madj = 46.2) and at-risk (M—adj = 49.4) depression reported significantly lower levels of PCS than controls (M—adj = 49.6, P’s < 0.05). Finally, those diagnosed with depression (M—adj = 39.2%), self-reported depression (M—adj = 30.9%), and at-risk for depression (M—adj = 23.3%) all reported significantly more overall work impairment than controls (M—adj = 12.8%). There was also a significant interaction, such that the difference in impairment between patients diagnosed with depression and controls was significantly greater in Japan (P < 0.05). CONCLUSIONS: Levels of diagnosed and self-reported depression were lower in Japan than in Europe, yet rates of at-risk depression were higher. Although the burden of depression was substantial, the results suggest that the work impairment burden in Japan is significantly greater than in the EU. PMH19 THE BURDEN OF DIAGNOSED, SELF-REPORTED, AND AT-RISK DEPRESSION IN CHINA AND EUROPE DiBonaventura MD1, Liu GG2, Wagner JS1, Stankus A3 1 Kantar Health, New York, NY, USA; 2Peking University, Beijing, China; 3Kantar Health, Princeton, NJ, USA OBJECTIVES: The objective of this project was to assess the burden of depression in terms of quality of life and work impairment in the EU and China. METHODS: Data were obtained from the 2008 EU and 2009 China National Health and Wellness Survey (NHWS). Patients were categorized into four groups: diagnosed depression, self-reported depression, at-risk for depression, and controls. Differences among these groups were examined on quality of life (mental (MCS) and physical component summary (PCS) scores of the SF-12v2), as well as overall work impairment, controlling for demographics and patient characteristics. RESULTS: In the EU, 5848 patients (10.9%) were diagnosed, 2037 (3.8%) were self-reported, 13168 (24.6%) were at-risk, and 32471 (60.7%) were controls. In China, 339 (2.5%) were diagnosed, 348 (2.6%) were self-reported, 3883 (29.2%) were at-risk, and 8752 (65.7%) were controls. Adjusting for demographics and comorbidities, those with diagnosed (b = −16.00, P < 0.0001) and self-reported (b = −14.42, P < 0.0001) depression, and those at-risk for depression (b = −9.22, P < 0.0001) reported significantly lower MCS than controls. Those diagnosed (b = −3.62, P < 0.0001) and at-risk (b = −0.47, P = 0.0022) for depression reported significantly lower PCS than controls. In addition, the MCS difference between those with diagnosed, self-reported, or at-risk depression relative to controls was significantly greater in Europe than in China (b = 5.88, P < 0.0001; b = 5.60, P < 0.0001; b = 3.49, P < 0.0001, respectively). However, the PCS gap between patients at-risk for depression and controls was significantly greater in China than in Europe (b = −0.99, P < 0.0001). Finally, the work impairment gap between patients with diagnosed, self-reported, or at-risk depression and controls was significantly greater in Europe than in China (b = −0.35, P = 0.0002; b = −0.40, P < 0.0001; b = −0.08, P = 0.0455, respectively). CONCLUSIONS: Depression was associated with a substantial burden on patients. The impact of depression on mental quality of life and work impairment was greatest in Europe while the impact of depression on physical quality of life was greatest in China.

MENTAL HEALTH – Health Care Use & Policy Studies PMH20 SWITCHING BETWEEN SSRI BRANDS IN AUSTRALIA 1 2 Ortiz MS , Calcino G 1 University of NSW, Darlinghurst, NSW, Australia; 2HI Connections, Canberra, ACT, Australia OBJECTIVES: To study the extent of brand substitution and switching between Selective Serotonin Reuptake Inhibitors (SSRI) available on the Pharmaceutical Benefit Scheme (PBS). METHODS: PBS prescription claims data provided by Medicare Australia of a 10% random sample of all Australian long-term concession card holders covering the time period August 2007 through July 2008 were assessed. Patients had to fill at least four prescriptions for an SSRI with generics over the 1-year period (fluoxetine [2 doses forms with up to 12 brands], fluvoxamine [2 strengths with up to 5 brands], paroxetine [1 strength with 11 brands] and sertraline [2 strengths with up to 13 brands]). The proportions of non-switchers (single brand only) and multiple switchers (two or more switches between brands) were determined for each strength of each SSRI molecule. RESULTS: A total of 18,691 Concessional patients filled at least four prescriptions for a SSRI in the 12-month time window. The majority of these patients received a single brand over the period: ranging from 49% for fluvoxamine to 67% for fluoxetine. Only a small proportion received three or more brands: ranging from 12% for fluoxetine to 16% for fluvoxamine. The proportions of multiple switchers varied slightly between molecules with: 22% for fluoxetine, 32% for fluvoxamine, 26% for paroxetine and 24% for sertraline. Switching was greater for

4th Asia-Pacific Abstracts products where more brands were available. The larger the brand premium, then the lower the market share of the premium brand. More multiple switchers were aged under 50 years (31%) than 70–79 years (20%). CONCLUSIONS: Multiple switches between SSRI brands occur in around one in four Australian Concessional patients and this has the potential to confuse some patients if not adequately monitored by prescribers and pharmacists. Switching rates increased where more brands are available on the PBS as well as in younger patients. PMH21 PHYSICIAN PRESCRIBING PATTERNS IN INNOVATIVE ANTIDEPRESSANTS IN THE UNITED STATES: THE CASE OF PATIENTS WITH MAJOR DEPRESSIVE DISORDER Lin HC, Balkrishnan R University of Michigan, Ann Arbor, MI, USA OBJECTIVES: The use of innovative antidepressants has been growing as many were introduced since 1990s. However, physician prescribing patterns in innovative antidepressants is unclear. This study disentangled the impacts of sociological factors, such as physician, patient, and health-care system factors, on the patterns and dynamics of adopting innovative antidepressants for major depressive disorder (MDD) in the US. METHODS: A retrospective 15-year cross-sectional study was conducted using the 1993–2007 National Ambulatory Medical Care Survey representing 125,605,444 MDD patients. The Heckman two-step model was used to capture physician’s two-stage prescribing decision and to correct sample selection bias. First, logistic regression was applied to examine the impact of sociological factors on whether an antidepressant is prescribed. Second, multinomial logistic regression was used to capture physician choices of innovative versus.older antidepressants. RESULTS: A total of 71.22% MDD patients were prescribed an antidepressant. Physicians who were primary care physicians, nonowner of health-care settings, practiced in Western and non-metropolitan areas, and had not seen patients before were more likely to prescribe innovative antidepressants (all P < 0.001. ) Patients who were younger, non-Hispanic, belong to HMOs, and took patient education were more likely to receive innovative antidepressants (all P < 0.001.) Trend analyses revealed that physicians who were owners of non-solo practice, practiced in Midwestern, Northeastern and metropolitan areas were decreasing likelihood of prescribing innovative antidepressants over time. Patients who were non-white, nonblack and non-Hispanic, and enrolled in Medicaid had increasing likelihood of receiving innovative antidepressants over time. CONCLUSIONS: The patterns of physician prescribing innovative antidepressants for MDD patients in the United States varied with sociological factors. Patient characteristics, physicians characteristics, and health-care system factors had impacts on physician’s decision on choosing innovative antidepressants. Prescribing patterns also varied over time. PMH22 HEALTH SCIENCE STUDENTS ATTITUDES TOWARD RESEARCH AND IT’S IMPLICATIONS TO TEACHING Mayya SS Manipal University, Manipal, Karnataka, India OBJECTIVES: To assess the attitudes toward research of health science students and discuss its implications to teaching. Attitudes about research are important because they may influence the quality of research. In particular, attitudes and beliefs about research can affect the extent to which they will develop useful critical thinking skills. Health science students usually tend to view research methods courses negatively. However, an understanding of these attitudes is necessary to help instructors facilitate the learning of research methodology, by enabling them to create more positive attitudes toward such courses. METHODS: The attitudes toward research of students of health sciences (n = 79), Manipal University, were measured using the instrument: the Attitudes toward research scale –ATR (Papanasiou, 2005). The sample of the study consisted of students enrolled for a course on “Biostatistics, Research Methodology and Epidemiology.” The ATR consist of 32 items constituting five subscales: research usefulness, research anxiety, positive attitudes, relevance to life, difficulty of research. Data was summarized by computing descriptive statistics. RESULTS: Students of health sciences scored more than 70% of the maximum score on the subscales -research usefulness, positive attitudes and relevance to life. They scored less than 50% of the maximum score on subscales -research anxiety and difficulty of research. Students expressed that estimating sample size, data analysis and interpretation are the most difficult aspects in research. CONCLUSIONS: To reduce the research anxiety and feeling of difficulty, topics relating to statistics are to be taught in a simpler way. They should be provided practical experience through project work. Usefulness of the topics is to be emphasised through actual data analysis and interpretation relating to their field.

MUSCULAR-SKELETAL DISORDERS – Clinical Outcomes Studies PMS1 THE LONG-TERM UTILIZATION AND SAFETY OF BIOLOGICAL AGENTS IN TREATING RHEUMATOID ARTHRITIS PATIENTS—A POPULATION-BASED CASE STUDY IN SOUTHERN TAIWAN Chang HC1, Chen LC1, Tseng HL2, Wu JS2 1 Kaohsiung Medical University, Kaohsiung, Taiwan, 2Bureau of National Health Insurance, Kaohsiung, Taiwan OBJECTIVES: This case study aims to evaluate the long-term utilization and safety of two biological agents- etanercept and adalimumab in treating Taiwanese rheuma-

A557 toid arthritis (RA) patients. METHODS: This cohort study used a population-based claim-data from the Kaopin Division of Taiwanese National Health Insurance during March 2006 to June 2009. Adult RA outpatients were identified by diagnosis code, and followed from the first prescription date (index date) of biological agents up to two months after discontinuation (prescription stop for more than 30 days) or the end of study. Prescription continuation rate between two drugs was evaluated using Kaplane-Meier survival curves and compared by the log-rank test. The frequencies of infectious events for new users of biological agents were also collected and the odds ratios of infectious events comparing etanercept against adalimumab were evaluated by logistic regression. RESULTS: Of all, 190 new users (mean age 53.43 ± 11.61 years; 83.68% female) were assessed. Most patients (70.53%) were prescribed biological agents in medical centers and 63.16% of them used etanercept. Total cumulative exposure time is 184.32 patient-years (145.86 patient-years in etanercept group, 38.46 patient-years in adalimumab group). Overall continuation rate for biological agents is 78.06% after 1 year, 55.04% after 2 years. The median continuation time is 836 days (95% CI 541, 972). Twenty-nine infectious events were detected; the most frequent events are pneumonia (seven events) and tuberculosis (six events). Patients who treated by etanercept had more infectious events than patients treated with adalimumab (OR: 5.99; 95% CI 1.55, 40.00). CONCLUSIONS: This study demonstrated that only half RA patients sustained their original biological therapy for more than 2 years. Furthermore, RA patients who treatd with etanercept acquired higher risk of infectious events. However, it is still necessary to further find out reasons for discontinuation and to explore the causality between biological agents and infectious events. PMS2 GASTROINTESTINAL AND CARDIOVASCULAR RISK OF NONSELECTIVE NSAIDS AND COX-2 INHIBITORS IN ELDERLY PATIENTS WITH KNEE OSTEOARTHRITIS Turajane T Police General Hospital, Bnagkok, Thailand OBJECTIVES: To evaluate the risk for gastrointestinal (GI) events and cardiovascular(CV) events in patients >60 years with knee osteoarthritis using tNSAIDs or coxibs users. METHODS: A hospital-based retrospective cohort study was applied. Data on prescription drug (NSAIDs, celecoxib, etoricoxib) was obtained from June 2004 to June 2007 were included if they were aged >60 years. Patients with a history of gastrointestinal disease or heart disease were excluded. RESULTS: A total 12,591 prescriptions from 1030 patients, an average of four prescriptions/patient/year, were screened—3982 (31.6%) prescriptions were for NSAIDs, 4426 (35.2%) were for celecoxib, and 4183 (33.2%) were for etoricoxib. The most common traditional NSAID prescribed was meloxicam (24%). The mean age of cohort was 69.6 years, with the majority being female (74%). Comparing celecoxib with NSAID use in logistic regression analysis, patients who received celecoxib were significantly less likely to suffer GI events than those who received NSAIDs; OR = 0.36 (95% CI 0.21–0.63, P = 0.00). Similarly, etoricoxib was less likely to cause GI events than NSAIDs; OR = 0.52 (95% CI 0.28–0.98, P = 0.04). Comparing to patients aged under 60 years, patients aged >70 years had a significantly higher chance of developing GI events, OR = 1.79 (95% CI 1.13–2.4) for patients aged 70–80 years and 3.36 (95% CI 1.78–5.81) for those aged >80 years. Drug exposure time, significantly increased the GI risks. For CV event, there were only three significantly associated with CV events-female (OR = 0.29, 95% CI 0.16–0.59, P = 0.00), age > 80 years (OR = 2.98, 95% CI 1.57–4.23, P = 0.00), and drug exposure time (OR = 1.05, 95% CI 1.02–1.54, P = 0.00). CONCLUSIONS: Incidence of GI and CV events was lower for coxibs than for NSAIDs and celecoxib had a lower incidence than etoricoxib. Patients with advanced age and higher drug exposure time had a significantly. PMS3 THE RISK OF REFRACTURE ASSOCIATED WITH THE COMPLIANCE AND THE PERSISTENCE WITH BISPHOSPHONATE THERAPY IN TAIWAN Su YJ1, Lin MJ1, Tsai IC1, Soong YK2 1 Novartis Pharmaceuticals Corporation, Taipei, Taiwan, Taiwan; 2Chang-Gung Memorial Hospital and The Taiwanese Osteoporosis Association, Taoyuan, Taiwan, Taiwan OBJECTIVES: To elucidate the relationship of the risk of refractures to compliance and persistence with bisphosphonate therapy in Taiwan. METHODS: We conducted a retrospective cohort study based on claims database of Taiwan’s National Health Insurance (NHI) between 2003–2006. The study subjects included all new users of bisphosphonates. Compliance was estimated using a medication possession ratio (MPR), calculated as the total days of supply divided by the number of days between the first and last dispensing plus the days of supply of the last dispensing. Persistence with bisphosphonate therapy was defined as continuous use, allowing for a refill gap of 30 days. RESULTS: The refracture rates of the osteoporosis patients increased with time. The refracture rate was 5.15%, 7.36%, and 8.49% at the first, second, and third years, respectively. The refracture rate of patients with over 80% compliance was significantly lower than those with a compliance below 80% (P < 0.05). The study found that nearly half of the patients were noncompliant with therapy (MPR < 80%) at as early as 3 months, and only around 30% of the patients were adherent at 1 year. The results also showed that the risk of refracture increased for patients with MPR < 80%, non-persistence, older patients and patients with comorbidities such as rheumatoid arthritis, diabetes mellitus or dementia. Patients with concomitant statin medication tended to have significantly lower refracture risks than those without. CONCLUSIONS: From the study, the compliance and persistence of Taiwanese

A558 patients is poor. In addition, the study demonstrated that the risk of refracture is associated with the compliance and persistence with bisphosphonate therapy in Taiwan. The compliance and persistence issues for osteoporosis treatment warrant much more attention.

MUSCULAR-SKELETAL DISORDERS – Cost Studies PMS4 BURDEN OF DISEASE IN PATIENTS WITH RHEUMATOID ARTHRITIS IN CHINA: RESULTS FROM 2009 NATIONAL HEALTH AND WELLNESS SURVEY Tang B1, Li Z2, Annunziata K3, Li X4, Dong P4 1 Pfizer, New York, NY, USA; 2Peking University People’s Hospital, Beijing, China; 3 KantarHealth, Princeton, NJ, USA; 4Pfizer China, Beijing, China OBJECTIVES: To assess comorbidity, quality of life (QOL), work/productivity loss, and medical resource utilization in patients with rheumatoid arthritis (RA). METHODS: Patients’ self-reported data were collected from 2009 National Health and Wellness Survey (NHWS). Survey samples represented major urban areas in China. QOL was measured by the physical component score (PCS) and mental component score (MCS) of the Short Form-12 (SF-12) (mean score of 50 for general population). Loss of work/productivity was measured by the validated Work Productivity and Activity Impairment (WPAI) instrument. Medical resource utilization was measured by health-care provider, emergency room (ER) visits and hospitalization in the past 6 months. Comparisons were made between respondents who suffer from RA (excluding psoriasis, psoriatic arthritis, or inflammatory bowel disease) versus respondents without RA (non-RA group). RESULTS: Of the 13,307 survey respondents, 276 (2.1%) were diagnosed with RA, and the average age was 46.6 years. There were more females in RA group (59.5%) versus non-RA group (50.7%). RA group reported more diagnosed comorbidities (insomnia 43%, hypertension 24.7%, migraine 24.1%, allergies 23.4%, arrhythmia 22.8%, gingivitis 22.5%, dermatitis 18.9%, osteoporosis 16.9%, anemia 15.5%, bronchitis 13.9%, asthma 13.7%), lower mean scores of PCS (42.7 ` 48.5) and MCS (42.4 vs. 46.0), more patients visited health-care providers (83.3% vs. 56.3%), ER (48.5% vs. 18.9%) and hospitalized (25.0% vs. 7.0%) in the past 6 months compared to non-RA group. Also, RA group reported 37.8% work/ productivity loss (absenteeism and presenteeism) and 36.6% impairment in daily activity compared to 22.7% and 22.6% in non-RA group. All comparisons between RA and non-RA groups were statistically significant at P < 0.05. CONCLUSIONS: From the China NHWS results, RA patients suffer from impairment in quality of life, work/productivity loss, more comorbidities and use of medical services. The findings indicate there is still an unmet medical need in RA patients in China. PMS5 MICRO-COSTING OF JUDO-THERAPY CLINICS IN JAPAN—MULTICENTERED COST ANALYSIS Tarumoto N1, Igarashi A2, Tarumoto E3, Harada T3, Fukuda T4 1 Teikyo Heisei University, Tokyo, Japan; 2Tokyo Univ. Faculty of Pharmacy, Tokyo, Japan; 3 Tarumoto Judo-Therapy Clinic, Tokyo, Japan; 4Tokyo University, Tokyo, Japan OBJECTIVES: In the Judo-therapeutic field, their costs are partly covered by National Health Insurance System via fee-for-service reimbursement, which does not necessarily reflect actual costs. Therefore, the actual situation of medical practice in Judo-therapy Clinics (bone-setting clinics) is analyzed by comparison between the actual costs calculated by micro-costing methods and .reimbursement cost calculated by fee-forservice system. METHODS: The basis of this study is a bidirectional evaluation of single medical intervention in terms of actual cost and reimbursed cost for single treatment with following methods: 1) All receipts are collected from the participating bone-setting clinics for a fixed period of time; 2) Along with collection of receipts, information on the medical interventions is recorded by region with regard to “how long” and “who” performed them; 3) The tentative cost of medical intervention is calculated based on the data in 2) above and the data of wage for the engaged persons. In other words, the tentative cost is calculated as the practitioner’s wage per unit time multiplied by the time consumed; 4) The cost-based amount of individual medical intervention is calculated by adding indirect costs (wage for other staff than practitioner, cost for operation/maintenance of judo-therapy clinic) to the data in 3); and 5) Standard critical paths are made for several diseases, based on which medical expense for each therapy model is calculated in terms of both receipt-based amount and costbased amount, and a comparison of amount is made between them. RESULTS: Therapies in a total of 1646 regions were recorded, based on which a comparison was made using four diseases (lumbar sprain, neck sprain, femoral contusion, back contusion) as models. The receipt-based medical expense was JPY9030 (US$100) for all diseases, while the cost-based medical expense widely varied from JPY22,057 to JPY22,950 (US$245 to US$255). CONCLUSIONS: There was a significant difference between the insurance benefit amount and the actual cost of medical intervention in judo-therapy.

4th Asia-Pacific Abstracts PMS6 HEALTH-CARE RESOURCE UTILIZATION AND OUTCOMES ANALYSIS OF OSTEOPOROTIC HIP FRACTURES AMONG THAI POSTMENOPAUSAL WOMEN: A PILOT INVESTIGATION Reungjarearnrung K1, Chonglerttham S2, Permsuwan U3 1 Bumrungrad Hospital, Bangkok, Bangkok, Thailand; 2Bumrungrad Hospital, Wattana, Bangkok, Thailand; 3Chiang Mai University, Muang, Chiang Mai, Thailand OBJECTIVES: Facing scarce data in Thailand on hospital costs of osteoporotic fractures, we examined repeated fractures and hospital costs incurred over 1 year from a primary osteoporotic hip fracture hospitalization occurred in postmenopausal women with or without comorbidities at Bumrungrad International Hospital in Bangkok. METHODS: Hospital admissions for primary hip fracture coded ICD-10 S 72.1 in post-menopausal women, aged 55 or older, were collected from January 2007 to January 2009. The hip fracture comorbidities considered were hypertension, type 2 diabetes, cardiovascular disease and their combinations. The repeated fractures within 6 or 12 months after first admission were documented. All direct medical costs were retrieved from the hospital database looking at 1-year follow-up period. Statistical analysis employed unpaired t-test, Mann–Whitney U test and multivariate analysis for comparison of outcomes, costs and correlations to comorbidities. RESULTS: A total of 59 hospital admissions for diagnosed osteoporotic primary hip fracture in postmenopausal women were screened. Patients’ mean age (SD) was 74.6 (7.7) years old, the mean hospital follow-up (SD) was 246.7 (77.5) days and the mean hospitalization duration (SD) was 17.0 (12.2) days. Patients at screening with and without comorbidities had similar characteristics including the types of fractures. The group with comorbidities was at higher relative risks of repeated fractures at 6 and 12 months follow-up compared to the group without comorbidities (1.13 vs. 0.91, P = 0.016; 1.10 vs. 0.94, P = 0.273, respectively) with consequently higher health-care utilization cost with 564,300 THB/patient/year versus 374,200 THB/patient/year, P = 0.043. The cost of all type drugs for both groups was only 8–10% of the total illness cost. CONCLUSIONS: In osteoporotic post-menopausal women admitted at hospital for primary hip fracture, comorbidities increased risks of repeated fractures hospitalization as well as hospital costs. The cost of drugs was marginal among hospital costs of osteoporosis. PMS7 A COST-EFFECTIVENESS ANALYSIS OF OSTEOPOROSIS TREATMENT FOR FRACTURE PREVENTION IN POSTMENOPAUSAL THAI WOMEN: A COMPARISON OF SEVEN TREATMENT OPTIONS Pongchaiyakul C1, Songpattanasilp T2, Taechakraichana N3 1 Khon Kaen University, Muang, Khon Kaen, Thailand; 2Phramongkutklao Hospital and College of Medicine, Ratchathewi, Bangkok, Thailand; 3Chulalongkorn University, Pathumwan, Bangkok, Thailand OBJECTIVES: To evaluate the cost-effectiveness of bisphosphonates (alendronate, risedronate, ibandronate, zoledronic acid), raloxifene, calcitonin and strontium ranelate, with a combination of calcium and Vitamin D as comparator, for the prevention of osteoporosis-related fractures in Thai post-menopausal women. METHODS: A Markov state transition model with 1-year cycle length was designed to simulate the cost-effectiveness of seven osteoporosis treatment interventions, compared with calcium and vitamin D, in Thai post-menopausal women aged 50–70 years. The model health states were categorized as osteopenia/no fracture, hip fracture, vertebral fracture, post-hip fracture, post-vertebral fracture and death. Treatment effects were obtained from published literature. The analysis was conducted using a societal perspective and included direct medical, direct nonmedical and indirect costs. Uncertainty was investigated by a probabilistic Monte Carlo simulation. Treatment outcomes were measured in terms of number of fractures avoided, number of life-years gained and quality-adjusted life-years (QALY) gained. Cost-effectiveness was defined as an ICER of less than 300,000 Baht (an incremental cost of ≤300,000 Baht for an outcome of no fracture for 1 year). RESULTS: 1) For patients with no prior fracture (primary prevention), zoledronic acid is cost-effective at ≥65 years, alendronate, risedronate and ibandronate at ≥70 years. 2) For patients with prior vertebral fracture (secondary prevention), zoledronic acid is cost-effective at ≥50 years, alendronate, risedronate and ibandronate at ≥55 years, raloxifene and strontium ranelate at ≥60 years; and 3) For patients with prior non-vertebral fractures, zoledronic acid is cost-effective at ≥60 years, alendronate at ≥65 years, and risedronate and ibandronate at ≥70 years. CONCLUSIONS: Zoledronic acid, followed by other bisphosphonates, is the most costeffective treatment option for both primary and secondary fracture prevention in Thai postmenopausal women with osteoporosis. These findings should be implemented in the government policy for selecting appropriate anti-osteoporotic drugs and reimbursement support strategy for Thai postmenopausal women with osteoporosis. PMS8 COST-EFFECTIVENESS OF RISEDRONATE FOR CORTICOSTEROIDINDUCED OSTEOPOROSIS IN AUSTRALIA Liew D1, Gonzalo F2, Lewis T3 1 The University of Melbourne, Fitzroy, Victoria, Australia; 2Sanofi-Aventis Australia Pty Limited, Macquarie Park, Australia; 3Sanofi-Aventis Australia Pty Ltd, Macquarie Park, NSW, Australia OBJECTIVES: We sought to determine, from the Australian health-care perspective, the cost-effectiveness of the bisphosphonate risedronate, relative to placebo, in preventing vertebral fractures among patients with corticosteroid-induced osteoporosis. METHODS: A decision-analytic Markov model was developed to simulate the onset

4th Asia-Pacific Abstracts of vertebral fractures and death among Australian patients with corticosteroid-induced osteoporosis. Age-specific underlying incidences of vertebral fractures and death were estimated from published data. The costs (AUD $1393 in Year 1 and AUD $34 in subsequent years) and utility (0.71) associated with a vertebral fracture were also drawn from published sources, as was the efficacy of risedronate in reducing vertebral fracture risk (relative risk 0.43, 95% CI 0.20–0.92). The model adopted a life-time horizon and an annual discount rate of 5% was applied to both costs and years of life/QALYs lived. Multiple sensitivity analyses were undertaken with reasonable variations to key data inputs. RESULTS: The base-case incremental cost-effectiveness ratio (ICER) with a life-time horizon was AUD $18,085 per QALY saved. Estimated ICERs remained below AUD $40,000 per QALY saved throughout all sensitivity analyses, except when assuming the upper limit of the 95% confidence interval for risedronate’s efficacy in reducing vertebral fracture risk (AUD $53,130 per QALY saved). CONCLUSIONS: Risedronate represents a cost-effective option for preventing vertebral fractures in Australian patients with corticosteroid-induced osteoporosis. This was confirmed by the recent recommendation by the Australian Pharmaceutical Benefits Advisory Committee (PBAC) that risedronate be reimbursed for this indication.

MUSCULAR-SKELETAL DISORDERS – Patient-Reported Outcomes Studies PMS9 HEALTH-RELATED QUALITY OF LIFE IN INDIAN PATIENTS WITH RHEUMATOID ARTHRITIS Khera K, Nagappa AN, Thunga G, Sam K Manipal College of Pharmaceutical Sciences, Manipal, India OBJECTIVES: Rheumatoid arthritis (RA) is a multisystem disease with various extraarticular manifestations (EAMs). Health-related quality of life (HRQOL) issues are assuming increasing importance in chronic rheumatic diseases like RA. Very less data on QOL in RA is available from the Indian subcontinent. There is also a paucity of literature on the impact of EAMs on HRQOL in RA. The objective of this study was to address these lacunae. METHODS: The study group comprised 166 patients with RA from a rheumatology clinic in India. Quality of life was estimated by the generic HRQOL measure: World Health Organization quality of life instrument (WHOQOLBref). Disease activity in RA was measured by calculating Disease Activity Score-28 (DAS28) and functional disability by the Health Assessment Questionnaire (HAQ). Extra-articular manifestations (EAMs) were diagnosed clinically. RESULTS: The mean duration of rheumatoid arthritis (RA) was 10 ± 3.1 years. The mean DAS28 and HAQ were 4.21 ± 1.1 and 1.01 ± 1.2, respectively. At least one ExRA was present in 46 patients. The mean HRQOL scores of the patients were 11.5 +/− 2.1, 12.2 +/− 2.5, 13.4 +/− 2.6, and 12.3 +/− 2.4 in the physical, psychological, social, and environmental domains of the WHOQOL-Bref respectively. Age, gender, disease duration, educational status, constitutional symptoms, rheumatoid factor positivity, erosions and deformities did not influence HRQOL. Disease activity had a negative influence on the physical and psychological domains. Patients with EAMs had significantly higher DAS28 scores compared to patients without EAMs. Even after adjustment for disease activity, patients with EAMs had lower HRQOL scores than patients without these features (statistically significant for physical domain). CONCLUSIONS: The physical domain of HRQOL is most affected in Indian patients with RA. Increasing disease activity and presence of EAMs worsens the quality of life.

MUSCULAR-SKELETAL DISORDERS – Health Care Use & Policy Studies PMS10 HEALTH-CARE UTILIZATION OF NEW PATIENTS WITH LOW BACK PAIN IN SOUTH KOREA Kim JH, Lee SM, Cheong C, Jang B, Son HJ National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea OBJECTIVES: The purpose of this study was to describe the pattern of health-care utilization among new patients with low back pain. METHODS: This study is based on the Korean National Health Insurance (NHI) claim data. Study subjects included those who 1) were newly diagnosed as low back pain; 2) received at least one lumbosacral spine simple x-ray; and 3) were not using medical service as a patient with low back pain for past 2 years before the first visit (index date). The follow-up period was 1 year from the index date. RESULTS: New patients with low back pain used the clinic (81.8%) most and orthopedics (62.2%) was the most visited department in 2006. Ninety-six percent of new patients used outpatient service and 5.9% used inpatient service. The medical cost (paid by insurers) per patient was $215.1. Male expenses more than female, and the cost increased as the age of patients increased. Average days including medication days and days of visit per new patient with low back pain was 6.7 days with the tendency that female’s average days including medication days and days of visit were more than those of male’s. In terms of outpatient service, average days including medication days and days of visit per new patients was 5.8 days and in inpatient service, it was 15.8 days. CONCLUSIONS: Efforts to identify the factors which may produce low back pain and to reduce the incidence is on need. Also, analysis on the factors influential over the health-care use of new patients with low back pain needs to be conducted in future studies.

A559 PMS11 HEALTH-CARE UTILIZATION OF PATIENTS WITH LOW BACK PAIN IN SOUTH KOREA (2006–2008) Kim JH, Lee SM, Cheong C, Jang B, Son HJ National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea OBJECTIVES: The objectives of this study were 1) to identify the pattern of healthcare use; and 2) to estimate the pattern change of health-care use for low back pain in each year of 2006, 2007, and 2008. METHODS: This study is based on the Korean National Health Insurance (NHI) claim data. Study subjects included those who 1) used medical service as a patient with low back pain from January 1 to December 31, and 2) received at least one lumbosacral spine simple x-ray each year. The identified number of the patients was 1,981,427 in 2006, 2,069,995 in 2007, and 2,119,408 in 2008. RESULTS: Patients with low back pain used the clinic most and orthopedics was the most visited department in each year of 2006, 2007, and 2008. The rate of outpatient service use declined and that of inpatient service use increased in compared 3 years. The medical cost (paid by insurers) showed increasing trend estimated as $569 million in 2006, $626 million in 2007, and $654 million in 2008. However, if the inflation rate of consumer price was considered (2.4% going up between 2006 and 2007, and 7.3% between 2006 and 2008), the cost in 2007 was estimated as $611 million still showing increase, but, in 2008, $610 million was estimated showing decrease. Medical cost per patient was $287.2 in 2006, $295.2 in 2007, and $287.8 in 2008. Average days including medication days and days of visit per patient was 9.4 days in 2006, 9.1 days in 2007, and 8.8 days in 2008 as declining. CONCLUSIONS: Efforts are needed to reduce prevalence and incidence of low back pain. And analysis on factors influential on the medical use of the patients needs to be conducted in future studies. PMS12 INCREASE OF INJECTION THERAPY ON PATIENTS WITH LOW BACK PAIN IN SOUTH KOREA (2006–2008) Kim JH, Lee SM, Cheong C, Jang B, Son HJ National Evidence-based Healthcare Collaborating Agency (NECA), Seoul, South Korea OBJECTIVES: The purposes of this study were 1) to analyze the pattern of injection therapy use for low back pain, and 2) to estimate the health-care use of low back pain patient who received injection therapy in 2006, 2007, and 2008. METHODS: This study is based on the Korean National Health Insurance (NHI) claim data. Study subjects included those who 1) used medical service as a patients with low back pain from January 1 to December 31; 2) received at least one lumbosacral spine simple x-ray; and 3) took more than one injection therapy each year. RESULTS: The percent of patients with low back pain who received injection therapy was increasing as followed: 10.8% in 2006 (213,796 among total 1,981,427 patients with low back pain), 11.0% in 2007 (227,432 among 2,069,995), and 11.5% (244,640 among 2,119,408). Number of patients with low back pain treated with injections per 100,000 general population was 436 in 2006, 462 in 2007, and 494 in 2008. Number of patients with low back pain treated with injections per 100,000 low back pain patients was 10,790 in 2006, 10,987 in 2007, and 11,543 in 2008. The rate of patients who received one injection in a year was 58.9% in 2006, 52.5% in 2007, and 48.1% in 2008 showing tendency of decrease, while more than two injections was increasing. The most conducted procedure was Epidural Nerve Blocks consistently for 3 years. The most department visited for service was anesthesiology in 2006 and orthopedics in 2007 and 2008. CONCLUSIONS: The injection therapy on patients with low back pain is increasing. But the efficacy of it has not reached to consensus so far, which means more careful conducting is needed. Also, the efficacy of injection therapy specifically in South Korea should be investigated in future studies. PMS14 FACTORS INFLUENCED ORAL OXICAM AND COXIBS’ EXPENDITURE AT A THAI TEACHING HOSPITAL, FISCAL YEAR 2006–2009 Kaojarern S1, Masaya-anon N1, Pongcharoensuk P2, Pattanaprateep O1 1 Ramathibodi Hospital, Mahidol University, Bangkok, Thailand; 2Mahidol University, Bangkok, Thailand OBJECTIVES: Oral Oxicam and Coxibs’ expenditure at a Thai teaching hospital was rapidly increased in recent years. We examined price and quantity factors that influenced the rising expenditure by deterministic model. METHODS: This study was a retrospective database analysis at a Thai teaching hospital in Bangkok. Prescription utilization data of one oral Oxicam and two Coxibs (Meloxicam, Celecoxib, and Etoricoxib) during fiscal year 2006 to 2009 were retrieved. In the model, changes in annual expenditure were attributed to two factors: cost per patient per year (P) and number of patients per year (Q). By measuring quantity as Defined Daily Dose (DDD), changes in cost per patient (P) comprised of two parts: cost per DDD (p) and DDD per patient (q). RESULTS: Overall expenditure of the three drugs was risen from 49.9 million Baht in 2006 to 64.9 million Baht in 2009 (adjusted for inflation). Thirty percent of total increase in expenditure was mainly the result of switching from Meloxicam (percent change, −44%) to Etoricoxib (119%). When decomposed into price and quantity factors, quantity had more effect than price, especially number of patients who moved from Meloxicam (percent change of Q, −35%) to Etoricoxib (112%) and Celecoxib (8%). DDD per patient for Celecoxib was a little increase (percent change of q, 4%), slightly decreased for Etoricoxib (−4%), and much lower for Meloxicam (−32%). Percent change in price of the three drugs was slightly positive (28%, 7%, and 8% for Meloxicam, Celecoxib, and Etoricoxib, respectively). CONCLUSIONS: The major increase of drug expenditure from 2006 to 2009 was from

A560 patients switching from older drug, Oxicam to new drugs, Coxibs. Deterministic model will help the hospital administrator understand pattern of drug use and its effect of financial management; however, clinical outcome should be included for further analysis.

NEUROLOGICAL DISORDERS – Clinical Outcomes Studies PND1 OUTCOME OF PHARMACIST INTERVENTION IN EDUCATION OF PATIENTS ON DUPLICATE PRESCRIPING HYPNOTIC-SEDATIVES Chu LL1, Lin HC1, Huang HY1, Chan AL2 1 Chimei Medical Center, Tainan, Taiwan; 2Chi Mei Medical Center, Tainan, Taiwan OBJECTIVES: The inappropriate uses of hypnotic-sedatives will likely cause drug abuse. It has been known that patients visit different hospitals back and forth to get a duplicate prescription of hypnotic-sedatives. The purpose of this study is to explore the decrease of refill frequency of prescribing sedative-hypnotics and drug cost. METHODS: This is a cohort study. Adult patients, who were prescribed two or more than two sedative-hypnotics with same pharmacological mechanism over a week by different physicians, will be retrieved from the claim database of a medical center in southern Taiwan. The retrieved patients were received a questionnaire to evaluate their knowledge about the appropriate use of hypnotic-sedatives before clinical pharmacists provided patients with education for appropriate use of hypnotic-sedatives. The primary outcome was the reduction of total consumption on hypnotic-sedatives and drug cost before and after pharmacist education. RESULTS: A total of 100 patients were included in this study. The results showed that about 62.8% of patients with duplicate use of hypnotic-sedatives were female and 59.1% of patients aged ³50 years. Their chief complains was insomnia and/or anxiety which cannot be relieved by taking only one hypnotic-sedatives. The most frequently prescribed drugs in their duplicate prescription were zolpidem (30.45%), Fludiazepam (22.34%), Alprazolam (14.9%), Flunitrazepam (8.45%), and Brotizolam (7.93%). The average total consumption on hypnotic-sedatives per day for each patient was decreased from 1.68 to 1.33 tablets. The monthly hypnotic-sedatives costs were reduced form NT$ 21,000 to NT $ 15,600. CONCLUSIONS: Pharmacist education on the appropriate use of hypnotic-sedatives and lifestyle management is likely to reduce the total consumption on hypnotic-sedatives and save drug costs. PND2 WARFARIN-ASSOCIATED INTRACRANIAL HEMORRHAGE Yang SC, Lin HL, Chan AL Chi Mei Medical Center, Tainan, Taiwan OBJECTIVES: To explore the possibility of international normalized ratio (INR),drug—warfarin interactions exaggerated intracerebral hemorrhage. METHODS: This is an observational study. Patients who had diagnoses of intracranial hemorrhage (ICH,ICD-9-CM codes 430, 431, 432.0, 432.1, 432.9) from January 2004 to December 2009 were retrieved from the claim database of a medical center in southern Taiwan. We reviewed medical records to identify the possible relationship with suspected risk factors, drug-warfarin interactions and ICH. Patients who were not taking warfarin at the time of hemorrhage and traumatic ICH were excluded. Potential drug interactions were defined according to online drug interaction database of Lexi-InteractTM. Data were analyzed by multivariate logistic modeling. RESULTS: During the study period, 24 patients were defined as ICH while patients were receiving warfarin. Forty-five percent of patients had an INR value >3.0 at admission to hospital and 25% were exceeded 5.0. The potential drug-warfarin interaction were antiplatelets (41.6%), allopurinol (16.6%), amiodarone (8.3%), NSAID (16.6%), others (16.6%).which may increased the bleeding risk. Five patients of 24 patients were died within 2 weeks. No correlation between the fatal cases and their gender, age, INR value, smoking, alcohol and comorbid diseases was founded. CONCLUSIONS: Although no correlation between drug–warfarin interaction, INR and ICH was found in this study, physicians should still need to inform patients, who are taking warfarin, to concern the potential interactions between herbs and over-the-counter medicines.

NEUROLOGICAL DISORDERS – Cost Studies PND3 A RAPID REVIEW OF ECONOMIC EVALAUTIONS OF DONEPEZIL, RIVASTIGMINE, GALANTAMINE AND MEMANTINE FOR ALZHEIMER’S DISEASE (AD) Feng J, Huang R, Pwee K Ministry of Health, Singapore OBJECTIVES: To review the cost-effectiveness of cholinesterase inhibitors (ChEIs; donepezil, galantamine and rivastigmine) and memantine for treating patients with AD. This review aims to inform policymakers in Singapore on the published costeffectiveness studies of these drug therapies. METHODS: Electronic databases (EMBASE, MEDLINE, NHS EED) were searched from inception to February 2010 for published economic evaluation studies. Additional references were identified through searching bibliographies of included studies and related publications. Review papers of economic evaluations either on individual drugs or more of these drugs were considered. An outline assessment of economic evaluations was undertaken using

4th Asia-Pacific Abstracts Drummond’s checklist. RESULTS: A systematic search of the literature identified 32 economic evaluations (2 of all three ChEIs, 11 of donepezil, 5 of rivastigmine, 7 of galantamine, and 7 of memantine). All studies used clinical trial data from single drug comparisons with placebo. Different approaches have been used to model disease progression for evaluating cost-effectiveness of each drug, making comparison between drugs more difficult. Caro et al. (2003) and Green et al. (2005) considered the three ChEIs using a common analytical framework. Most of the studies reported that drugs provided health benefits and cost savings over time, suggesting that their use was cost-effective. However, Green et al. (2005) concluded that treatment with ChEIs may not be cost effective from the UK NHS’s perspective. CONCLUSIONS: The majority of economic evaluations of ChEIs and memantine reported that these treatments were the dominant strategies compared with no treatment. However, conclusions from the reviews of economic analyses were mixed, and the authors commented that the findings from economic analyses of these four drugs should be interpreted with caution due to the assumptions and methodology used. Comparison between studies is difficult due to difference in study types, geographic locations, perspectives and different assumptions within the models.

RESPIRATORY-RELATED DISORDERS – Clinical Outcomes Studies PRS1 RISK OF SERIOUS ASTHMA EXACERBATIONS ASSOCIATED WITH LONG-ACTING-BETA-AGONISTS AMONG MEDICAID PATIENTS WITH ASTHMA: A RETROSPECTIVE COHORT STUDY Guo JJ1, Tsai K2, Kelton CM3, Bian B1, Wigle P1 1 University of Cincinnati, Cincinnati, OH, USA; 2Novartis Pharmaceutical Company, East Hanover,, NJ, USA; 3University of Cincinnati College of Business, Cincinnati, OH, USA OBJECTIVES: Concerns have been raised that the use of long-acting-beta-agonists (LABAs) may be associated with an increased risk of serious exacerbations (SAEs). We proposed to examine the association between LABAs and SAEs among patients with asthma in a large Medicaid population. METHODS: A total of 1,117,850 patients with asthma were selected for this study using a multi-state Medicaid claims database. The study period was January 1, 2002 to December 31, 2007. SAEs included any primary asthma diagnosis for either hospitalization, emergency room visits, or intubation. Major asthma medications included in the analysis were inhaled corticosteroids (ICS), LABAs, ICS/LABA combinations, short-acting-beta-agonists, leukotriene modifiers, and methylxanthines. Asthma severity levels were defined by patient’s treatment regimens based on the national asthma guideline. Cox proportional hazard regression was conducted to assess the risk of SAEs associated with the use of LABAs, ICS/LABA combinations, and other covariates. RESULTS: Based on the first 6 months of wash-out period, we identified 550,392 patients with newly diagnosed and 567,458 patients with preexisting asthma. A total of 197,283 (35.8%) newly diagnosed and 242,207 (42.7%) preexisting patients had at least one SAE during the study period. After controlling for covariates, the risk of SAEs was relatively low among newly diagnosed patients taking salmeterol (hazard ratio [HR] = 0.93, 95% confidence intervals [CI] 0.89–0.97), formoterol (HR = 0.88, 0.83–0.94), and ICS/LABA combination (HR = 0.88, 0.87–0.90). Meanwhile, the risk of SAEs was mixed among preexisting asthmatic patients taking salmeterol (HR = 1.03, 1.01–1.04), formoterol (HR = 0.94, 0.90–0.97), and ICS/LABA combination (HR = 1.01, 1.00–1.02). Other key risk factors (P < 0.0001) were associated with African-American, severe or moderate asthma, use of oral steroids, alcohol or substance disorder, depressive disorder, pregnant women, obesity, and upper respiratory tract infection. CONCLUSIONS: The risk of SAEs should be considered when treating patients with asthma, especially for patients with key comorbidities as well as African-American patients and pregnant women. PRS4 ASSESSMENT OF KNOWLEDGE ABOUT TUBERCULOSIS AMONG TB PATIENTS IN NORTH EAST LIBYA Solliman M1, Ahmad Hassali MA1, Al-Haddad M1, Hadidan M2 1 Universiti Sains Malaysia, Penang, Penang, Malaysia, 2Alfatih Medical Sciences University, Tripoli, Libya OBJECTIVES: To evaluate the knowledge toward the etiology and treatment of Tuberculosis (TB) among TB patients in North Eastern Libya. METHODS: Face to face interview with all TB Patients in 2009 was conducted at Kuwaifia and Shahat Chest Hospitals in North Eastern Libya. Questionnaire included questions related to the knowledge about transmission, diagnosis; risk factors, treatment and prevention of TB. A total of 140 patients participated in this study. Knowledge scores ranged from zero (minimum score) to 23 (maximum score). All data was analyzed using SPSS version 15.00®. Inferential statistics were used as whenever appropriate. P-value of less than 0.05 was considered significant. RESULTS: Majority of the respondents were Libyans (n = 92, 65.7%). Fifty-five (39.3%) of the respondents were males, and up to 90 subjects (64.3%) were urban residents. Almost all of the respondents have heard about TB. Nearly (n = 70, 50.0%) shared a thought that the main source of their knowledge regarding TB were health workers, followed by television (n = 63, 45.0%), and family members (n = 34, 24.3%). Libyans have significantly better knowledge (12.26 ± 4.33) than non Libyans (7.85 ± 4.83, t = 30.12, P < 0.001). In addition those respondents with tertiary educations scored significantly higher knowledge score (15.7 ± 4.2) compared to those of intermediate (15.3 ± 3.8) and illiterate (9.4 ± 3.5), [F = 30.12, P =

4th Asia-Pacific Abstracts 0.001]. Moreover, respondents from urban areas were found to have better knowledge (12.06 ± 3.82) than rural residents (8.38 ± 4.92, t = 20.0), (P = < 0.001). While, in term of level of income those with high income (12.37 ± 3.79) have significantly better knowledge than low income (8.00 ± 4.85), [F = 5.5, P < 0.001]. CONCLUSIONS: The level of knowledge about TB among TB patients was a low, Therefore, there is a need for massive health education campaign to be undertaken by policymakers in order to improve TB patients’ knowledge toward TB. PRS5 EVALUATING SAFETY OUTCOMES OF LONG-ACTING BETA AGONISTS (LABAS) IN PATIENTS WITH ASTHMA Bian B1, Guo JJ1, Kelton CM2, Wigle P1 1 University of Cincinnati, Cincinnati, OH, USA, 2University of Cincinnati College of Business, Cincinnati, OH, USA OBJECTIVES: The use of long-acting beta agonists (LABAs) has raised safety concerns, especially the potential for severe asthma exacerbations (SAEs). The purpose of this review was to review the published safety outcomes researches about LABA therapy and compare disease epidemiologic trends. METHODS: A systematic literature review was conducted with the specific focus on safety outcomes research from clinical trials, metaanalyses, and post-marketing studies. Patients with asthma were reviewed and compared between races, and between age groups. SAEs were defined as either asthma-related death, hospitalization or emergency room visits. Seven major studies were identified and reviewed. Annual mortality rate and SAE trends for asthmatic patients and subgroup of African American patients were compared. RESULTS: The prevalence of asthma increased from 5% in 1992 to 7.8% in 2006, while mortality rate decreased from 1.7 per 100,000 in 1999 to 1.2 per 100,000 in 2006. Multiple clinical trial data showed that LABA monotherapy may be associated with increased risk of asthma exacerbation or asthma related death. On the other hand, meta-analyses showed that the use of LABA with ICS may be associated with positive outcomes including symptoms reduction, quality of life improvement, reduced rate of exacerbations and severe exacerbation. Due to the limited data, the relationship between LABAs and an increased risk of severe asthma exacerbation is unclear, especially in pediatric and African American patients. Due to safety concerns, the US FDA made several label changes to warn the public about LABA therapy and has requested more assessments of LABA use in asthma patients. CONCLUSIONS: The safety outcomes of the use of LABAs for the asthma treatment remain controversial. More well-design research is required to assess the risk of SAEs related to LABA therapy in different populations.

RESPIRATORY-RELATED DISORDERS – Cost Studies PRS6 BURDEN OF SMOKING RELATED MAJOR DISEASES IN INDIA: ESTIMATION OF DIRECT OUT-OF-POCKET COST Kochhar P1, Arora M2, Menon S1, Selvaraj S2 1 Pfizer India, Mumbai, India, 2PHFI, New Delhi, India OBJECTIVES: To estimate the direct out-of-pocket medical costs of treating major diseases attributable to smoking in India. METHODS: Indian data on burden of smoking related illnesses was compiled on lines of the BENESCO (Benefits of smoking cessation on outcomes) model by secondary analysis of existing datasets from census and National Sample Survey (NSS), 2004. The prevalence of smoking(cigarettes and bidis) was estimated by adjusting census projected population data with smoking rate found in the National Family & Health Survey–3 (NFHS-3), 2005–2006. Data on age and gender-wise disease prevalence and unit cost of treatment of diseases was obtained from NSS. A prevalence-based, disease-specific approach was used to estimate the direct medical costs of treating cancer and other tumors, bronchial asthma, tuberculosis, respiratory diseases, hypertension and coronary heart disease attributable to smoking. RESULTS: Analysis of data from NFHS–3 estimated that there were 98.2 million male and 4 million female smokers in India between 15 to 49 years of age. A total of 1.1 million hospitalizations and 90.5 million outpatient visits for 6 major diseases among adults were attributed to smoking annually. The total annual direct out-of-pocket medical cost for treating 6 major diseases was found to be Indian National Rupees 198 billion (US$ 4.4b). The total cost attributable to smoking was estimated to be INR 46 billion (US$ 1b) (i.e., 23% of the total); INR 33.8 billion (US$ 761m) for outpatient care and INR 12.3 billion (US$ 276m) for inpatient care. The total annual smoking attributable cost of treatment was highest for heart disease at INR 14.5 billion (US$ 327m) followed by bronchial asthma INR 8.8 billion (US$ 199m) and hypertension INR 8.6 billion (US$ 195m). CONCLUSIONS: The smoking epidemic in India is a major preventable public health problem. The direct out-ofpocket medical costs of treating major diseases attributable to smoking are high. PRS8 THE EFFECT OF UNCONTROLLED ASTHMA ON HEALTH-RELATED QUALITY OF LIFE AND RESOURCE USE IN JAPAN AND THE UNITED STATES DiBonaventura MD1, Arakawa I2, Fukuda T3, Nagae T4, Wagner JS1, Stankus A5 1 Kantar Health, New York, NY, USA; 2GlaxoSmithKline K.K., Tokyo, Japan; 3Tokyo University, Tokyo, Japan 4York Pharma KK, Tokyo, Japan; 5Kantar Health, Princeton, NJ, USA OBJECTIVES: The aim of the current study was to determine the burden of illness associated with uncontrolled asthma in both the United States and Japan. METHODS: Data were obtained from the 2009 US and 2008 Japan National Health and Wellness

A561 Surveys (NHWS), cross-sectional Internet-based surveys of adults. Only patients who were diagnosed with asthma were included in the analyses. Using the Asthma Control Test (ACT), each asthma patient was categorized as having controlled or uncontrolled asthma and these groups were compared on health-related quality of life (mental component summary (MCS) and physical component summary (PCS) scores of the SF-12v2), the number of emergency room (ER) visits, and the number of hospitalizations controlling for demographics (country, age, gender, ethnicity, income, education) and patient characteristics (BMI and Charlson comorbidity index) in multiple regression and generalized linear models. RESULTS: In Japan, 648 patients (3.2%) were diagnosed with asthma while 6481 patients (8.5%) were diagnosed with asthma in the United States. Even after adjusting for demographics and patient characteristics, those with uncontrolled asthma reported significantly lower levels of MCS scores (M-adj = 42.5 vs. 45.9, P < 0.0001). However, not only were PCS scores significantly lower among uncontrolled asthma patients (M-adj = 41.7 vs. 47.4, P < 0.0001), but the differences between controlled and uncontrolled patients was significantly greater in the United States (b = 4.4, P < 0.0001). The number of hospitalizations (M-adj = 0.61 vs. 0.23, P < 0.0001) and ER visits (M-adj = 0.51 vs. 0.24, P < 0.0001) were both significantly higher for those with uncontrolled asthma. The effect on ER visits was significantly stronger in the United States than in Japan (b = 8.2, P < 0.01). CONCLUSIONS: Although the prevalence of asthma in the United States and uncontrolled asthma was greater than that of Japan, uncontrolled asthma patients in both countries exhibited a significantly worse health outcome profile. PRS9 DRUG UTILIZATION EVALUATION OF COMBINED INHALER THERAPIES AMONG PATIENTS WITH AIRWAY DISEASES IN A 2000-BED MEDICAL CENTER IN TAIWAN Lin HW1, Yu IW2, Cheng WE2 1 China Medical University, Taichung, Taiwan; 2China Medical University Hospital, Taichung, Taiwan OBJECTIVES: Appropriate use of combined inhalation therapies for airway disease patients is likely to reduce the undesirable health utilization, improve patients’ healthrelated quality of life, smooth the decline of pulmonary function and prevent the mortality. A secondary data analysis was performed to describe the utilization of two types of combined inhaler therapies among the patients diagnosed with airway diseases for quality improvement of medical care. METHODS: The claimed administration data with visits using two types of combined inhalers (four various dosages and dosage forms of Seretide and one Symbicort turbulaler) and the corresponding visits’ diseases and demographic information from the 2000-bed medical center affiliated to a medical university in Taiwan were retrieved and analyzed. There were up to five diagnoses listed in the data set. The descriptive analysis was performed to examine the trends of different types of combined inhalers being prescribed for patients diagnosed with corresponding airway diseases (i.e., ICD9: 491, 492,493,494,496, 518) in the year of 2009. RESULTS: Of 2495 patients being prescribed with the combined inhaler therapies (accounted for 9273 total visits), male was dominated (57.7% vs. 42.3%), average age was 56.7 ( ± 22.1) year-old and more than 80% were prescribed by the pulmonary physicians. Less than 2% had experience of switching inhalers within 1 year. Only 55.4% were prescribed with coded airway diseases (asthma and COPD accounted for 25.2% and 21.96%, respectively). Further, 50.6% of them were prescribed with inhalers once or twice within 1 year. These prescribing patterns might encounter the “drug choice problems.” CONCLUSIONS: The prescribing patters of combined inhaler therapies seem problematic in this health setting and need to be examined persistently and periodically. To ensure appropriate usages of these inhalers, it is necessary to further study the contributing factors and to investigate the possible resolutions to improve the quality of medical care.

RESPIRATORY-RELATED DISORDERS – Health Care Use & Policy Studies PRS10 RACE AND ETHNICITY IN EMERGENCY ROOM AND NON-URGENT CARE VISITS AMONG LATINO CHILDREN SUFFERING FROM ASTHMA Chang J1, Patel I1, Ortega A2, Park YS1, Balkrishnan R1 1 University of Michigan, Ann Arbor, MI, USA; 2University of California, Los Angeles, CA, USA OBJECTIVES: To identify differences in health-care utilization by race and Latino ethnicity among children with asthma and determine the relative contribution of predisposing, enabling, and unobserved factors in explaining these differences. METHODS: The California Health Interview Survey (CHIS) is a non-institutionalized household survey of California residents conducted bi-annually since 2001. Crosssectional survey data were a0nalyzed to identify differences by race and Latino ethnicity among children with asthma. After adjustment for sample weights and design effects, over 400,000 samples in the 2007 CHIS were calculated for children with asthma. Disparities in health-care utilization across all race/ethnicity categories were analyzed using multivariate logistic regression and negative binomial regression analysis. RESULTS: Mexican children with asthma are over 100% more likely to have an emergency room (ER) visit in the previous year (P < 0.04) and 58% less likely to have a doctor visit (P < 0.049) compared to White children with asthma after controlling for confounding variables. Among ethnicity groups, Latino ethnicity is 72% more likely to have an ER visit in the previous year and 38% less likely to have a doctor visit compared to non-Latino ethnicity. Negative binomial regression results show that

A562 Mexican race is negatively associated with number of doctor visits (P < 0.024). CONCLUSIONS: The CHIS database offers an opportunity to investigate the occurrence of ER and doctor visits among children with asthma by race and ethnicity. This study shows that Mexican children are more likely to have an ER visit and less likely to have a doctor visit compared to White children. PRS11 A CASE STUDY EVALUATING MALAYSIAN COMMUNITY PHARMACISTS’ PROFESSIONAL PRACTICES AND SERVICES Alamin A1, Izham M2, Hassali MAA1 1 Universiti Sains Malaysia, Penang, Malaysia; 2Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia OBJECTIVES: To evaluate the professional practices and services of the Malaysian community pharmacists, using a ‘simulated client’ interview method. The interview was carried at two community pharmacists in the state of Penang. METHODS: The simulated client (PhD student at the School of Pharmacy, Universiti Sains Malaysia) visited two community pharmacists “Pharmacy A” and “Pharmacy B” and asked for medications for his and his son’s “viral fever symptoms” respectively. The details were documented in a form developed by the researchers. RESULTS: The findings suggested a poor professional behavior and dispensing practices among both the pharmacies studied. Although, both of them asked the duration of the symptoms, they did not ask the age of the patient (s), presence of blood in sputum, previous history of presenting medical complaints, past medical history, drug history and smoking status. The nature of the sputum was asked only by the pharmacists at pharmacy B. The pharmacy A dispensed Tab. Clarithromycin 500 mg, Tab. Paracetamol 500 mg, and Azithromycin powder for suspension 200 mg/5 ml for his son’s viral fever symptoms. The pharmacy B dispensed Tab. Cefuroxime 250 mg, Amoxicillin + clavulanic acid powder for suspension 125 mg + 31.25 mg/5 ml, Linctus Dextromethorphan + Pro-methacine + Pseudo-ephedrine + Paraben sebagai pengawet, and Tab. Green oats + Sea Buckthom + Stinging Nettle + Tongat Ali + Damiana. The cost of prescription was RM 56.50 (US$ 16.14) for pharmacy A and RM 129 (US$ 45.15) at pharmacy B. CONCLUSIONS: Our findings suggested the need for continuing pharmacy education (CPE) programs for the community pharmacists in the country and need for curriculum changes incorporating more topics related to community pharmacy practice. Though our survey included only two community pharmacies, it can serve as a pilot study for future in depth survey encompassing various regions in Malaysia.

SENSORY SYSTEMS DISORDERS – Clinical Outcomes Studies PSS1 EPIDEMIOLOGY AND COMORBIDITIES OF PATIENTS WITH PSORIASIS N TAIWAN Tsai TF1, Tang CH2, Hung ST2, Tsai IC3, Schenkel B4, Zhang M4 1 National Taiwan University Hospital, Taipei, Taiwan, Taiwan; 2Taipei Medical University, Taipei, Taiwan; 3Janssen Cilag Taiwan, Taipei, Taiwan; 4Johnson & Johnson Pharmaceutical Services, LLC, Horsham, PA, USA OBJECTIVES: Psoriasis is a chronic immune-mediated inflammatory disorder and is associated with concurrence of other complex disorders. This study aims to describe the epidemiology of psoriasis and the prevalence of comorbidities in patients with psoriasis in Taiwan. METHODS: Patients who had at least one outpatient visit or admission diagnosed with ICD-9-CM code 696.0–1 in the Taiwan National Health Insurance (NHI) claims database during 2006 were identified as psoriasis cases. The cases were further classified into moderate to severe psoriasis (sPsO) for those who had previously received systemic therapy and mild psoriasis (mPsO) for those who had not. The cases were matched with controls in a 1:4 ratio from one million beneficiaries in the NHI dataset representing the Taiwan population. Matching variables included age, gender and residential area. Prevalence of comorbidities was assessed using prevalence relative risk (RR) based upon a Cox proportional regression model. RESULTS: The prevalence of psoriasis in the overall population was 2.35‰ (n = 53,761). Mean age (SD) was 46.2 (± 18.6) and 60% were male. Of the 53,761 psoriasis cases, 17.2% (n = 9,265) were the sPsO type.sPsO patients had a significantly increased prevalence ratio (RR; [95% confidence interval ])for hypertension (1.33; [1.26, 1.40]), diabetes (1.52; [1.42, 1.63]), hyperglyceridemia (1.56; [1.44, 1.69]), heart disease (1.24; [1.14, 1.34]), depression (1.44; [1.23, 1.69]), hepatitis B (1.95; [1.44, 2.65]) and hepatitis C (2.19; [1.49, 3.22]). Increased risk of these comorbidities was also observed in mPsO patients compared to controls, but with a smaller magnitude compared to sPsO. CONCLUSIONS: Psoriasis was associated with a significantly increased risk of comorbidities, especially for those patients with moderate to severe disease. These health consequences should be taken into consideration when evaluating the burdens of psoriasis and designing effective treatment plans.

4th Asia-Pacific Abstracts SENSORY SYSTEMS DISORDERS – Cost Studies PSS2 COMPARISON OF COSTING METHODS: COST TO CHARGE RATIO AND DIAGNOSIS RELATED GROUP USING STEVENS-JOHNSON SYNDROME AND TOXIC EPIDERMAL NECROLYSIS AS A CASE STUDY Dilokthornsakul P, Chaiyakunapruk N Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: Several methods have been to estimate costs in economic evaluation studies. Different methods may give different values. The magnitudes of difference of estimates based on cost-to-charge ratio and diagnosis related group (DRG) methods were not much investigated. This study aimed to assess the differences of estimates of direct medical cost calculated by cost-to-charge ratio and DRG method using Stevens—Johnson syndrome (SJS) and Toxic epidermal necrolysis (TEN) as a case study. METHODS: This study was undertaken with retrospective cohort approach using electronic databases of a hospital located in the northern part of Thailand. We included patients who came to the hospital between January 1, 2003 and December 31, 2007 and were coded with International Classification of Disease 10th Edition (ICD-10) of L51.1 (SJS) or L51.2 (TEN). We used cost-to-charge ratio and DRG methods to estimate direct medical cost of SJS/TEN. All cost estimates were converted to reference year (2008) using Thai consumer price index in health product sectors. RESULTS: One hundred and fifty-four patients were included. Mean age was 45.3 years. Fifty-six percent were male and 93% were SJS patients. The mean and median direct medical cost estimated by cost-to-charge ratio was US$918 and US$308, respectively. The mean and median values estimated by DRG were much lower with cost of US$554 and US$264, respectively. The mean estimate based on cost-to-charge ratio was higher than that of DRG by US$364 or 1.65 times. CONCLUSIONS: The cost estimates using cost-to-charge ratio and DRG are different. Selection of cost estimations may impact the cost estimates and cost-effectiveness results. Researchers should consider the perspective of study while choosing the method used for estimating costs. PSS3 THE ECONOMIC BURDENS FOR PATIENTS WITH PSORIASIS IN TAIWAN Tang CH1, Tsai TF2, Hung ST1, Schenkel B3, Tsai IC4, Zhang M3, Chang CW1 1 Taipei Medical University, Taipei, Taiwan; 2National Taiwan University Hospital, Taipei, Taiwan, Taiwan; 3Johnson & Johnson Pharmaceutical Services, LLC, Horsham, PA, USA; 4 Janssen Cilag Taiwan, Taipei, Taiwan OBJECTIVES: Psoriasis is a chronic autoimmune disease that affects approximately 2.4% of the Taiwanese population. However, little is known about its economic burdens. This study estimated those burdens for patients with psoriasis in Taiwan. METHODS: Patients who had at least one outpatient visit or admission diagnosed with ICD-9-CM code 696.0–1 in the Taiwan National Health Insurance (NHI) claims database during 2006 were identified as psoriasis cases. They were further classified into moderate to severe psoriasis (sPsO) for those who had previously received systemic therapy and mild psoriasis (mPsO) for those who had not. The cases were matched with controls in a 1:4 ratio by age, gender and residential area from a representative NHI dataset. Differences in health-care utilization between cases and controls were assumed to be attributable to psoriasis. NHI costs were assessed based on a two-part model: a logistic regression to predict the probability of use of services and a generalized linear model to predict utilization for users of services. Out-ofpocket costs, transportation costs, and productivity loss from patients and caregivers were calculated based on face-to-face patient interviews (n = 327) during July 2009 to March 2010. All costs were reported in 2009 New Taiwan dollars (1 US$ = 31.91 NT$). RESULTS: The case-control analysis showed that the NHI cost for patients with sPsO and mPsO was 2.34 times and 1.58 times the cost of controls, respectively. The total cost due to sPsO was NT$56,885 per patient (70%, 18%, and 12% for NHI cost, patient out-of-pocket cost, and productivity loss, respectively). For mPsO, the total cost was NT$28,573 (56%, 33%, and 11% for NHI cost, patient out-ofpocket cost and productivity loss, respectively). Total direct and indirect costs of psoriasis are estimated to be NT$1.80 billion annually. CONCLUSIONS: Psoriasis is associated with a significant economic disease burden in Taiwan. PSS4 BURDEN OF DISEASE IN PATIENTS WITH GLAUCOMA IN CHINA: RESULTS FFROM 2009 NATIONAL HEALTH AND WELLNESS SURVEY Tang B1, Annunziata K2, Zhong Z3, Dong P3, Xuan J1 1 Pfizer, New York, NY, USA; 2KantarHealth, Princeton, NJ, USA; 3Pfizer China, Beijing, China OBJECTIVES: To assess comorbidity, quality of life (QOL), work/productivity loss, and medical resource utilization in patients suffer from glaucoma. METHODS: Patients’ self-reported data were collected from 2009 National Health and Wellness Survey (NHWS). Survey samples represented major urban areas in China. QOL was measured by the physical component score (PCS) and mental component score (MCS) of the Short Form-12 (SF-12) (mean score of 50 for general population). Loss of work/ productivity was measured by the validated Work Productivity and Activity Impairment (WPAI) instrument. Medical resource utilization (MRU) was measured by health-care provider, emergency room (ER) visits and hospitalization in the past 6 months. Comparisons were made between respondents who were diagnosed glaucoma versus respondents without glaucoma (non-glaucoma group). Since glaucoma typically affects adult population, respondents with age 35 and above were included in the analysis. RESULTS: Of the 7949 survey respondents with age 35 and above, 85

4th Asia-Pacific Abstracts (1.07%) respondents were diagnosed glaucoma. The average age in glaucoma group was 62.9 years and 43.3% were female compared to 51.9 years old with 49.3% female in non-glaucoma group. Glaucoma group reported more comorbidities, lower mean scores of PCS (39.1 vs. 47.1) and MCS (41.8 vs. 46.9), more patients visited healthcare providers (84.7% vs. 56.6%), ER (35.0% vs. 18.3%) and hospitalization (20.6% vs. 8.5%) in the past 6 months compared to non-glaucoma group. Also, glaucoma group reported 39.5% work/productivity loss (absenteeism and presenteeism) and 42.1% impairment in daily activity compared to 21.8% and 23.3% in non-glaucoma group. All comparisons in QOL, MRU, and work/productivity loss between two groups were statistically significant at P < 0.05. CONCLUSIONS: From the China NHWS results, glaucoma patients suffer from impairment in quality of life, work/ productivity loss, more comorbidities and use of medical services. The findings indicate there is still an unmet medical need in glaucoma patients in China. PSS5 HEALTH RESOURCE UTILIZATION AND COSTS IN PATIENTS WITH AGE-RELATED MACULAR DEGENERATION IN THAILAND: PRELIMINARY FINDINGS Ruamviboonsuk P1, Ratanasukon M2, Dilokthornsakul P3, Chanatittarat C4, Pokawattana N1, Tungsomeroengwong A1, Chaiyakunapruk N3 1 Rajavithi Hospital, Rajavithi, Bangkok, Thailand; 2Prince of Songkla University, Muang, Songkla, Thailand; 3Naresuan University, Muang, Phitsanulok, Thailand; 4Norvatis Pharmaceutical, Klongtoey, Bangkok, Thailand OBJECTIVES: Age-related macular degeneration (AMD) is the leading cause of blindness and visual impairment. The prevalence increases proportionally to the number of aging population. Little is known about its economic consequences, therefore; this study is undertaken to determine economic outcomes associated with AMD in Thailand from the societal perspective. METHODS: Health resource utilization and direct medical costs were estimated based on a retrospective medical chart review of AMD patients at a University hospital during 2008–2009. We collected outpatient drug, intervention, medical visit and laboratory costs. Charges were converted to costs using a cost-to-charge ratio. Direct nonmedical costs and indirect costs were obtained from patient’s interview using a structured questionnaire at a University-affiliated hospital. Patients were inquired about costs of transportation, additional food, medical device, house modification, and opportunity costs of care giver. Descriptive statistics were used. All costs were adjusted to year 2009 values. RESULTS: Twenty-two patient’s medical records were reviewed. Mean age was 67 years with 64% of male. The average number of medical visits was 11, while the cost for each outpatient visit was 233 USD. Drug costs accounted for 79% of total direct medical costs. The annual mean direct medical costs of AMD were USD2963. For direct nonmedical and indirect costs, 31 patients were included. Mean age was 62 years with 39% of male. Four patients (12.9%) bought extra medical devices. A quarter of them (26%) needed a care giver to take care at home. On average, the amount of time care givers spent for caring patients was 11.3 hours/month. The mean annual direct nonmedical cost was USD168 and USD85 for the opportunity costs of care giver. CONCLUSIONS: The economic consequences AMD were substantial. These estimates can be used as input parameters in a study evaluating the cost-effectiveness of interventions in patients with AMD. PSS6 ASTIGMATISM AND COST OF POST-CATARACT SURGERY SPECTACLE WEAR IN KOREA Kim H1, Lim S2, Cho B3, Sohn Y4, Kim T5, Tan R6, Gibbs M7, Shin H8 1 Korea University Medical Center, Korea, South Korea; 2Seran Eye Hospital, Seoul, South Korea; 3Hangil Eye Hospital, Korea, South Korea; 4Kim’s Eye Hospital, Korea, South Korea; 5 Yonsei University Severence Hospital, Korea, South Korea; 6Alcon Laboratories R&D International, Singapore; 7IFL, Shanghai, China; 8Alcon Laboratories R&D International, Seooul, South Korea OBJECTIVES: To estimate the direct and indirect economic costs of spectacle wear in Korea in patients with and without astigmatism. Intraocular lenses (IOLs) improve visual acuity after cataract surgery; however, patients must sometimes continue to wear spectacles. To date, there is a paucity of published information on the economic consequences of spectacle wear in Korea and the impact of wearing spectacles on an individual’s quality of life. METHODS: Participants to this cross-sectional survey met the following criteria: 1) Age > 45 years at the time of cataract surgery; 2) subject having had senile cataract surgery in both eyes; and 3) subject needing spectacles after cataract surgery. Patients completed a self-administered questionnaire on: demographics, spectacle and surgical costs, indirect costs and impact on daily activities. Lifetime spectacle costs were calculated from discounted average cost of spectacles, frequency of replacement and average life expectancy. The economic perspective was of the patient. RESULTS: A total of 150 patients were recruited from five eye hospitals in Korea. 41% of patients spend 30–60 minutes traveling to and from an appointment with 15% spend over 90 minutes. Indirect costs associated with traveling are also incurred by the relatives who accompany 67% of patients to their appointments. Patients with post cataract surgery astigmatism paid approximately twice as much for their spectacles as patients with no post surgery astigmatism (KRW431,531 and 212,815, respectively.) The additional lifetime cost of spectacles for patients with postsurgery astigmatism is KRW2,065,152 which is higher than the associated additional out of pocket cost. CONCLUSIONS: Post surgery astigmatism can result in higher out of pocket (OOP) spectacle costs in Korea. These lifetime OOP costs are greater than the OOP costs associated with implanting premium IOLs. A better understanding of local spectacle costs has meaningful implications for patient decisionmaking related to IOL choice.

A563 PSS7 ECONOMIC EVALUATION ON VISION SCREENING FOR PRESCHOOL CHILDREN Huang CH1, Chen MJ2, Tan HY3, Chow IHI1, Pwu RF4, Tang CH1 1 Taipei Medical University, Taipei, Taiwan; 2Taipei City Hospital, Taipei, Taiwan; 3Chang Gung Memorial Hospital, TaoYuan, Taiwan; 4Division of Health Technology Assessment, Taipei, Taiwan OBJECTIVES: To evaluate the cost-effectiveness of a vision screening program for early detection of preschool children with amblyopia in Taipei City, Taiwan. METHODS: A decision analytic model was established to simulate the natural history of amblyopia in the context of Health Bureau of Taipei City Government. A societal perspective was undertaken to perform the cost-utility analysis and cost-benefit analysis. Cumulative quality-adjusted life-years, lifetime benefits, and lifetime costs for the following two strategies were compared: 1) A vision screening program was implemented toward preschool children aged 4–6; 2) No vision screening program. Net benefits and incremental cost-effectiveness ratio (ICER) were calculated to assess whether vision screening programs is cost-effective and cost-benefit. All costs and outcomes were discounted at an annual rate of 3%. One-way sensitivity analysis and probabilistic sensitivity analysis were performed to account for the uncertainty of the parameters. RESULTS: The model predicted the discounted average life-time QALYs were 29.2067 QALYs per child for strategy 1, and 29.2051 QALYs for strategy 2. The discounted average total lifetime costs were NT$723.9 and NT$593.6 per child for strategy 1 and strategy 2, respectively. ICER for the screening strategy was NT$80,932/QALY. The net social benefits was NT$160.7 per child and the vision screening program for preschool children could potentially bring approximately NT$6,900,297 net benefit to the society. Results were most sensitive to the prevalence rate of amblyopic. The vision screening strategy was always preferred when a threshold of willingness-to-pay for a QALY was >= NT$50,000. CONCLUSIONS: In comparison with no screening program, ICER for the vision screening program was found to be less than NT$555,760 (three times GDP per capita in 2007 in Taiwan). The vision screening program for preschool children aged 4–6 were cost-effective in Taipei City, Taiwan. PSS8 COST-EFFECTIVENESS OF BRINZOLAMIDE/TIMOLOL FIXED COMBINATION (BTFC) VERSUS DORZOLAMIDE/TIMOLOL FIXED COMBINATION (DTFC) FOR PATIENTS WITH OPEN ANGLE GLAUCOMA IN KOREA AND TAIWAN Tan R Alcon Laboratories R&D International, Singapore OBJECTIVES: To assess the cost-effectiveness of BTFC compared with DTFC using a payer perspective in Korea and Taiwan. METHODS: A 1-year decision analytic model simulated the effects on cohorts of open angle glaucoma patients in Korea and Taiwan from a payer perspective. Clinical effectiveness data (responder rates: percentage of patients that reach target intraocular pressure [IOP] of 18 mmHg) were derived from a non-inferiority phase III randomized clinical trial (RCT) of BTFC versus DTFC (Manni 2009). Comfort data were taken from a RCT comparing BTFC versus DTFC (Vold 2008). Compliance rates in relation to comfort (Barnebey 2000), risk of visual field defect (VFD) progression (Stewart 1993), and subsequent incremental medication utilization were estimated from literature. Medication costs were obtained from list prices. QALYs related to VFD were estimated from Korean-specific and other Asian published data sources. Costs and effects were discounted at three percent and expressed in local currency. Cost-effectiveness threshold level was set at equivalent to the respective country’s gross domestic product (GDP) per capita. Sensitivity analyses were performed. RESULTS: The BTFC responder rate was 61% versus 59% for DTFC. Possibly due to the more physiological pH of the BTFC formulation, the discomfort rate was significantly better at 51% for BTFC and 83% for DTFC. BTFC economically dominated DTFC in Korea with 76,980 KRW savings and a gain of 0.02 QALYs. In Taiwan, Azarga was highly cost-effective with an incremental cost of 800 TWD, a QALY gain of 0.0259, and an incremental cost effectiveness ratio (ICER) of approximately 5% of GDP per capita. Sensitivity analyses demonstrated the robustness of these findings. CONCLUSIONS: BTFC is non-inferior to DTFC with regards to IOP lowering potential, has a significantly better comfort profile, and is highly cost-effective in both Korea and Taiwan.

SENSORY SYSTEMS DISORDERS – Patient-Reported Outcomes Studies PSS9 SUMMARIZING UTILITY VALUES OF PATIENTS WITH AGE-RELATED MACULAR DEGENERATION: A SYSTEMATIC REVIEW AND META-ANALYSIS Dilokthornsakul P, Chaiyakunapruk N Naresuan University, Muang, Phitsanulok, Thailand OBJECTIVES: Age-related macular degeneration (AMD) is the leading cause of blindness and visual impairment which is associated with poor health-related quality of life. However, there remains a lack of good summary of pooled estimates of health utilities for AMD. This study aims to systematically review all studies estimating utility of AMD patients and meta-analyze them according to different visual acuity level. METHODS: A systematic search was performed using computerized databases including Pubmed, EMBASE, CINAHL, Cochrane library, IPA, Psych Info, CEA registry

A564 and Central of review and dissemination (CRD databases) from 1966–June 2009. Keywords included macular degeneration with preference or utility or names of all possible techniques/tools used to derive utility or other synonyms. Original articles reporting health state utilities of patients with AMD were included. All articles were reviewed and extracted independently by two investigators for study design, population and outcomes. Meta-analyses were performed using DerSimonian and Laird method under a random-effects model. RESULTS: Out of 1887 studies identified, 24 studies met inclusion criteria. Twenty-one studies obtained utility directly from patients, while the others inquired ophthalmologists or community members to estimate utility. Standard Gamble (SG) and time-trade off (TTO) were the most commonly techniques used in these studies (11 studies used TTO, while 5 studies used SG). Based on a meta-analysis of three studies which obtained utility from 510 patients and provided sufficient data for calculation, the utility values with 95% confidence intervals for visual acuity of 20/20–20/25, 20/30–20/50, 20/60–20/100, 20/200–20/400 and