Summary. Introduction: Venous thromboembolism (VTE) poses a significant health ...... HIPAA. Medical Privacy â National Standards to Protect. Hospital-based ...
Journal of Thrombosis and Haemostasis, 6: 1077–1086
DOI: 10.1111/j.1538-7836.2008.02997.x
ORIGINAL ARTICLE
Hospital-based costs associated with venous thromboembolism treatment regimens G . M E R L I , * C . F E R R U F I N O , J . L I N , à M . H U S S E I N and D . B A T T L E M A N *Thomas Jefferson University, Philadelphia, PA; IMS Health, Falls Church, VA; and àSanofi-Aventis, Bridgewater, NJ, USA
To cite this article: Merli G, Ferrufino C, Lin J, Hussein M, Battleman D. Hospital-based costs associated with venous thromboembolism treatment regimens. J Thromb Haemost 2008; 6: 1077–86.
Introduction Summary. Introduction: Venous thromboembolism (VTE) poses a significant health and economic burden in US hospitals. Clinical guidelines for acute VTE treatment recommend antithrombotic therapy (at least 5 days) with low molecular weight heparin (LMWH) or unfractionated heparin (UFH). With upcoming US national performance measures requiring successful implementation of evidence-based therapy, cost considerations for anticoagulant choice are of increasing importance to hospitals. Methods: This retrospective cohort analysis utilizes discharge records from a large real-world US population (January 2002 to December 2006) to provide total, direct, inpatient medical costs associated with LMWH and UFH for acute VTE treatment. Furthermore, for both LMWH and UFH discharges, we compare VTE-related readmission rates at 30 and 90 days after discharge. Results: In total, 57 131 discharges were identified (57.7% LMWH; 42.3% UFH). After adjustment for covariates, including age, severity of illness, and length of stay, total direct medical costs per hospital discharge for UFH were $3476.22 vs. $3056.42 for LMWH (P < 0.0001; difference $420). Costs were significantly higher in the UFH group for most cost categories. Notably, drug acquisition cost was higher for LMWH. LMWH treatment was 12% [odds ratio (OR) 0.876; P < 0.001] and 10% (OR 0.895; P = 0.0006) less likely to result in VTE readmission within 30 and 90 days, respectively. Conclusions: This study provides the first large, real-world analysis of the total direct medical costs of treating VTE in-hospital. It confirms that, despite higher drug acquisition costs, LMWH is cost-saving compared with UFH in the inpatient setting, and is associated with a lower VTE readmission rate at 30 and 90 days than is UFH. Keywords: deep vein thrombosis, heparins, hospital cost, low molecular weight heparin, pulmonary embolism. Correspondence: Geno Merli, Jefferson Center for Vascular Diseases, Jefferson Medical College, Thomas Jefferson University Hospital, 833 Chestnut Street, Philadelphia, PA, USA. Tel.: +1 215 955 6540; fax: +1 215 503 2203. E-mail: geno.merli@jefferson.edu Received 25 September 2007, accepted 6 April 2008 Ó 2008 International Society on Thrombosis and Haemostasis
Venous thromboembolism (VTE) encompasses both deep vein thrombosis (DVT) and pulmonary embolism (PE), and is a leading cause of morbidity and mortality [1]. In the USA, there are an estimated 300 000 VTE-related deaths annually [2]. This common and potentially lethal disease recurs frequently and is associated with serious long-term complications. Up to 30% of patients who experience VTE develop a recurrent VTE episode within 10 years [3]. VTE poses not only a tremendous health challenge but also an economic burden on the healthcare system. Each year in the USA, nearly one million people develop symptomatic VTE [2], and the predicted annual cost of managing these events is high. National expenditures associated with DVT events diagnosed in the hospital setting alone are estimated at $1.5 billion per year [4]. The economic burden of VTE is not only restricted to initial events but is amplified by the costs and utilization of medical resources for managing recurrent VTE events [5,6]. As many as 25% of patients presenting with an initial VTE event require rehospitalization [5], and often readmissions incur greater costs than the initial episode [6]. Anticoagulant therapies are used as the frontline treatment for both acute DVT and PE [7]. Evidence-based guidelines from the American College of Chest Physicians provide physicians with recommendations for appropriate treatment regimens for patients with confirmed or suspected VTE [7]. These guidelines highlight a key role for low molecular weight heparin (LMWH) and unfractionated heparin (UFH) in the treatment of acute VTE [7], with clinical evidence demonstrating that LMWH treatment is as effective and well-tolerated as UFH in the treatment of VTE [7,8]. A meta-analysis of randomized controlled trials of patients with objectively diagnosed DVT found that the rates of recurrent VTE at 3 months after acute treatment were 4.5% for LMWH and 5.7% for UFH. The rates of major bleeding at 3 months were 2.9% in the LMWH group and 4.3% in the UFH group [9]. Current clinical guidelines on the initial treatment of VTE provide Grade 1a recommendations that patients are treated for at least 5 days with subcutaneous (s.c.) LMWH or intravenous (i.v.) or s.c. UFH, together with vitamin K antagonist therapy started on day 1 [7]. Newer anticoagulant therapies, such as the synthetic selective anti-factor Xa agent
1078 G. Merli et al
fondaparinux, have been evaluated in clinical trials for initial treatment of DVT in combination with warfarin, but are not yet included within current practice guidelines as a first-line treatment option [7]. In addition to clinical guidelines, the Joint Commission and the National Quality Forum have recently introduced a set of quality assurance measures within the ÔNational Consensus Standards for the Prevention and Care of Venous ThromboembolismÕ project [10], with the specific goal of improving inhospital VTE assessment, diagnosis, prophylaxis, and treatment. With national performance measures highlighting the need for hospitals to examine their VTE treatment practices, the relative costs of different regimens are of particular importance to healthcare decision-makers. Prior studies have focused predominantly on reimbursed costs and thus the payerÕs perspective [5,6]. However, few studies have assessed the total direct hospital costs of acute VTE treatment from drug acquisition to hospitalization costs for patient care. This study therefore compares the total direct hospital costs associated with UFH or LMWH for acute VTE treatment in the inpatient setting. The study utilizes discharge records from the Premier Perspective database, a large, hospital-based transactional database in the USA, and examines the total, direct, inpatient medical costs and readmission rates associated with LMWH and UFH regimens for acute VTE treatment.
VTE treatment during the index hospitalization; received more than one class of drug (apart from LMWH or UFH in conjunction with warfarin), or a regimen involving crossover between different regimens during the index hospitalization; and were discharged with comorbid conditions considered to be contraindicative to anticoagulant therapy, including malignant hypertension, blood disease, active peptic ulcer, liver disease, renal insufficiency, and HIV (see Appendix B for a complete list of exclusion criteria ICD-9 codes). The discharge records fulfilling all the study inclusion criteria and none of the exclusion criteria were then stratified according to the treatment regimen received during the index hospitalization, namely LMWH, UFH, fondaparinux, or other anticoagulants. Discharged patients could have been receiving concurrent warfarin with either of these regimens. An additional, post hoc decision was made to exclude discharges where the index regimen was fondaparinux or other anticoagulant therapies from the final analysis, due to their having insufficient sample sizes for comparison. The All-Patient Refined Diagnoses-Related-Group (APRDRG) Disease Severity Level Scale was used to control for disease categories and severity of illness [13]. This scale categorizes patients into similar disease groups and, within categories, rates the severity of disease into minor, moderate, major, extreme severity of illness or risk of mortality, or unknown.
Materials and methods The study was a retrospective analysis of inpatient hospital discharges and billing records from the Premier Perspective database from January 2002 through December 2006. The Premier Perspective database contains information from approximately 5.5 million patient discharges per year across the USA, from not-for-profit, non-governmental, community and teaching hospitals and health systems. All patient records used in the study were de-identified in accordance with the Health Insurance Portability and Accountability Act of 1996 [11]. This study did not involve Ôhuman subjectsÕ and was therefore exempt from Institutional Review Boards review under the Common Rule [45 CFR §46.001(b)(4)] [12]. Study population
Discharge records were included in the analysis if patients met all of the following criteria: age 18 years or older at the time of their index (first VTE hospitalization during the study period) hospitalization; had a complete hospital stay (i.e. both an admission and discharge date, and a discharge status other than ÔexpiredÕ or ÔunknownÕ); had a primary diagnosis of VTE during their index hospitalization (see Appendix A for the complete list of ICD-9 codes used to identify VTE diagnoses); and received treatment with an LMWH or UFH alone, or in combination with warfarin. Discharge records were excluded if patients: had an inpatient or outpatient diagnosis of VTE at any time during the 12month preindex period; did not receive LMWH or UFH for
Study period
Three discrete time periods were used in this study: the identification period (January 2003 to September 2006); the preindex period for each discharge (12 months prior to the admission date for the index hospitalization); and the follow-up period for each discharge (the first 90 days following discharge from the index hospitalization). All records pertaining to the index hospitalization and any subsequent hospitalization(s) during the dischargeÕs follow-up period were extracted for further processing. Costs
Total direct hospital costs of VTE treatment was calculated as the sum of in-hospital direct costs associated with VTE treatment. For discharges with a primary diagnosis of VTE for the index hospitalization, all costs associated with the hospitalization were included. Costs per category (cardiology/ electrocardiography, emergency room, laboratory, operating room, other, pharmacy – anticoagulant, pharmacy – other, professional fees, radiology, respiratory, patient room and board, medical supply, therapy) were also collected and compared between groups. Readmissions
Records pertaining to the first readmission within 30 or 90 days of the date of discharge from the index hospitalization were Ó 2008 International Society on Thrombosis and Haemostasis
Hospital-based costs associated with VTE 1079
used in readmission analyses. Readmission was classified as either VTE-related (those with a primary or secondary diagnosis of VTE) or as VTE-unrelated (those without any diagnosis of VTE) based on ICD-9 codes (see Appendix A). Statistical analysis
The analyses included descriptive statistics with measures of central tendency (mean, median) and measures of dispersion (range and standard deviation) calculated for continuous variables for each group. Clinical and demographic characteristics were calculated and compared across study cohorts using univariate statistics. Continuous variables were compared with t-tests, and categorical variables were compared with chisquare tests. Total direct medical costs in this study were compared using generalized linear models (GENMOD), adjusting for patient and hospital characteristics such as race, age group, sex, admission status, discharge status, payer, physician specialty, APR-DRG severity level, hospital teaching status, urban or rural status, provider geographic area, and
length of stay (LOS). Outliers (the top 1%) in the cost distributions were excluded. VTE-related readmission rates, across the treatment regimens, were compared using logistical regression analysis. Results were considered statistically significant if the statistical test yielded a P-value < 0.05. All analyses were conducted using SASÒ Release 9.1 (SAS Institute Inc., Cary, NC, USA) statistical software. Results In total, 57 131 discharge records meeting the study inclusion criteria were identified from 499 hospitals. These comprised 32 959 (57.7%) discharges in the LMWH group and 24 172 (42.3%) in the UFH group (Fig. 1). Within the LMWH group, enoxaparin was the most commonly prescribed LMWH therapy, corresponding to 32 141 (97.5%) discharges. The demographics and clinical characteristics were generally similar between the LMWH and UFH groups (Table 1). However, on the basis of the APR-DRG severity level, the UFH cohort appeared to be a slightly sicker population.
Total discharges n = 147 464 452 Excluded – outpatient discharges and out of study period (n = 134 746 342) Inpatient index discharges in study period n = 12 718 110 Excluded – discharges without primary VTE (n = 12 590 331) 1° VTE diagnosis during index hospitalization n = 127 779
Discharges ≥ 18 years with complete hospital stay n = 119 461
Discharges meeting all inclusion criteria n = 57 131
LMWH group n = 32 959
Excluded – age
