Impact of a pharmacist-directed intervention in ... - Writing in Pharmacy

Practice Reports Pharmacist-directed intervention

PRACtice RePORTS

Impact of a pharmacist-directed intervention in postmenopausal women after fracture Rachel M. F. Heilmann, Cari R. Friesleben, and Sarah J. Billups

O

steoporosis affects an estimated 44 million Americans age 50 years or older.1 One of every 2 women over age 50 years will have an osteoporosis-related fracture in her lifetime, substantially affecting her quality of life and health care costs. In 2005, approximately 2 million women in the United States experienced a fracture, costing a total of approximately $19 billion.1 This number is expected to increase to 3 million fractures annually by 2025 due to continued growth in the elderly population. Data from the National Osteoporosis Foundation (NOF) suggest that persons who have had a hip fracture have a fourfold greater risk of having a second hip fracture compared with individuals without a previous fracture.1 The frequency of second hip fractures is 2–11% in the four years following the initial fracture.1,2 Research has also found that women with a history of vertebral fracture are four times more likely to have another vertebral fracture, regardless of their bone mineral density (BMD), compared with

Purpose. The impact of decentralized clinical pharmacy services on the implementation of appropriate care in postmenopausal women with a recent history of a fracture was assessed. Methods. Women 67 years of age or older with a documented fracture between January 1 and December 31, 2007, were identified in two geographic regions. At the intervention site, a decentralized clinical-pharmacy-based osteoporosis management service (CPOMS) intervened on postmenopausal women following fracture, while the comparison group utilized a centralized registered nurse to manage this population. In both groups, interventions included initiation of either osteoporosis medication or bone mineral density (BMD) screening. Results. Of the 827 women in the CPOMS group, 65% (523) initiated a medication for osteoporosis or completed BMD screening within 6 months of the fracture, compared with 46% (139) of the 302 women in the comparison group (p < 0.001; cumulative incidence ratio [CIR], 1.75; 95% confidence interval [CI], 1.44–2.12). CPOMS patients were nearly twice as likely as

women with no history of vertebral fracture.3

Rachel M. F. Heilmann, Pharm.D., BCPS, is Clinical Pharmacy Specialist, Pharmacy Department, Kaiser Permanente Skyline Clinic, Denver, CO. Cari R. Friesleben, Pharm.D., BCPS, is Clinical Pharmacy Specialist, Pharmacy Department, Kaiser Permanente East Denver Clinic, Denver. Sarah J. Billups, Pharm.D., BCPS, is Clinical Pharmacy Specialist, Pharmacy Department, Kaiser Permanente Central Support Services, Aurora, CO. Address correspondence to Dr. Heilmann at the Pharmacy Department, Kaiser Permanente Skyline Clinic, 1375 East 20th Avenue, Denver, CO 80205 ([email protected]).

504

Am J Health-Syst Pharm—Vol 69 Mar 15, 2012

comparison group patients to purchase osteoporotic medications (42% [347 of 827] versus 24% [73 of 302]; CIR, 1.89; 95% CI, 1.50–2.40) and equally likely to undergo BMD testing (35% [289 of 827] versus 31% [94 of 302]; CIR, 1.13; 95% CI, 0.91–1.39). Both interventions achieved results substantially higher than national averages. Of patients receiving a medication and continuous membership for 12 months, adherence was also significantly higher in the CPOMS group (46% [39 of 291] versus 28% [20 of 71], respectively; p = 0.007). Conclusion. An integrated pharmacist-run osteoporosis management service demonstrated a substantial increase in the rate of osteoporosis drug initiation among postmenopausal women who experienced a fracture compared with a centrally located nurse-run service. BMD screening rates did not significantly differ between groups. Index terms: Bone density; Bone resorption inhibitors; Clinical pharmacists; Clinical pharmacy; Compliance; Diagnosis; Fractures; Osteoporosis; Patients; Pharmaceutical services; Postmenopause Am J Health-Syst Pharm. 69; 2012:504-9

Although NOF and others have established practice guidelines rec-

Presented at the Western States Pharmacy Residency Conference, Monterey, CA, May 19–22, 2009. The Kaiser Permanente Colorado clinical pharmacy specialists in primary care and continuing care are acknowledged for their hard work and dedication to improving the care for postmenopausal women after fracture. The authors have declared no potential conflicts of interest. Copyright © 2012, American Society of Health-System Pharmacists, Inc. All rights reserved. 1079-2082/12/0302-0504$06.00. DOI 10.2146/ajhp110309

Practice reports Pharmacist-directed intervention

ommending osteoporosis therapy after a vertebral or hip fracture or with a T-score (i.e., the standard deviation from the average bone mass of a healthy adult) of ≤–2.5 in postmenopausal women,4-7 fewer than 15% of Americans with a diagnosis receive treatment.8-10 Osteoporosis therapies, such as bisphosphonates, have been shown to decrease fracture risk by half but are widely underutilized.3 Many patients for whom bisphosphonates are prescribed either do not initiate therapy or discontinue treatment within a year.11,12 The National Committee for Quality Assurance has developed Healthcare Effectiveness Data and Information Sets (HEDIS) to establish objective measures that allow quality-of-care comparisons among health plans.13 In 2004, the percentage of women age 67 years or older with a fracture (excluding a fracture of the face, skull, fingers, or toes) who have a BMD test performed or begin osteoporosis therapy within six months after a fracture was established as a Medicare-reported HEDIS measure assessing the quality of postfracture care provided to postmenopausal women. Managed care organizations have implemented various interventions to improve their performance on this Medicare-reported HEDIS measure.14-17 The purpose of this study was to compare the capability of a decentralized clinical-pharmacybased osteoporosis management service (CPOMS) with that of a centralized nurse-based service to affect outcomes in postmenopausal women after a fracture. Methods Setting. Kaiser Permanente Colorado (KPCO) is a nonprofit, group model, integrated health care delivery system that provides integrated health care services for over 480,000 members, including 63,000 Medicare recipients age 65 years or older, at 17 outpatient medical offices in

the Denver–Boulder area. Patients’ health encounter information, medication history, and laboratory test results were obtained from electronic medical records (EMRs). KPCO employs decentralized primary care clinical pharmacy specialists and clinical pharmacy specialists in specialties (e.g., nephrology, continuing care, palliative care, cardiology) who work in consultation with primary care providers and specialty providers, nurses, and other health care professionals at each medical office or skilled-nursing facility. The clinical pharmacy specialists in primary care and specialty areas assist patients, providers, and clinic staff with disease management; drug therapy management, including identifying and resolving drugrelated problems; drug information requests; quality initiatives involving drug therapy; and drug affordability issues. Between 0.5 and 2.6 full-timeequivalent (FTE) primary care clinical pharmacy specialists work in each medical office building. The number of FTE clinical pharmacy specialists in specialties varies based on specialty provider needs. CPOMS. KPCO initiated the CPOMS in January 2007. The service was integrated into the workflow of the existing clinical pharmacy specialists in primary care clinics (26 FTEs) and at skilled-nursing facilities (3 FTEs). An analyst within the pharmacy department generated a monthly report that identified women age 67 years or older with Medicare coverage who suffered a HEDIS-defined fracture. One primary care clinical pharmacy specialist served as the coordinator, distributing a clinic-specific report to each clinic and tracking interventions to ensure consistency of care throughout the service. After the clinical pharmacy specialist at each facility reviewed the patient’s medical record, he or she developed a therapeutic plan and sent it to the primary care provider for input or approval.

The therapeutic plan, which was documented in the EMR, could consist of recommendations for BMD screening, initiation of osteoporosis therapy (i.e., bisphosphonates, estrogen replacement therapy, raloxifene, calcitonin, teriparatide), or calcium and vitamin D supplementation as indicated. Once the primary care provider authorized a final plan, the clinical pharmacy specialist contacted the patient to implement the plan. Patients had BMD testing performed at one of two medical offices within the KPCO region. All patients were followed monthly as needed by telephone, postal mail, or e-mail for six months after their fracture to ensure adherence with recommendations. Approximately 400 patients were evaluated monthly by the 29 clinical pharmacy specialists. Comparison group. A geographically distinct integrated health maintenance organization serving approximately 32,000 Medicare members age 65 years or older served as the comparison group. One centrally located registered nurse (RN) reviewed medical records for all women in the comparison group (age 67 years or older with Medicare coverage who had a HEDIS-defined fracture) each month, assessed the appropriateness of either BMD screening or initiation of osteoporosis therapy, and sent recommendations to each patient’s primary care provider via the EMR. As in the CPOMS, the primary care provider reviewed and approved the plan, and the RN implemented it via telephone. The RN evaluated approximately 45 patients per month. One difference in the comparison group was the availability of mobile vans capable of performing dual-energy x-ray absorptiometry scans; the vans traveled between selected medical offices, making BMD testing more convenient for these women. Patients were followed until BMD test results were available or the patient picked up her first osteoporosis medication, if indicated.


505


Patient population. All women with Medicare coverage 67 years of age or older with a diagnosis code for a fracture during a hospital, an urgent care, or an outpatient visit between January 1 and December 31, 2007, were eligible for inclusion. In accordance with the HEDIS measure, patients were excluded if (1) their fracture was limited to the face, skull, fingers, or toes, as these fractures are not deemed to be associated with osteoporosis, (2) their BMD had been measured or a prescription for an osteoporosis medication had been filled in the year before the date of the fracture, or (3) continuous membership in the health plan could not be verified. Continuous membership was defined as no membership gaps greater than 45 days one month before and six months after the fracture date. Outcomes measures. The primary outcome evaluated was either completion of BMD screening or initiation of osteoporosis medication in the six months after the fracture date. The number of patients in each group reaching the primary endpoint was divided by the total number of patients in the group to determine the rate of success. Secondary outcomes included the percentage of patients adhering to osteoporosis medication instructions and the proportion of patients persisting with osteoporosis medication 365 days after the fracture date. Adherence was defined as a medication possession ratio of at least 80%. The medication possession ratio was calculated by dividing the total medication days’ supply purchased by the total number of days between the first medication purchase and 365 days from the fracture date. Medication persistence was defined based on the last medication purchase prior to the 365-day cutoff date. If the days’ supply multiplied by 1.2 was greater than or equal to the number of days between the final prescription purchase and the cutoff date, medication use was con506

sidered persistent. A factor of 1.2 was used to allow for an adherence rate of less than 100% but at least 80%. Data sources. Data were collected from the EMR and pharmacy and membership databases. Fracture dates, fracture types (classified as vertebral, hip, wrist, or other), BMD screening dates, and T-scores were validated by manual review of the medical record. Information regarding medication purchases was obtained from electronic queries of internal pharmacy records using generic product identifier codes. Information on health plan membership eligibility and patient age was obtained from electronic queries of membership databases. This retrospective, parallel-group, cohort study was approved by institutional review boards at both study sites. Statistical analysis. A sample size of 136 women per group was estimated to have 80% power to detect a 15% difference in the primary outcome with an alpha of 0.05. Continuous variables were analyzed using independent t-tests. Categorical variables were evaluated using chi-square analysis. Cumulative incidence ratios (CIRs, determined by dividing the percentage of CPOMS patients with positive outcomes by the percentage of comparison-group patients with positive outcomes) were calculated, along with 95% confidence intervals (CIs). A predetermined subset analysis of the primary outcome was performed for patients with hip or vertebral fractures. An additional subanalysis of patients with a T-score of ≤–2.5 compared the percentage of patients initiating osteoporosis therapy within 60 days of the BMD screening date. All analyses were performed using SAS software, version 9.1.3 (SAS Institute, Inc., Cary, NC). Results Administrative queries identified 1996 and 805 women meeting inclusion criteria in the CPOMS and


comparison groups, respectively. After excluding patients with noncontinuous membership or a BMD test or osteoporosis medication purchase within 12 months before the fracture date, 1056 and 365 patients remained in the CPOMS and comparison groups, respectively. An additional 229 patients in the CPOMS and 63 patients in the comparison group were excluded due to fractures that could not be validated (e.g., fractures that occurred in the distant past or suspected fractures that were later diagnosed as sprains). The final study group included 827 women in the CPOMS group and 302 in the comparison group (Figure 1). Baseline characteristics of age and fracture types of both study groups were similar (Table 1). The composite of osteoporosis therapy initiation or BMD screening during the six months after their fracture was more prevalent in CPOMS patients compared with the comparison group (65% [523 of 827] versus 46% [139 of 302]) (CIR, 1.75; 95% CI, 1.44–2.12). Women in the CPOMS group were nearly twice as likely as women in the comparison group to purchase osteoporosis medications (42% [347 of 827] versus 24% [73 of 302]; CIR, 1.89; 95% CI, 1.50–2.40) and equally likely to undergo BMD testing (35% [290 of 827] versus 31% [94 of 302], respectively; CIR, 1.13; 95% CI, 0.91–1.39) (Table 2). The type of medication initiated did not significantly differ between groups (p = 0.158). Bisphosphonates were most commonly prescribed (93% [272 of 294] in the CPOMS group versus 89% [63 of 71] in the comparison group). In the CPOMS group, calcitonin was initiated in 4% of women (11 of 291) versus 10% in the comparison group (7 of 71). Raloxifene was initiated in 2% of CPOMS patients (6 of 291) and in none of the women in the comparison group. Finally, estrogen was initiated in 1% of pa-

Practice reports Pharmacist-directed intervention

Figure 1. Schematic demonstrating the selection of patients for study inclusion. BMD = bone mineral density.

Pharmacist-Managed Group

Comparison Group

1996 women met screening criteria

805 women met screening criteria

Reasons for Exclusion Noncontinuous membership

255

125

BMD testing performed or drug therapy purchased during 12 months before

170 BMD 515 drug therapy

115 BMD 200 drug therapy

Miscoded fracture or BMD date per chart review

229

63

827 women evaluated for primary outcome

tients in both groups (6 patients in the CPOMS group and 1 patient in the comparison group). Subset analysis of the 338 patients with hip or vertebral fractures found that women in the CPOMS group were more likely to undergo BMD testing or initiate osteoporosis therapy (CIR, 2.41; 95% CI, 1.75–3.32) and nearly three times more likely to initiate osteoporosis therapy than were women in the comparison group (CIR, 2.91; 95% CI, 2.06–4.12) (Table 2). BMD screening results showing osteoporosis with a T-score of ≤–2.5 at the hip, spine, femoral neck, or forearm did not differ significantly between groups (31% [90 of 290 in CPOMS and 23% [22 of 94] in the comparison group, p = 0.224). Of patients with osteoporotic T-scores, 81% of CPOMS patients (73 of 90) and 59% (13 of 22) of comparison group patients started osteoporosis medication within 60 days after the date of fracture (CIR, 2.29; 95% CI, 1.06–4.74) (Table 2).

302 women evaluated for primary outcome

Table 1.

Baseline Characteristics

Variable

PharmacistManaged Group (n = 827)

Comparison Group (n = 302)

Mean ± S.D. age, yr 77.5 ± 7.7 77.6 ± 7.5 Primary fracture site, no. (%) pts Hip 132 (16) 54 (13) Vertebra 100 (12) 52 (17) Wrist 157 (19) 45 (15) Other 438 (53) 151 (50)

Due to gaps in membership in the 12 months after the fracture date, only 291 women in the CPOMS group and 71 women in the comparison group were evaluated for medication adherence and persistence. Adherence was significantly higher in the CPOMS group than in the comparison group (46% [139 of 291] versus 28% [20 of 71], respectively; p = 0.007). Medication persistence did not differ significantly between groups (54% [159 of 291] versus 45% [32 of 71], p = 0.19).

p 0.416 0.097

Discussion Osteoporosis is often underdiagnosed and undertreated.8-10 In this retrospective, parallel-group, cohort study, an osteoporosis intervention service run by decentralized clinical pharmacy specialists was associated with a significantly greater proportion of patients undergoing BMD testing or initiating drug therapy after a fracture when compared with an intervention run by a centralized nurse. In particular, women with the highest risk for recurrent fractures—


507


Table 2.

Proportion of Patients in Treatment Groups Achieving HEDIS Management Targets for Women Who Have Had A Fracturea

Target

No. (%) Patients Comparison CPOMS Group (n = 302) (n = 827)

p

Cumulative Incidence Ratio (95% Confidence Interval)

Outcome 538 (65) 139 (46)