Pharmacy & Therapeutics - September 2008 - (Page 532)

Effect of Prescription Copayments on Adherence and Treatment Failure with Oral Antidiabetic Medications John Barron, PharmD; Peter Wahl, MLA, MS; Maxine Fisher, PhD; and Craig Plauschinat, PharmD, MPH ABSTRACT Background: Previous studies have shown that an increase in cost sharing by patients for medications results in reduced medication use. The purpose of our study was to determine whether the amount of members’ copayments predicted oral antidiabetic treatment failure in a managed care population and to analyze the relationship between copayments, adherence to therapy, and hemoglobin A1c levels in patients with type-2 diabetes. Methods: Health plan members 18 years of age or older with type-2 diabetes and who were newly initiated on an oral antidiabetic drug (OAD) between January 1, 2002, and January 31, 2006, were identified from a managed care population in the U.S. Members were required to have continuous eligibility for six or more months before and 12 months after the index prescription and were placed into four treatment groups. These patients were followed for one year. The time during which they took the index OAD was measured until treatment failure (discontinuation of therapy or a switch of the index OAD) or censoring of patient data. The medication possession ratio (MPR), which was used to calculate adherence, was defined as the number of days that a patient had a supply of the index OAD during the year after the index fill, divided by 365 days. Copays were identified for every prescription, and the amount for each 30 days of treatment was calculated. We used multivariate analyses to assess the impact of copays on treatment failure, adjusting for differences in member characteristics. Results: Adherence was poor, with an overall mean MPR of 0.57, decreasing from 0.58 for the group with a copay of less than $10 to 0.52 for patients with a copay of $20 or more. Initial treatment failed for 13,091 patients (70%), with approximately 60% of all members discontinuing treatment. This discontinuation rate ranged from 55% in the lowest copayment group (below $10) to 67% in those with a copay of $30 or more. For every $10 increase in copay, OAD treatment was 26% more likely to fail (95% confidence interval, 22.3–29.8%; P < 0.0001). Conclusion: Higher copays for health plan members were a significant predictor of treatment failure. With the increased trend toward more cost sharing by members of health plans, ongoing evaluation will be necessary to determine the impact on treatment persistence in patients with chronic conditions. INTRODUCTION Diabetes is a highly prevalent disease that affects 7% of the population and consumes one of every 10 health care dollars spent in the U.S.1,2 Type-2 diabetes affects 90% to 95% of all patients who have diabetes.1 In addition to relying on diet and exercise to achieve adequate glycemic control, most patients with diabetes use insulin, oral antidiabetic drugs (OADs), or both.3 Adherence to OADs is a key strategy in achieving longterm glycemic control.4 However, overall adherence levels to these drugs remain poor, exacerbated by complex treatment regimens in patients with multiple comorbid conditions.5,6 Studies have shown that increased adherence to OADs is correlated with improved outcomes—namely decreased glycosylated hemoglobin (HbA1c) levels—and health economic benefits in these patients.7–13 Furthermore, in numerous studies of patients with chronic conditions, higher copayments predict decreased adherence to medication regimens.14–18 We sought to determine whether the amount of members’ copays predicted treatment failure with OADs. Treatment failure was defined as discontinuation of or a therapy switch from the initial drug regimen in a managed care population. We also analyzed the relationship between copay, adherence, and HbA1c levels in a sample of patients with type-2 diabetes. METHODS Data Our retrospective study included patients identified from administrative claims databases from seven commercial health plans in the southeastern, mid-Atlantic, central, and western regions of the U.S. Automated health plan enrollment, medical, and pharmacy administrative claims files for more than nine million individuals were used. We deleted denied, rejected, and duplicate claims from the analytic file to ensure that the analytic file reflected only the true direct medical costs related to the delivery of health care. A limited data set, as defined by the Health Insurance Portability and Accountability Act (HIPAA) of 1996, was used for the data analysis. HealthCore, Inc., had all HIPAA required business associate and data use agreements in place before the research was conducted. Dr. Barron is Research Operations Director and Dr. Wahl and Dr. Fisher are Senior Research Analysts, all at HealthCore, Inc., in Wilmington, Delaware. Dr. Plauschinat is Associate Director of Evidence-Based Medicine at Novartis Pharmaceuticals Corporation in East Hanover, New Jersey. Patients Patients with type-2 diabetes were identified from review of medical claims and pharmacy data between January 1, 2002, Disclosure: Novartis provided funding to HealthCore for this research and for other research projects during the past three years. Accepted for publication April 16, 2008. 532 P&T® • September 2008 • Vol. 33 No. 9

Table of Contents Feed for the Digital Edition of Pharmacy & Therapeutics - September 2008

Pharmacy & Therapeutics - September 2008 Contents Editorial Medication Errors Prescription: Washington New Drugs/Drug News/New Medical Devices Drug Forecast Effect of Prescription Copays on Adherence and Treatment Failure with Oral Antidiabetic Medications Vaccine Declinations Present New Challenges for Public Health Universal Health Care in America Digestive Disease Week and American Diabetes Association Pharmaceutical Approval Update

Pharmacy & Therapeutics - September 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.