Pharmacy & Therapeutics - November 2008 - (Page 654)

Switching from Fixed-Dose to Free-Combination Antihypertensive Medications for 30-day supplies for each prescription fill and then filled the ARB prescriptions on January 1, 2004, February 15, 2004, and March 31, 2004. The patient then filled the HCTZ prescriptions on January 1, 2004, February 15, 2004, March 31, 2004, and April 30, 2004 with no further refills. Such a patient was defined as being “no longer persistent” as of April 30, 2004—the day on which both drugs were not available to the patient. Medication compliance was measured by the medication– possession ratio (MPR) for the one-year follow-up period: MPR = (days supply of the medication filled during one year) divided by 365 × 100. For patients in the FC cohort, both medications had to be available to the patient on the same day for that day to be included in the numerator of the medication–possession ratio. A drug was considered to be available to the patient on all calendar days from the fill date to end of day’s supply for the prescription fill (e.g., from January 1 through January 30 for a prescription filled on January 1 with a 30-day supply). Thus, for the FC cohort, all calendar days on which both drugs were available to patients were identified and included in the numerator of the medication–possession ratio. During the 12-month follow-up period, we identified hypertension-related health care services received in a hospital, an emergency department, and physician office settings from medical claims with a primary diagnosis code for hypertension (ICD-9-CM 401.XX–404.XX). For each of the three service settings, we measured hypertension-related utilization of resources as the percentage of patients receiving hypertensionrelated health care in that setting. Using total reimbursements from claims data, we created two hypertension-related health care expenditure variables: (1) total health care expenditures over the 12-month follow-up period for services with a primary diagnosis of hypertension, and (2) total expenditures for hypertension-related services and medications (i.e., study drugs and all other hypertensionrelated agents). To estimate the difference between the FDC and FC cohorts in annual hypertension-related costs associated with the difference in compliance for the cohorts, we computed the product of (1) the percentage point difference in compliance for FDC and FC cohorts estimated in the generalized linear model of compliance, and (2) the change in hypertension-related costs for each percentage point change in compliance estimated in the generalized linear model of hypertension-related expenditures. We also computed similar estimates for the annual risks of hypertension-related hospitalization, emergency admissions, or physician-office visits. Outcomes were analyzed separately for Medicare beneficiaries, commercially insured patients (“commercial”), and the two groups combined (“total”). Statistical software (Stata and SAS) was used to conduct all analyses. RESULTS A total of 14,449 patients taking either antihypertensive FDC or FC agents within the same drug classes were enrolled (Table 1, page 660). The sample included 1,216 patients switching from an FDC of an ARB/HCTZ; 1,331 patients switching from an FDC of an ACE-inhibitor/HCTZ; and 4,678 patients switching from an FDC of an ACE-inhibitor/calcium-channel blocker and their respective matched controls (N = 7,224), who continued with their corresponding FDC medications. Overall, the treatment groups were closely matched between FC and FDC cohorts for all demographic variables and risk factors. This method verified that the propensity scorematching algorithm was successful in selecting cohorts that were balanced on these characteristics. Of the two cohorts, 8,217 patients were commercially insured and 6,232 patients had Medicare coverage. Mean patient age was 62.06 years (standard deviation [SD] ± 12.67) for the FDC cohort and 62.86 years (SD ± 13.10) for the FC cohort. Women were similarly represented in 56.9% of both cohorts. Prevalence rates for comorbid conditions and risk factors were well matched across the two groups but were relatively low in the overall study population based on the six-month time frame assessed. Statistical Methods We used chi-square tests for proportions and t-tests for means to compare descriptive statistics of outcomes for the FDC and FC cohorts. We estimated the FDC–FC differences in compliance using generalized linear models with the log–link function and gamma distribution,44,45 and we estimated differences in persistence for FDC and FC using logistic regression. All models included patient demographics, medical comorbidities, and risk factors as control variables. We used multivariate logistic regression models to estimate the effect of improved compliance on the risk of hospitalization, emergency admissions, and physician office use for hypertension-related services, and we used generalized linear models to estimate the effect of improved compliance on expenditures for hypertension-related services and medications. The log–link function and gamma distribution were used for the generalized linear models to address the skewed distribution of the expenditure data.44,45 The utilization and expenditure models controlled for the patient’s study cohort, demographics, health care expenditures (measured six months prior to the index date), and medical comorbidities and risk factors (measured six months prior to the index date). Persistence Persistence with therapy declined more rapidly over time for patients who switched from the FDC to the FC regimen. The FDC–FC difference in persistence was greater for Medicare patients than for commercially patients (Table 2, page 661). At the end of the 12 months of follow-up, persistence rates for FDC and FC were 58.3% and 14.9%, respectively (P < 0.001) for the total sample; 56.2% and 15.2%; for commercial coverage; and 61.2% and 14.4% for Medicare coverage (P < 0.001 for all contrasts) (see Table 2 for unadjusted rates). Multivariate regression-adjusted differences in persistence for FDC, compared with FC regimens, were 42.5% for the total sample, 40.4% for commercial patients, and 45.2% for Medicare patients (P < 0.001 for all contrasts) (see Table 2 for regressionadjusted differences). Compliance Patients who continued taking the FDC regimen also had significantly higher rates of compliance, compared with those continued on page 660 654 P&T® • November 2008 • Vol. 33 No. 11

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

Pharmacy & Therapeutics - November 2008 Contents Editorial Medication Errors Prescription: Washington New Drugs/Drug News/New Medical Devices Drug Forecast Heparin-Induced Thrombocytopenia Medication Utilization Patterns and Hypertension-Related Expenditures among Patients Who Were Switched From Fixed-Dose to Free-Combination Antihypertensive Therapy European Society for Medical Oncology and Association for the Study of Bone and Mineral Research Pharmaceutical Approval Update

Pharmacy & Therapeutics - November 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.