Publications

BACKGROUND: Clinical trials often compare hypoglycemic medications on the basis of glycemic control but do not examine long-term outcomes (e.g., mortality). This study demonstrates an alternative approach to lengthening clinical trials to assess these long-term outcomes.

OBJECTIVE: To use observational quasi-experimental methods using instrumental variables (IVs) to compare the effect of two hypoglycemic medications, sulfonylureas (SUs) and thiazolidinediones (TZDs), on long-term outcomes.

METHODS: This study used administrative data from the Veterans Health Administration and Medicare from 2000 to 2010. The study population included US veterans dually enrolled in Medicare who received a prescription for metformin and then initiated SUs or TZDs. Patients could either continue on or discontinue metformin after the initiation of the second agent. Treatment was defined as starting either a SU or a TZD. Local variations in SU prescribing rates were used as instruments in IV models to control for selection bias. Survival models predicted all-cause mortality, ambulatory care sensitive condition hospitalizations, and stroke or heart attack (acute myocardial infarction).

RESULTS: Starting on SUs compared to TZDs significantly increased the likelihood of experiencing mortality and ACSC hospitalization. The estimated hazard ratio for the effect of starting on SUs compared to TZDs was 1.50 (95% confidence interval [CI] 1.09-2.09) for all-cause mortality, 1.68 (95% CI 1.31-2.15) for ambulatory care sensitive condition hospitalization, and 1.15 (95% CI 0.80-1.66) for acute myocardial infarction or stroke.

CONCLUSIONS: Our findings suggest increased risk of major adverse events associated with SUs as a second-line agent. Quasi-experimental IV methods may be an important alternative to lengthening clinical trials to assess long-term outcomes.

OBJECTIVES: To assess the association between multiple pharmacy use and medication adherence and potential drug-drug interactions (DDIs) in older adults.

DESIGN: Cross-sectional propensity score-weighted analysis.

SETTING: 2009 claims data.

PARTICIPANTS: A nationally representative sample of 926,956 Medicare Part D beneficiaries aged 65 and older continuously enrolled in fee-for-service Medicare and Part D that year who filled one or more prescriptions at a community retail or mail order pharmacy.

MEASUREMENTS: Multiple pharmacy use was defined as concurrent (overlapping time periods) or sequential use (non-overlapping time periods) of ≥ 2 pharmacies in the year. Medication adherence was calculated using a proportion of days covered of 0.80 or greater for eight therapeutic categories (beta-blockers, renin angiotensin system antagonists, calcium channel blockers, statins, sulfonylureas, biguanides (metformin), thiazolidinediones, and dipeptidyl peptidase-IV inhibitors). Potential DDIs arising from use of certain drugs across a broad set of classes were defined as the concurrent filling of two interacting drugs.

RESULTS: Overall, 38.1% of the sample used multiple pharmacies. Those using multiple pharmacies (concurrently or sequentially) consistently had higher adjusted odds of nonadherence (ranging from 1.10 to 1.31, P < .001) across all chronic medication classes assessed after controlling for sociodemographic, health status, and access to care factors than single pharmacy users. The adjusted predicted probability of exposure to a DDI was also slightly higher for those using multiple pharmacies concurrently (3.6%) than for single pharmacy users (3.2%, adjusted odds ratio (AOR) = 1.11, 95% confidence interval (CI) = 1.08-1.15) but lower in individuals using multiple pharmacies sequentially (2.8%, AOR = 0.85, 95% CI = 0.81-0.91).

CONCLUSIONS: Filling prescriptions at multiple pharmacies was associated with lower medication adherence across multiple chronic medications and a small but statistically significant greater likelihood of DDIs in concurrent pharmacy users.

OBJECTIVES: To describe changes in rosiglitazone prescribing following the US Food and Drug Administration (FDA) warning of potentially increased risk of myocardial infarction and removal from the Department of Veterans Affairs National Formulary (VANF), assess patient-level factors associated with rosiglitazone discontinuation, and evaluate changes in glucose control.

STUDY DESIGN: Historical cohort.

METHODS: Veterans with an active outpatient prescription for rosiglitazone on April 1, 2007, were followed until June 30, 2008. Incidence rate ratios (IRRs) of rosiglitazone discontinuation were compared over time using Poisson methods. We identified patient-level factors associated with stopping rosiglitazone using multivariable Poisson regression and compared glycated hemoglobin (A1C) values across time among patients who discontinued/continued rosiglitazone using linear mixed models.

RESULTS: Of 95,539 veterans with an active outpatient rosiglitazone prescription, 86.7% discontinued rosiglitazone. Discontinuation rates increased significantly after the FDA warning, with IRRs from 1.6 to 1.8. After removal from the VANF, rosiglitazone discontinuation rates again increased significantly. Discontinuing rosiglitazone was associated with the FDA warning, removal from the VANF, female sex, black race, Hispanic ethnicity, comorbidity, A1C greater than 9%, and use of rosiglitazone as first- or second-line therapy. Among patients who did and did not receive a replacement medication, the mean changes in A1C from baseline were 0.12% and 0.46%, respectively. For those who continued rosiglitazone, the mean change in A1C was -0.02%.

CONCLUSION: The rosiglitazone discontinuation rate increased following the FDA warning and increased further following removal of rosiglitazone from the VANF. Glucose control may have declined among those who discontinued rosiglitazone.

BACKGROUND: Clinical validation studies of the Healthcare Effectiveness Data and Information Set (HEDIS®) measures of inappropriate prescribing in the elderly are limited.

OBJECTIVES: The objective of this study was to examine associations of new exposure to high-risk medication in the elderly (HRME) and drug-disease interaction (Rx-DIS) with mortality, hospital admission, and emergency care.

METHODS: A retrospective database study was conducted examining new use of HRME and Rx-DIS in fiscal year 2006 (Oct 2005-Sep 2006; FY06), with index date being the date of first HRME/Rx-DIS exposure, or first day of FY07 if no HRME/Rx-DIS exposure. Outcomes were assessed 1 year after the index date. The participants were veterans who were ≥65 years old in FY06 and received Veterans Health Administration (VA) care in FY05-06. A history of falls/hip fracture, chronic renal failure, and/or dementia per diagnosis codes defined the Rx-DIS subsample. The variables included a number of new unique HRME drug exposures and new unique Rx-DIS drug exposure (0, 1, >1) in FY06, and outcomes (i.e., 1-year mortality, hospital admission, and emergency care) up to 1 year after exposure. Descriptive statistics summarized variables for the overall HRME cohort and the Rx-DIS subset. Multivariable statistical analyses using generalized estimating equations (GEE) models with a logit link accounted for nesting of patients within facilities. For these latter analyses, we controlled for demographic characteristics, chronic disease states, and indicators of disease burden the previous year (e.g., number of prescriptions, emergency/hospital care).

RESULTS: Among the 1,807,404 veterans who met inclusion criteria, 5.2 % had new HRME exposure. Of the 256,388 in the Rx-DIS cohort, 3.6 % had new Rx-DIS exposure. Multivariable analyses found that HRME was significantly associated with mortality [1: adjusted odds ratio (AOR) = 1.62, 95 % CI 1.56-1.68; >1: AOR = 1.80, 95 % CI 1.45-2.23], hospital admission (1: AOR = 2.31, 95 % CI 2.22-2.40; >1: AOR = 3.44, 95 % CI 3.06-3.87), and emergency care (1: AOR = 2.59, 95 % CI 2.49-2.70; >1: AOR = 4.18, 95 % CI 3.71-4.71). Rx-DIS exposure was significantly associated with mortality (1: AOR = 1.60, 95 % CI 1.51-1.71; >1: AOR = 2.00, 95 % CI 1.38-2.91), hospital admission for one exposure (1: AOR = 1.12, 95 % CI 1.03-1.27; >1: AOR = 1.18, 95 % CI 0.71-1.95), and emergency care for two or more exposures (1: AOR = 1.06, 95 % CI 0.97-1.15; >1: AOR = 2.0, 95 % CI 1.35-3.10).

CONCLUSIONS: Analyses support the link between HRME/Rx-DIS exposure and clinically significant outcomes in older veterans. Now is the time to begin incorporating input from both patients who receive these medications and providers who prescribe to develop approaches to reduce exposure to these agents.

OBJECTIVE: The objective of the study was to assess the adherence to hormonal contraception (pill, patch, ring, or injectable) among women veterans and examine the relationships between race/ethnicity and the months of contraceptive supply dispensed with contraceptive adherence.

STUDY DESIGN: We conducted a retrospective analysis of the Department of Veterans Affairs (VA) national databases to examine the adherence to hormonal contraception over 12 months among women aged 18-45 years who had hormonal contraceptive coverage during the first week of fiscal year 2008. We examined several adherence indicators including gaps between refills and months of contraceptive coverage. Descriptive statistics and multivariable models were used to examine the associations between race/ethnicity and contraceptive supply dispensed with adherence.

RESULTS: Our cohort included 6946 women: 47% were white, 6% were Hispanic, 22% were black, and 25% were other race or had missing race information. Most women (83%) received a 3 month supply of contraception at each fill. More than 64% of women had at least 1 gap in coverage of 7 days or longer. Only 22% of women received a full 12 months of contraception without any gaps (perfect adherence). Compared with whites, Hispanics were significantly more likely to experience gaps (64% vs 70%; P = .02), and Hispanics and blacks received fewer months of contraceptive coverage (9.3 vs 8.9 and 9.0, P < .001). Compared with women receiving 3 month supplies, those receiving 1 month supplies had a higher likelihood of a gap (63% vs 72%, P < .001), fewer months of coverage (9.3 vs 6.9, P < .001), and a lower likelihood of perfect adherence (22% vs 11%, P < .001).

CONCLUSION: Adherence to hormonal contraception among women veterans is poor. Efforts to improve contraceptive adherence and lower risk of unintended pregnancy are needed; dispensing more months of supply for hormonal contraception may be a promising strategy.

PURPOSE: A Web-based application for coordinating medication-use evaluation (MUE) initiatives within the Veterans Affairs (VA) health care system is described.

SUMMARY: The MUE Tracker (MUET) software program was created to improve VA's ability to conduct national medication-related interventions throughout its network of 147 medical centers. MUET initiatives are centrally coordinated by the VA Center for Medication Safety (VAMedSAFE), which monitors the agency's integrated databases for indications of suboptimal prescribing or drug therapy monitoring and adverse treatment outcomes. When a pharmacovigilance signal is detected, VAMedSAFE identifies "trigger groups" of at-risk veterans and uploads patient lists to the secure MUET application, where locally designated personnel (typically pharmacists) can access and use the data to target risk-reduction efforts. Local data on patient-specific interventions are stored in a centralized database and regularly updated to enable tracking and reporting for surveillance and quality-improvement purposes; aggregated data can be further analyzed for provider education and benchmarking. In a three-year pilot project, the MUET program was found effective in promoting improved prescribing of erythropoiesis-stimulating agents (ESAs) and enhanced laboratory monitoring of ESA-treated patients in all specified trigger groups. The MUET initiative has since been expanded to target other high-risk drugs, and efforts are underway to refine the tool for broader utility.

CONCLUSION: The MUET application has enabled the increased standardization of medication safety initiatives across the VA system and may serve as a useful model for the development of pharmacovigilance tools by other large integrated health care systems.