Understanding primary nonadherence to medications and its associated healthcare outcomes: A retrospective analysis of electronic medical records in an integrated healthcare setting.
详细信息
文摘
OBJECTIVES: Phase I objectives were to measure the extent of primary nonadherence PNA) in an integrated healthcare system and identify factors significantly associated with PNA to chronic and acute medications. The objective of phase II was to examine healthcare outcomes associated with PNA to cholesterol and diabetes medications. METHODS: This retrospective cohort study was conducted at Kaiser Permanente Southern California using data from its integrated electronic medical record system. All new prescriptions for 10 therapeutic drug groups antiinfectives,analgesics,migraine medications,antidiabetics,osteoporosis medications,cardiovascular agents,antihyperlipidemics,antiasthmatics,antidepressants,and anticoagulants) written for patients who satisfied membership and other criteria were included in phase I. PNA rates were calculated by therapeutic drug group as the number of prescriptions not filled within 14 days over the total number of prescriptions. A multivariate logistic regression was used to identify patient,prescriber,and prescription factors significantly associated with PNA. Patients who filled an antidiabetic or antihyperlipidemic were eligible for inclusion in Phase II if they were at least 18 years of age and satisfied additional inclusion and exclusion criteria. A more conservative definition of PNA was applied in phase II as the failure to fill within 180 days. Descriptive statistics were used to summarize baseline differences among patient groups and to describe secondary nonadherence and unadjusted rates of event outcomes. The paired t-test and a multivariate difference-in-difference model were used to evaluate changes in lab values. A multivariate Cox Proportional Hazards model was used to measure the effect of PNA and secondary nonadherence on event outcomes of new disease complications,ER visits,and hospitalization. RESULTS: In phase I,a total of 569,095 new prescriptions for 398,025 patients satisfied study criteria. The overall PNA rate was 9.8% but individual PNA rates varied widely by therapeutic drug group. The effects of patient race,baseline comorbidities,prior fills,and being treatment-naive on PNA were consistent across drug groups. The strongest predictors were prior fills and being treatment-naive. Some factors mostly prescriber and prescription factors) depended on whether the treatment was acute or chronic. Phase II included 21,291 cholesterol patients and 10,137 diabetes patients. PNA rates were12.0% and 5.1% for cholesterol and diabetes,respectively. Secondary nonadherence was relatively worse in cholesterol than diabetes with about 20% of primary adherent patients only filling the first prescription. Adherent patients were on average older and sicker than nonadherent patients. Patients who were nonadherent to their cholesterol or diabetes medication were significantly associated with worse health outcomes in terms of lab values and ER visits. In cholesterol,the risk for ER visits was significantly higher in secondary nonadherent patients HR＝1.24,95% CI: 1.14,1.34) and the risk of hospitalization was significantly higher in primary nonadherent patients with prior cholesterol medication use HR＝1.69,95% CI: 1.19,2.41) in comparison to adherent patients. For diabetes,secondary nonadherent patients were associated with significantly increased risk for ER visits HR＝1.30,95% CI: 1.11,1.51). CONCLUSIONS: The results of this study help clinicians and healthcare decision-makers to understand the extent of PNA and how PNA varies by therapeutic drug group in an integrated healthcare setting. Most factors of PNA were consistent across drug groups,but some factors depended on whether the treatment was acute or chronic. In addition,these results can facilitate the decision-making process in designing and implementing cost-effective patient interventions to improve overall adherence to cholesterol and diabetes medications. Future research should examine the effect of PNA on longer-term health outcomes such as mortality.