Longitudinal evaluation of medication underuse in older outpatients and its association with quality of life.

PURPOSE: We investigated the factors promoting prescribing omissions (medication underuse) in long-term medical care and the impact of withholding indicated medications on quality of life. METHODS: In a population-based cohort study of older ambulatory patients (ESTHER), we collected data with sequential questionnaires from participants and from their GPs. Concurrently, in two consecutive home visits, trained study physicians performed comprehensive geriatric assessments and recorded all medicines currently taken. Each patient's medication was screened for underuse using the START-2 criteria. RESULTS: Medication underuse (absence of ≥1 indicated medication) was present in 70.3 and 73.2 % of 989 participants at two consecutive home visit assessments, respectively. Following variable selection accounting for subject-specific heterogeneity over time, multivariate results revealed that more drugs (odds ratio with 95 % confidence intervals: 0.83 [0.78;0.87] per drug) and better cognitive status (0.93 [0.87;0.99] per point on the MMSE scale) were preventive factors, while worse self-reported health status (1.33 [1.05;1.67] per point on an 5-point scale) and increasing frequency of GP consultations (1.07 [1.00;1.15] per visit within the preceding 3 months) were positively associated with medication underuse. An increase in omitted medications over time was associated with worse quality of life as determined on the EuroQuol EQ-Vas and EQ-5D scales. CONCLUSION: In addition to general and physician-related factors, also patient-related aspects, such as individual health appraisal, were associated with medication underuse. Because withholding indicated drugs was associated with substantially reduced quality of life, controlled intervention studies are necessary to confirm the notion that pharmacological appropriateness improves personal wellbeing.