Association Between Potentially Inappropriate Medications and 30-Day Post–Hospital Discharge Outcomes in US Veterans

BACKGROUND: Potentially inappropriate medication (PIMs) use is common in older inpatients and it may lead to increased risk of adverse drug events. OBJECTIVES: To examine prevalence of PIMs at hospital discharge and its contribution to health care utilization and mortality within 30-days of hospital discharge. METHODS: This was a prospective cohort of 117 570 veterans aged ≥65 years and hospitalized in 2013. PIMs at discharge were categorized into central nervous system acting (CNS) and non-CNS. Outcomes within 30-days of hospital discharge were: (1) time to first acute care hospital readmission, and all-cause mortality, (2) an emergency room visit, and (3) ≥3 primary care clinic visits. RESULTS: The cohort's mean age was 74.3 years (SD 8.1), with 51.3% exposed to CNS and 62.8% to non-CNS PIMs. Use of CNS and non-CNS PIMs, respectively, was associated with a reduced risk of readmission, with an adjusted hazard ratio (aHR) of 0.93 (95% CI = 0.89-0.96) for ≥2 (vs 0) CNS PIMs and an aHR of 0.85 (95% CI = 0.82-0.88) for ≥2 (vs 0) non-CNS PIMs. Use of CNS PIMs (≥2 vs 0) was associated with increased risk of mortality (aHR = 1.37 [95% CI = 1.25-1.51]), whereas non-CNS PIMs use was associated with a reduced risk of mortality (aHR = 0.75 [95% CI = 0.69-0.82]). CONCLUSION AND RELEVANCE: PIMs were highly common in this veteran cohort, and the association with outcomes differed by PIMs. Thus, it is important to consider whether PIMs are CNS acting to optimize short-term posthospitalization outcomes.