Restarting Medications After Deprescribing in Adults Discharged From Hospital to Skilled Nursing

Importance: Deprescribing interventions safely reduce polypharmacy in hospitalized older adults, but the durability of medication reductions and factors associated with medication restart are poorly understood. Objective: To characterize medication restart patterns after hospital-initiated deprescribing, identify factors associated with restart, and examine associations between restart and acute health care utilization. Design, Setting, and Participants: This cohort study analyzed data from 2 randomized trials: Best Possible Medication History, Evaluate, Deprescribing Recommendations, and Synthesis trial (conducted March 2016 to October 2020 at Vanderbilt University Medical Center and 22 affiliated postacute care facilities) and Veterans Affairs-Drug Reduction in Older Patients trial (conducted October 2019 to March 2023 at VA Tennessee Valley Healthcare System). Participants were aged 50 years or older, taking 5 or more prehospital medications, and discharged to a skilled nursing facility (SNF) for postacute care; analyses included all participants with 1 or more deprescribed medication. Data were analyzed from September 2024 to December 2025. Exposures: Trial interventions included comprehensive medication history, identification of deprescribing opportunities, participant or surrogate interview to assess agreement, and collaboration with clinicians. Patients in the control group received a comprehensive medication history followed by usual care, which included routine deprescribing by their practitioners. Main Outcomes and Measures: Medication restart (reinitiation or return to previous or higher dose) was determined from discharge documentation and structured patient or surrogate interviews at SNF discharge, 7 and 90 days after SNF discharge. The following were assessed: restart frequency, timing, and medication class; factors associated with restart; and associations between restart and emergency department visits and hospital readmissions. Results: Among 598 participants, the median (IQR) age was 74.0 (67.2-82.0) years, 354 (59.2%) were male, and 417 (69.7%) restarted 1 or more deprescribed medication. Of 8734 total deprescribed medications, 1385 (15.9%) were restarted across 1461 restart episodes; 330 (22.6%) occurred during the SNF stay, 423 (29.0%) between SNF discharge and 7-day follow-up, and 708 (48.5%) between 7- and 90-day follow-ups. Overall, approximately 84% of deprescribed medications remained discontinued at 90 days. Restart rates were similar between intervention and usual care groups. Higher health literacy and longer intervention duration were associated with lower restart risk, whereas higher baseline medications, more prescribers, and use of 1 to 2 (vs ≥3) pharmacies were associated with a higher risk. Restart during the SNF stay was associated with higher 90-day hospital readmission. Conclusions and Relevance: In this cohort study, approximately 1 in 6 deprescribed medications were restarted within 90 days, with nearly half occurring soon after SNF discharge. Patient factors and markers of care fragmentation were associated with restart, suggesting that enhancing transitional care and postdischarge support may improve the durability of hospital-initiated deprescribing.