A deprescribing intervention for hospitalized older adults reduces polypharmacy (Shed-MEDS)

Référence

Vasilevskis EE, Shah AS, Hollingsworth EK, et al. Deprescribing medications among older adults from end of hospitalization through postacute care. JAMA Intern Med 2023;183(3):223-231.

Sommaire

Hospitalized patients 50 years and older who had at least 5 pre-hospital medications and were being discharged to a post-acute care (PAC) facility were randomized to receive either the Shed-MEDS intervention or usual care. Of 1353 eligible patients, 911 declined participation and 70 were not approached, resulting in 186 patients in each study group. In the hospital, all patients underwent the Best Possible Medication History process, which involves a review of medical records, pharmacy refill history, and the controlled substance monitoring database, as well as a patient/surrogate interview, to obtain an accurate medication list. The intervention group also received 4 additional steps conducted by a research clinician (either a pharmacist or a nurse practitioner): (1) a review of medications for deprescribing, (2) a conversation with the patient/surrogate about deprescribing, (3) a discussion of deprescribing actions with outpatient prescribers and the inpatient team, and (4) a review of medications and deprescribing recommendations with the PAC facility upon transfer. After discharge, the research clinician called the PAC facility on a weekly basis to review medications and sent a reconciled medication list and ongoing deprescribing recommendations to outpatient prescribers. The primary outcome was the total medication count at hospital discharge, at PAC facility discharge, and at 90 days following PAC facility discharge. A total of 142 patients in each study group were included in the intention-to-treat analysis. The groups were similar at baseline: average age was 76 years, 62% were female, 84% were of White race and ethnicity, and the median number of prehospital medications was 16. Although the mean total medication count at hospital discharge was similar in the 2 groups, the intervention group had 14% fewer medications at PAC facility discharge (0.86; 95% CI 0.80 - 0.93) and 15% fewer medications at 90-day follow-up (0.85; 0.78 - 0.92). Vitamins, supplements, laxatives, and antihypertensives were the most frequently deprescribed classes of medications. There were no differences in overall adverse events, adverse drug events, or adverse drug withdrawal events between the 2 groups. Of note, given the large percentage of patients who declined participation in the trial, there was a possibility of enrollment bias. Those who enrolled may have been more inclined to deprescribe, although this behavior could be seen in both the intervention and control groups.