Results

The study included 196 patients randomized to medication review and 212 to usual care. The included patients had a median age of 81 years (IQR 75 to 85 years), were prescribed a median of 12 medicines (IQR 10 to 14) at baseline, and 71% were females. Detailed patient characteristics and patient flowchart have previously been reported.3 Medicines prescribed at baseline are listed in Supporting Information Table S1. As reported in the primary publication,3 medication review reduced the number of prescribed medicines by 18% after the first visit, by 16% after 4 months, and by 11% after 13 months compared with 5%, 5% and 2% in the usual care group. The medicine changes, and their corresponding persistence, that led to these differences in the number of prescribed medicines are described in detail in Table 1. The number of overprescribed and underprescribed medicines as a function of the number of baseline medicines is visualized in Figure \ref{633516}. The primary reasons for discontinuations in the medication review group are listed in Table 2. The medicines that were more often discontinued in the medication review group compared with usual care are listed in Table 3. The medicines that were most often restarted after being discontinued in the medication review group are listed in Table 4. The proportions of prescribed medicines in pharmacological subgroups in the two groups at the different time points are shown in Figure \ref{222994}. The numbers from the plots in Figure \ref{222994} are listed in Supporting Information Table S2. The factors that predicted the number of overprescribed medicines in the two models are listed in Table 5. The effect of the patient-related factors on predicted number of overprescribed medicines is visualized in Figure \ref{907792}. For reference, the number of overprescribed medicines for patients referred from the geriatric department was a median (IQR) of 3 (2 to 5) compared with 4 (2 to 6) for patients referred from the GP, and for patients not motivated for medicine changes 3 (1 to 4) compared with 4 (2 to 6) for patients motivated for medicine changes.