———————————————————Table 1 near here—————————————————-

Medication changes

Total population

From baseline to follow-up, 255 medication changes were effectuated, of which 83% (n=212) were deprescribing, 15% (n=38) were new prescriptions, and 2% (n=5) involved other medication changes (i.e., dose increase or change in dosing interval). The median ( [IQR]) medication changes per patient was 2 [2-4]. Medication changes were maintained for 88.5% (n=77) of patients at follow-up and partly maintained for 10.3% (n=9) (e.g., if several changes were made for one patient, but only some of these changes were maintained). For one patient, the suggested medication changes were not implemented for unknown reasons.

Care home residents

In the care home residents, 93 medication changes were made, including 83% deprescribing (n=77), 15% new prescriptions (n=14), and 2% other medication changes (n=2) (dose increase). Medication changes were maintained for 92.5% (n=25) of patients at follow-up and partly maintained for 7.5% (n=2). For one care home resident, treatment with an antidepressant was stopped as part of the intervention. However, as this resulted in reduced functional level, the antidepressant was re-prescribed, and the patient returned to a stable functional level.

Community-dwelling patients with chronic disease

In the community-dwelling patients, 162 medication changes were made, hereof 83% deprescribing (n=135), 15% new prescription (n=24), and 2% other medication changes (n=3) (dose increase or change in dosing interval). Medication changes were maintained for 87% (n=52) of patients at follow-up and partly maintained for 12% (n=7). For one patient, the changes were not executed for unknown reasons.

Health-related outcomes

Health-related outcomes for the total study population and the two sub-groups are presented in Table 2. In the total study population, mean self-reported health status significantly increased, while the proportions of patients with general condition rated as “average or above” and with functional level rated as “without any disability” remained stable. Overall, the subgroup analyses showed similar trends in self-reported health status and functional level, although statistical significance was only reached for self-reported health status in community-dwelling patients due to the small sample size of the sub-groups. Noteworthy, among care home residents, the proportion of patients with general condition rated as ”average or above” increase non-significantly with 18.5 percentage points, in contrast to no difference among community-dwelling patients.