Identifying characteristics of patients requiring proactive pharmaceutical interventions in the recovery period and assessing the effect of rehabilitation and drugs: a retrospective study.

Background: The aging of the population in many countries has made rehabilitation an essential part of improving the quality of life of older individuals. The risk factors for falls during rehabilitation include a history of falls, gait disturbances, dizziness, and medication use. Although numerous studies have explored various fall prevention measures, stratified or detailed analyses of the relationship between the activities of daily living (ADL) and drugs have not been performed. This study aimed to examine the effect of drugs on ADLs in patients undergoing rehabilitation and explored the factors affecting patients' ADLs identify the characteristics of patients requiring proactive pharmaceutical interventions.

Methods: Participants aged ≥ 20 years admitted to the Kaifukuki Rehabilitation Ward at Hyogo Medical University Sasayama Medical Center underwent functional independence measure (FIM) assessments and were evaluated for medication use. The complexity of the medication regimen was assessed using the Japanese version of the medication regimen complexity index (MRCI-J) based on prescription data. Hierarchical cluster analysis was used to classify the participants based on their FIM scores.

Results: No correlation was found between FIM motor gain and MRCI-J differences among all participants. Hierarchical cluster analysis was used to classify participants into four groups based on their FIM motor and cognitive scores at admission and discharge. Decision tree analysis was performed using the four identified groups as objective variables and yielded eight nodes. The algorithm included length of hospital stay, sex, age, units of rehabilitation performed, and the MRCI-J score. The group with a hospital stay < 74 days, aged < 90 years, and who underwent > 77 units of rehabilitation during the study period was further divided into fourth tiers based on the MRCI-J scores, with the non-increased MRCI-J group assigned as Node 7 and the increased MRCI-J group as Node 8.

Conclusions: No relationship was found between ADLs and prescribed drugs in the overall participant population. In participants from Nodes 7 and 8, who had a relatively short length of hospital stay and were discharged with preserved physical and cognitive functions, prescription changes appeared to have some effects on patient's ADLs.