Abstract W P131: Efficacy of Self-exercise on General Wards to Improve Activities of Daily Living: A Propensity Score Approach

Abstract

Objective: This study aimed to determine whether self-exercise (SE) correlates with an improvement in activities of daily living during hospitalization. Methods: The subjects were 1490 patients on general ward in 7centres. Items surveyed were age, sex, stroke type, length of stay, post-onset day of admission, motor functional independence measure (mFIM) and cognitive FIM(cFIM) at admission and at discharge, modified Rankin Scale (mRS) score before onset and at admission, Glasgow Coma Scale score at admission, presence/absence of SE, at least a 30% SE implementation rate(30%SE), and the presence/absence of hospital ward staff training(ST). In addition, the difference between mFIM at admission and at discharge (mFIM gain) was calculated. For analysis, a correlation between m FIM gain and the presence/absence of SE implementation was examined. SE implementation probability was subjected to stratification analysis using the propensity score (PS) and adjusted for confounding factors. Results: A significant difference in all items was noted based on the presence/absence of SE implementation. Among these, mFIM at discharge, cFIM at discharge, mFIM gain, and general wards having at least a 30% SE had a moderate to strong effect (0.53, 0.47, 0.49, and 0.44). PS was calculated for cFIM at admission, 30% SE , sex, age, mRS score at onset, and the presence/absence of ST .Stratification analysis revealed that the common odds ratio according to the Mantel-Haenszel test was 2.1, and therefore the probability for mFIM gain to be at least 22 points as a result of SE implementation was at least 2.1 times. Conclusion: This study used data from multicentres and therefore has excellent external validity. This study also determined the effect of SE implementation using the PS. Furthermore, the results suggest that SE implementation contributes to improvement in activities of daily living.