Abstract TP95: Concurrent Recovery of Language and Motor Function during Inpatient Stroke Rehabilitation

Background: Stroke impairments are variable and multidimensional, encompassing language, motor, sensory, and cognition. Mechanisms underlying recovery across motor and language domains may be additive or competitive, but studies are lacking. We aimed to examine differences in change of language function in two patient groups: those who did and those who did not meet the minimal clinically important difference (MCID) in motor function after inpatient rehabilitation. We also evaluated differences in discharge disposition. Methods: This retrospective cross-sectional study used data from the eRehabData repository. Patients who discharged from 4 Charlotte, NC area rehab hospitals between 2016 and 2019, received therapy for stroke, and had an admission Functional Independence Measure (FIM) score of 5 or lower on at least 1 of 2 language-related FIM items [comprehension or expression; N=2774; 46.6% female, mean age 67(SD=14)] were included. Items on the FIM are scored on a 7-point scale (1=total assist, 7=complete independence). The primary outcome measure was the change in language FIM subscore. Patients were grouped by whether they met the MCID for the motor FIM subtotal (which combines the scores of 13 motor FIM items), defined as a 17-point change. The secondary outcome was discharge disposition with 67% discharged home (N=1863). We used a generalized linear model to examine the relationship between magnitude of language score change and motor FIM status (i.e., meeting the MCID or not), and we used logistic regression to examine factors predictive of discharge home. All models were adjusted for motor FIM status, length of stay, therapy hours, race, age, and sex. Results: Approximately 60% of patients (N=1654) either met or exceeded the motor MCID. Patients who did not meet the motor FIM MCID were less likely to show larger change scores in expression (OR=0.603, 95% CI=0.551-0.658, p<.001) and comprehension (OR=0.657, 95% CI=0.601-0.719, p<.001). Patients who met the motor FIM MCID were more likely to discharge home (OR=6.605, 95% CI=5.495-7.940, p<.001). Conclusion: Patients who met the motor FIM MCID were more likely to have larger comprehension and expression change scores and discharge home, even after accounting for other demographic and clinical factors. If underlying mechanisms are indeed additive, there is much potential to build upon this natural synergy to further improve rehabilitation outcomes for stroke survivors.