Domenico Scrutinio1, Pietro Guida2, Bernardo Lanzillo3, Chiara Ferretti4, Anna Loverre2, Nicola Montrone2, Simona Spaccavento2. 1. Istituti Clinici Scientifici Maugeri-SPA SB. I.R.C.C.S., Cassano Murge, Italy. Electronic address: domenico.scrutinio@icsmaugeri.it. 2. Istituti Clinici Scientifici Maugeri-SPA SB. I.R.C.C.S., Cassano Murge, Italy. 3. Istituti Clinici Scientifici Maugeri-SPA SB. I.R.C.C.S., Telese Terme, Italy. 4. Istituti Clinici Scientifici Maugeri-SPA SB. I.R.C.C.S., Montescano, Italy.
Abstract
OBJECTIVE: To characterize rehabilitation outcomes of patients with severe poststroke motor impairment (MI) and develop a predictive model for treatment failure. DESIGN: Retrospective cohort study. Correlates of treatment failure, defined as the persistence of severe MI after rehabilitation, were identified using logistic regression analysis. Then, an integer-based scoring rule was developed from the logistic model. SETTING: Three specialized inpatient rehabilitation facilities. PARTICIPANTS: Patients (N=1265) classified as case-mix groups (CMGs) 0108, 0109, and 0110 of the Medicare classification system. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Change in the severity of MI, as assessed by the FIM, from admission to discharge. RESULTS: Median FIM-motor (FIM-M) score increased from 17 (interquartile range [IQR] 14-23) to 38 (IQR, 25-55) points. Median proportional recovery, as expressed by FIM-M effectiveness, was 26% (IQR, 12-47). Median FIM-M change was 18 (IQR, 9-34) points. About 38.5% patients achieved the minimal clinically important difference. Eighteen point six percent and 32.0% of the patients recovered to a stage of either mild (FIM-M ≥62) or moderate (FIM-M 38-61) MI, respectively. All between-CMG differences were statistically significant. Outcomes have also been analyzed according to classification systems used in Australia and Canada. The scoring rule had an area under the curve of 0.833 (95% confidence interval, 0.808-0.858). Decision curve analysis displayed large net benefit of using the risk score compared with the treat all strategy. CONCLUSIONS: This study provides a snapshot of rehabilitation outcomes in a large cohort of patients with severe poststroke MI, thus filling a gap in knowledge. The scoring rule accurately identified the patients at risk for treatment failure.
OBJECTIVE: To characterize rehabilitation outcomes of patients with severe poststroke motor impairment (MI) and develop a predictive model for treatment failure. DESIGN: Retrospective cohort study. Correlates of treatment failure, defined as the persistence of severe MI after rehabilitation, were identified using logistic regression analysis. Then, an integer-based scoring rule was developed from the logistic model. SETTING: Three specialized inpatient rehabilitation facilities. PARTICIPANTS: Patients (N=1265) classified as case-mix groups (CMGs) 0108, 0109, and 0110 of the Medicare classification system. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Change in the severity of MI, as assessed by the FIM, from admission to discharge. RESULTS: Median FIM-motor (FIM-M) score increased from 17 (interquartile range [IQR] 14-23) to 38 (IQR, 25-55) points. Median proportional recovery, as expressed by FIM-M effectiveness, was 26% (IQR, 12-47). Median FIM-M change was 18 (IQR, 9-34) points. About 38.5% patients achieved the minimal clinically important difference. Eighteen point six percent and 32.0% of the patients recovered to a stage of either mild (FIM-M ≥62) or moderate (FIM-M 38-61) MI, respectively. All between-CMG differences were statistically significant. Outcomes have also been analyzed according to classification systems used in Australia and Canada. The scoring rule had an area under the curve of 0.833 (95% confidence interval, 0.808-0.858). Decision curve analysis displayed large net benefit of using the risk score compared with the treat all strategy. CONCLUSIONS: This study provides a snapshot of rehabilitation outcomes in a large cohort of patients with severe poststroke MI, thus filling a gap in knowledge. The scoring rule accurately identified the patients at risk for treatment failure.