Predicting follow-up living setting in patients with stroke

Objective: To examine living setting at 3- to 6-month follow-up for inpatients with functional impairments discharged from medical rehabilitation. Design: Retrospective performed by using information from the Uniform Data System for Medical Rehabilitation (UDSMR) representing medical rehabilitation patients across the United States. Setting: National survey data. Participants: Information submitted in 1997 and 1998 to the UDSMR by 167 facilities from 40 states was examined. A total of 9587 patient records were included in the final sample. The mean age ± standard deviation was 70.2 ± 12.4 years. The sample included 51.6% women and was comprised of 77.5% non-Hispanic white patients, with an average length of stay of 22.3 ± 4.6 days. Interventions: Not applicable. Main Outcome Measure: Living setting (home vs not at home) at 3- to 6-month follow-up. Results: A discriminant function training model including 8 statistically significant variables correctly classified 85.1% of the patients (n=8149). The total FIM™ instrument score, patient age, function-related group, and marital status were found to be useful classification variables. Wilks λ for the model was.924 (X²=1031.49, P<.000). The area under the receiver operating characteristics curve was.85. Conclusion: The association among functional abilities, demographic characteristics, and follow-up living setting in patients with stroke is complex. Functional variables can be used to help predict follow-up living setting. These variables change based on patient severity level.