Hospital readmission in persons with stroke following postacute inpatient rehabilitation

Abstract

Background: Readmission is an important quality indicator following acute care hospitalization. We examined factors associated with hospital readmission in persons with stroke following postacute inpatient rehabilitation.

Methods: Prospective cohort study including 674 persons with stroke who received rehabilitation at 11 facilities located in eight states and the District of Columbia. Measures included hospital readmission within 3 months of discharge, sociodemographic characteristics, length of stay, primary payment source, comorbidities, stroke type, standardized assessments of motor and cognitive function, depressive symptoms, and social support.

Results: Mean age was 71.5 years (SD = 10.5). Twenty-five percent of patients reported high depressive symptoms. Overall, 18% (n = 122) of the sample was rehospitalized. Univariate analyses showed that people who were rehospitalized were more likely (p < .05) to be non-Hispanic white, married, demonstrate less functional independence at discharge, experience longer lengths of stay in rehabilitation, and report more depressive symptoms and lower social support. In the fully adjusted multivariable hierarchical generalized linear model, motor functional status (OR = 0.98, 95% CI 0.96-0.99), depressive symptoms (OR = 1.80, 95% CI 1.06-3.05), and social support (OR = 2.28, 95% CI 1.29-4.03) remained statistically significant. In addition, a minority-by-depressive symptoms interaction term also reached statistical significance.

Conclusion: Functional status, depressive symptoms, and social support were important predictors of hospital readmission. These variables are not included in most administrative data sets. Future research to develop useful risk-adjustment models for rehospitalization following postacute inpatient rehabilitation services should include large diverse samples and explore practical sources for additional meaningful information.

Figures

Figure 1.

Unadjusted rehospitalization rates by minority status and Center for Epidemiologic Studies-Depression scores.

Figure 2.

Predicted rehospitalization rates by minority status and Center for Epidemiologic Studies-Depression scores based on the fully adjusted model shown in Table 2.