Predicting length of stay in patients admitted to stroke rehabilitation with high levels of functional independence

Abstract

Purpose. In Ontario, stroke patients with relatively mild functional deficits are admitted to inpatient rehabilitation programmes. Despite apparently minor impairments, many of these patients remain in rehabilitation for prolonged periods of time. The objective of the present study is to identify variables that predict length of stay (LOS) within this population of high functioning stroke patients.

Methods. Stroke patients with Functional Independence Measure (FIM™) scores ≥100 admitted to a specialised inpatient rehabilitation programme in Ontario, Canada, from April 2005 to March 2008 were included in this study. Data from the National Rehabilitation Reporting System and chart review were used. Associations with LOS were examined among 25 potential predictor variables using bivariate correlations. Significantly correlated (p <0.002) variables were entered into a multiple linear regression.

Results. Twenty-four percent (n  ==  134) of patients admitted during the study period were identified as presenting with mild functional deficits. These patients had a mean admission age of 63.5 (SD 14.6) years, a mean admission FIM score of 109.9 (SD 6.6) and a mean LOS of 22.3 (SD 10.5) days. FIM motor and cognitive subscale scores, Mini Mental State Exam scores and five single items assessing orientation, financial independence, and verbal, written and auditory communication ability were significantly correlated with LOS. The regression model, which included all eight variables, explained 60%% of the variance associated with LOS (r  ==  0.80).

Conclusions. For patients admitted to stroke rehabilitation programmes with mild functional impairment, 60%% of the variance in LOS can be explained by a small number of variables indicative of deficits in the areas of motor function, cognition and communication.

Keywords:

• Stroke
• mild stroke
• length of stay

Declaration of interest:

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.