Multivariate analysis of the factors affecting duration of acute inpatient rehabilitation after hip and knee arthroplasty.

Abstract

To determine which factors predict length of stay at a rehabilitation hospital for total hip and knee arthroplasty patients. Length of stay at an acute rehabilitation hospital was analyzed with respect to demographic, surgical, and comorbidity factors using univariate and multivariate statistical techniques. Data were retrospectively collected of 808 hip and knee arthroplasty rehabilitation patients from years 2000 and 2001 at a large, urban, university-affiliated acute rehabilitation hospital. Demographic factors included age, insurance, marital status, sex, race, and body mass index. Surgical factors included knee/hip arthroplasty, unilateral/bilateral arthroplasty, and indication for surgery. Also, the 22 most prevalent comorbid illnesses for the patient group were analyzed for their association with length of stay. Significant factors for predicting length of stay were indication for surgery, number of comorbid illnesses, unmarried marital status, and black race. Also, advanced age and male sex approached significance. Length of stay in rehabilitation for total hip arthroplasty and total knee arthroplasty can be statistically modeled using various factors. These findings will help clinicians address prolonged length of stay and resource allocation issues.