Predicting outcomes after hip fracture repair.

Abstract

To compare the activities of daily living before and after hip fracture and construct a statistical model for discharge destination and independent walking. The classification accuracy of the model was determined from an independent sample. Prospective study: FIM prefracture, at discharge, and at 6-mo follow-up were obtained from 63 patients who underwent operations for acute hip fractures. A statistical model for discharge destination and independent walking was made and classification accuracy was checked using 78 independent samples. The motor FIM scores at prefracture decreased significantly at discharge (P < 0.0001) and at 6-mo follow-up (P < 0.0001), but at 6-mo follow-up, they had increased significantly compared with those at discharge (P = 0.0103). A mobility subscale was used to predict discharge destination, and mobility and social cognition subscales were related to independent walking. The predictive accuracy was 87%. Motor FIM scores increase for at least 6 mos after hip fracture, and discharge destination and independent walking were highly predictable from FIM mobility and social cognition subscales.