Measuring and predicting early functioning post-stroke.

摘要

Functioning, the most important outcome of stroke interventions, is complex to characterize. Stroke survivors measure functioning compared to what they did in their prestroke lives; hence equating functioning with recovery. To optimize the recovery of functioning, research suggests that rehabilitation interventions should start early post-stroke. To quantify changes in function owing to different interventions, the average value on an index comprised of multiple items related to function is compared or the proportions of people categorized into different functional levels are compared. Currently, there is no agreed upon method of quantifying improvements in functioning and using multiple indices is problematic. This thesis examined combining a method of quantifying behaviours, Rasch analysis that produces measures with interval properties from ordinal observations, with the component of the International Classification of Functioning (ICF), to conceptualize, define, and quantify functioning in a single measure. The ensuing prototype measure was limited in scope. Therefore, using data from a longitudinal prognostic study involving people with acute stroke assessed at three days and followed up at three months, a Functioning measure at three months, the F3m, was developed. The F3m is a valid and reliable measure that amalgamates tests where performance is observed and self-report questionnaires where people rate their difficulties in performing physical activities. The F3m covers all ICF components and can be used to quantify recovery at three months. Interventions to improve early functioning post-stroke must impact favourably on the factors that affect early recovery. The most influential factor related to recovery, to date, has been early functioning. As a measure of such early functioning did not exist, a measure of functioning at three days, the FM, was constructed in a manner similar to the F3m. Univariate and multivariate analyses were then used to identify strong early factors collected 24-72 hours post-stroke, and link the factors to function at 3 months. A seven-variable predictive model of functioning was derived. The most important influential predictor of functioning in the model, the comprehensive FM measure, can now be used to evaluate and develop early interventions to enhance functioning, and to act as a covariate explaining the recovery of functioning.