Responsiveness of the SF-36 and FIM in lower extremity amputees undergoing a multidisciplinary inpatient rehabilitation

Background: SF-36 and FIM are common outcome measures in in-patient amputee rehabilitation. in order to conduct high-quality research into effectiveness of rehabilitation following lower extremity amputation, researchers need to be confident that their selected outcome instruments will be sufficiently sensitive to change over time. The aim of the present study was to evaluate the short-term responsiveness of SF-36 and FIM in lower extremity amputees undergoing in-patient rehabilitation. Methods: Sixty-eight of seventy-four consecutive lower extremity amputees underwent a multidisciplinary inpatient rehabilitation program. SF-36 for evaluating of the health-related quality-of-life and FIM as measure of the functional independence were administered on admission and at discharge. Eight subscales and both summary scores of SF-36 were calculated norm-based and adjusted for age and gender. Six FIM dimensions were linear transformed into a 0-100 scale and aggregated to two domains (motor and cognitive) as well as to a summary score. Sensitivity was measured by the standardized effect size (SES) and standardized response mean (SRM) for all dimensions. Additionally floor and ceiling effects were calculated for both measures. Results: Moderate to high responsiveness was found for all SF-36 subscales reaching SES between 2.22 for physical functioning and 0.55 for general health. SRM was, on average, lower for nearly all dimensions. The component scores were not more sensitive as the most subscales. The FIM showed high sensitivity in all dimensions, both motor and cognitive domains as well as in the total score. There were substantial floor and ceiling effects in some subscales of SF-36 and FIM, but not on SF-36 summary scores or total FIM score. Discussion: Short-term responsiveness of both measurements used in the current study have not been reported conclusively yet in amputee population. The changes in the physical scales of the SF-36 were comparable but these in the mental scales and in terms of functional independence were lower then reported elsewhere. Co-morbidity seems to have a major impact on the results. Floor and ceiling effects may have biased the calculated results limiting their generalisability. Conclusions: There is a poor evidence in terms of responsiveness of widely used outcome instruments in amputee rehabilitation, despite the importance of this problem. Both SF-36 and FIM were found sensitive enough to detect longitudinal changes in health-related quality-of-life and functional independence in the investigated sample. The results should be interpreted with caution due to possible biasing. SF-36 and FIM could be recommended in combination for further research needed to evaluate the effectiveness of the rehabilitation after lower extremity amputation.