Trade-offs between effectiveness and efficiency in stroke rehabilitation

Background Most stroke research has studied rehabilitation effectiveness and rehabilitation efficiency separately and not investigated the potential trade-offs between these two indices of rehabilitation. Aims To determine whether there is a trade-off between independent factors of rehabilitation effectiveness and rehabilitation efficiency. Methods Using a retrospective cohort study design, we studied all stroke patients (n=2810) from two sub-acute rehabilitation hospitals from 1996 to 2005, representing 87.5% of national bed-years during the same period. Results Independent predictors of poorer rehabilitation effectiveness and log rehabilitation efficiency were older age race-ethnicity caregiver availability ischemic stroke longer time to admission dementia admission Barthel Index score, and length of stay. Rehabilitation effectiveness was lower in females, and the gender differences were significantly lower in those aged =70 years (beta -4.7 (95% confidence interval -7.4 to -2.0)). There were trade-offs between effectiveness and efficiency with respect to admission Barthel Index score and length of stay. An increase of 10 in admission Barthel Index score predicted an increase of 3.6% (95% confidence interval 3.24.0) in effectiveness but a decrease of 0.04 (95% confidence interval -0.05 to -0.02) in log efficiency (a reduction of efficiency by 1.0 per 30 days). An increase in log length of stay by 1 (length of stay of 2.7 days) predicted an increase of 8.0% (95% confidence interval 5.710.3) in effectiveness but a decrease of 0.82 (95% confidence interval -0.90 to -0.74) in log efficiency (equivalent to a reduction in efficiency by 2.3 per 30 days). For optimal rehabilitation effectiveness and rehabilitation efficiency, the admission Barthel Index score was 3062 and length of stay was 3741 days. Conclusions There are trade-offs between effectiveness and efficiency during inpatient sub-acute stroke rehabilitation with respect to admission functional status and length of stay.