Cost Utility Analysis of Multidisciplinary Postacute Care for Stroke: A Prospective Six-Hospital Cohort Study.

Background: Few studies have compared the optimal duration and intensity of organized multidisciplinary neurological/rehabilitative care delivered in a regional/district hospital with the standard rehabilitative care delivered in the general neurology/rehabilitation ward of a medical center. This study measured functional outcomes and conducted cost-utility analysis of an organized multidisciplinary postacute care (PAC) project in secondary care compared with standard rehabilitative care delivered in tertiary care.

Methods: This prospective cohort study enrolled 1,476 patients who had a stroke between March 2014 and March 2018 and had a modified Rankin scale score of 2-4. After exact matching for age ± 1 year, sex, year of stroke diagnosis, nasogastric tube, and Foley catheter and propensity score matching for the other covariates, we obtained 120 patients receiving PAC (the PAC group) from four regional/district hospitals and 120 patients not receiving PAC (the non-PAC group) from two medical centers.

Results: At baseline, the non-PAC group showed significantly better functional outcomes than the PAC group, including EuroQol-5 dimensions (EQ-5D), Mini-Mental State Examination (MMSE) and Barthel index (BI). During weeks 7-12 of rehabilitation, improvements in all functional outcomes were significantly larger in the PAC group ( < 0.001) except for Functional Oral Intake Scale (FOIS). Cost-utility analysis revealed that the PAC group had a significantly lower mean (± standard deviation) of direct medical costs (US$3,480 ± $1,758 vs. US$3,785 ± $3,840, < 0.001) and a significantly higher average gain of quality-adjusted life years (0.1993 vs. 0.1233, < 0.001). The PAC project was an economically "dominant" strategy.

Conclusions: The PAC project saved costs and significantly improved the functional outcomes of patients with stroke with slight to moderately severe disabilities. Randomized control trials are required to corroborate these results.