Abstract

Objectives

The 21st century has an increasing elderly population at risk of cerebrovascular disease (CVD). Efficient care for recovering functional status is emphasized among policy makers. We investigated whether rehabilitation and its early initiation provided for CVD patients produced functional recovery in acute care hospitals.

Methods

Using a Japanese administrative database during a 4-month interval from 2004 to 2008 in patients ages ≥15 years, we measured the demographics, consciousness level at admission, comorbidities, complications, procedures, ventilation administration, initiation day of rehabilitation, and hospital characteristics. Outcomes included total charges (TC) and functional status measured by the Barthel index (BI). Multivariate analysis measured the impact of rehabilitation and its early initiation on outcomes. To reduce the selection bias of rehabilitation and the ecological fallacy, we used propensity score matching and the linear mixed model.

Results

Excluding 488 deceased patients, we analyzed 45,014 CVD patients. Rehabilitation at a generalized unit produced greater BI improvement than no rehabilitation or at intensive care units. A longer hospitalization, but not a 1-day delay of rehabilitation initiation, resulted in less BI improvement and more TC. A higher patient volume and academic hospitals were associated with more TC but not with BI improvement.

Conclusions

Rehabilitation, but not the timing of rehabilitation, might accompany functional recovery in acute care hospitals. Because the hospital mix or medical units can explain the variation in the quality of rehabilitation, policy makers, along with monitoring unnecessary long hospitalizations, should encourage a referral policy for rehabilitation-intensive facilities and develop effective rehabilitation using technology to optimize functional outcomes.