The association of hospital volume with mortality and costs of care for stroke in Japan

Abstract

Background: The association between hospital volume and patient outcomes remains unclear for stroke. Little is known about whether these relationships differ by stroke subtypes.

Objectives: To examine the association of hospital volume with in-hospital mortality and costs of care for stroke.

Research design: Secondary data analysis of national hospital database.

Subjects: A total of 66,406 patients admitted between July 1 and December 31, 2010 with primary diagnosis of stroke at 796 acute care hospitals in Japan were included.

Measures: We used a locally weighted scatter-plot smoothing method to test the relationship between hospital volume and outcomes. On the basis of these results, we categorized patient volume into 3 groups (10-50, 51-100, and >100 discharges/6 mo). We tested the volume-outcome relationship using multivariable regression models adjusting for patient and hospital characteristics. Subgroup analysis was conducted by stratifying on stroke subtype.

Results: Compared with those treated at high-volume hospitals (>100 discharges), patients admitted to low-volume hospitals (10-50 discharges) had higher in-hospital mortality (adjusted odds ratio, 1.45; 95% CI, 1.23-1.71, P<0.0001). In the lowest volume hospitals, adjusted costs of care per discharge were 8.0% lower (95% CI, -14.1% to -1.8%, P=0.01) compared with the highest volume hospitals. The volume-mortality association was significant across all stroke subtypes. Highest volume hospitals had higher costs than lowest volume hospitals for subarachnoid hemorrhage, but this association was nonsignificant for ischemic and hemorrhagic stroke.

Conclusions: Highest volume hospitals had lower mortality than the lowest volume hospitals for stroke in Japan. Highest volume hospitals had higher costs for subarachnoid hemorrhage, but not for ischemic and hemorrhagic stroke.