Hospital specialization and hospital charge, length of stay, and mortality for lumbar spine disease inpatients

Abstract

Background

In South Korea, notable recent health policy changes such as designation of specialty hospitals and implementation of diagnosis related groups (DRGs) have occurred, hospitals have begun exploring and planning specialization strategies in order to achieve competitive advantage and attract more patients. Thus, precise, valid, and reliable measures of hospital specialization for competitive advantage have become increasingly necessary. However, there exist various limitations on the case-mix hospital specialization index based on patient proportion. Thus, the purpose of this study is to modify the category medical specialization index and to investigate the association between this modified category medical specialization index and length of stay (LOS), mortality, and hospital charge.

Methods

A representative national sample dataset provided by the National Health Insurance Service-Cohort Sample Data (NHIS-CSD) for 12 years, 2002 to 2013, was employed. To extract lumbar spine disease patients within hospitals from these data, diseases were classified into 267 categories based on the International Classification of Diseases, Tenth Revision (ICD-10). Modified Category Medical Specialization (Modified CMS) was incorporated into log transformation to the denominator of CMS, to reduce between-hospital variation in number of medical categories. Associations with LOS, mortality, hospital cost per case and hospital cost per diem. This study included 56,622 cases; the primary analysis was based on a generalized estimating equation (GEE) regression model accounting for correlation among hospitals within each region to examine our hypothesis.

Results

Our modified CMS shows a U-shaped trend of hospital specialization from small to large hospitals, with excellent goodness-of-fit. Among lumbar spine disease surgery patients, the adjusted effects of association between modified CMS and length of stay, mortality, hospital cost per case, and total cost per diem show that LOS in hospitals with high modified CMS was -2.539 days shorter (p: <.0001), mortality of hospitals with high modified CMS was 0.972 times lower (p: 0.715), total cost per case of hospitals with high modified CMS was -152,060 won lower (p: <.0001), and total cost per diem of hospitals with high modified CMS was -42,362 won higher (p: <.0001).

Conclusions

To our knowledge, this is the first study to develop and implement modified CMS for a specific disease (lumbar spine disease). The results show that increase in hospital specialization has a substantial effect on decrease in hospital costs per case, LOS, mortality, and increase in hospital cost per diem. Therefore, in the context of increasing competition and recent policy changes by the Korean government, our results may help Korean and non-Korean hospital policymakers understand the effects of hospital specialization strategy on hospital profitability and efficiency.