Changing the Relative Weights of Tw-DRGs by Applying MS-DRG Major Complications and Comorbidity Standards

Abstract

Objectives: Adjusting the severity of illness (SOI) for each inpatient service is important for reducing financial risks faced by different hospitals and for maintaining patients' accessibility to and equity in an inpatient payment system. In this study, we analyzed four Tw-DRGs (Taiwan version of Diagnosis-Related Groups) by applying the new Major Complications and Co-morbidity (MCC) standards in the new Medicare Severity-adjusted DRGs (MS-DRG). Variations in medical expenses and length of stay (LOS) among hospitals with different accreditation and ownership status were examined after employing more appropriate SOI standards. Methods: A retrospective study was conducted using the 2006 Taiwan NHI inpatient claim database. After re-grouping Tw-DRGs by MCC, it was found that four DRGs had MCC in over 60% of their claims. These included Tw-DRG 12101 (Circulatory disorders with acute myocardial infarction and major complications, discharged alive with Complications and Comorbidities (CC), 20201 (Cirrhosis and alcoholic hepatitis with CC), 41601 (Septicemia age≧18 with CC) and 47501 (Respiratory system diagnoses with ventilator support with CC) with 2,294, 19,411, 28,576 and 33,178 cases respectively. These four DRGs were selected for comparison of the differences in medical expenses for inpatients with MCC versus those without MCC. The coefficient of variation (CV) of medical expenses was applied to measure intra DRG homogeneity. The hospital case mix index (CMI) and medical expenses were also calculated and compared to examine changes in CMI across hospitals. Results: Public medical centers had the highest average expense per discharge (relative value unit 95,611.20), followed by non-profit medical centers (89,338.56), regional hospitals and district hospitals. The CMI for public medical centers under the Tw-DRGs incorporating the MCCs standards increased by 1.099% more than did the CMI for regional hospitals; however, the CMI for non-profit regional hospitals decreased slightly by -0.02%. CMI reduction corresponded with lower expenses for district hospitals. MCC-split DRGs had a lower CV with respect to medical expenses. Conclusions: The findings of this study indicated that MCCs accounted for variations in medical expenses for patients in the same DRG across different hospitals. Our findings were limited to four Tw-DRGs and should be interpreted cautiously.