Diagnosis Related Groups and Inflammatory Bowel Disease: Variation in Outcome Measures

Abstract

Introduction: Centers for Medicare and Medicaid Services (CMS) uses diagnosis related group (DRG) codes to classify and reimburse inpatient admissions. DRG for inflammatory bowel disease (IBD) were selected for this study from the database maintained by University HealthSystems Consortium (UHC), an alliance of 120 academic medical centers. The DRGs were divided into major complication or comorbidity (MCC - DRG 385), with complication or comorbidity (CC - DRG 386) and without CC or MCC (DRG 387) based on the severity. The purpose of this study was to compare outcomes including length of stay and mortality for patients admitted under these DRGs. Methods: The cases submitted to the UHC Clinical Data Base/Resource Manager v 1.5.0.10 database by member institutions were analyzed for the year 2011. The data included observed outcomes as well as expected outcomes. Variables used for analysis included case volume at each institution, average and expected length of stay (LOS) with standard deviation (SD) and mortality. Descriptive statistics were incorporated for the analysis and conclusions. The numbers were rounded. Analysis was conducted for each DRG separately. Results: DRG 385: total of 1,298 cases were included in the database from 193 institutions. The average number of cases per facility was 6.73 (1-54). The mean LOS was 8.12 (1-49.5) with range of LOS SD from 0-67.18. The death rate was 1.39%. Mortality index was 2.51 (0-197.78). DRG 386: 217 institutions discharged 6,206 cases. The average number of cases per facility was 28.6 (1-181). The mean LOS was 4.96 (1-25) with SD range of 0-16.97. Death rate was 0.02% and average mortality index 0.85 (1-183.55). DRG 387: 216 institutions had total of 5,145 discharge. The average number of cases was 23.86 (1-179) with mean LOS 3.66 (1-15) having SD range 0-15.56. Death rate was 0%. Only 11.4% cases were with MCC. LOS increased with the severity of DRG, however the MCC had significantly higher LOS which was likely explained by few outlier cases as deduced from the range of standard deviation. All DRGs were associated with very low mortality. Conclusion: There is an increasing focus on tying outcomes with DRGs. The data set provided DRG specific insights into outcome variables for IBD. However, there is a significant limitation of lack of follow on data or correlation with therapies used. Future research using this data will identify correlation with predictive variables and association with DRG payments.