807 Superior Outcomes in Crohn’s But Not UC Admissions to Academic Hospitals

Abstract

INTRODUCTION: Inflammatory bowel disease (IBD) affects an estimated 3.1 million US adults and is increasing in prevalence. Recent data found lower mortality rates for common conditions in Medicare beneficiaries admitted to academic vs non-teaching hospitals. This study compares outcomes for patients admitted with IBD to academic vs non-academic hospitals. METHODS: The national inpatient sample (NIS) from 2016 was queried for all cases of Crohn’s disease (CD) or ulcerative colitis (UC) where this was the principle cause for admission (first ICD-10 code). Outcomes for patients admitted to academic (teaching) hospitals were compared to those admitted to non-academic rural or urban hospitals using survey-adjusted chi square and linear regression to examine mortality, length of stay (LOS), abdominal surgery, and hospitalization costs. Multivariable models adjusting for Elixahauser mortality score (an NIS-specified validated comorbidity measure) and surgery during admission were created. CD and UC patients were analyzed separately. RESULTS: In 2016, there were 35,950 admissions for UC and 71,040 for CD. No difference was seen in mortality for academic (0.4%) versus community (0.5%) hospitals in UC (P = 0.56). Mortality for CD patients was 0.2% in academic vs 0.4% in non-teaching hospitals (P = 0.06). Adjusting for surgery and Elixhauser score, patients admitted to a non-academic hospital had an OR for death of 0.95 in UC (P = 0.90) but 2.2 in CD (P = 0.03). Patients at academic hospitals were more likely to undergo surgery for both UC (19% vs 5.1%, P < 0.0001) and CD (22.9% vs 11.7%, P < 0.0001). Elixhauser scores were higher for UC (6.0 vs 5.4, P = 0.01) as well as CD (4.3 vs 3.8, P = 0.002) patients at non-academic hospitals. Regression analysis showed that UC patients at non-academic hospitals had shorter LOS (P = 0.001) with no significant differences in cost. In the CD regression, non-academic hospitals had shorter LOS (P < 0.0001) with lower cost (P = 0.01). CONCLUSION: After adjustment for comorbidities and surgery during admission, patients admitted to community hospitals had twice the odds of death in CD; no similar effect was seen with UC. Surgical interventions and LOS were significantly lower for all IBD patients at community hospitals despite higher Elixhauser scores. Further research is needed to investigate outcomes in IBD patients admitted to academic vs community hospitals.