OBJECTIVES: To characterize resource utilization and health outcomes for hospitalized Crohn’s disease (CD) patients.

METHODS: Hospitalized CD patients were identified using de-identified administrative claims from the Optum Clinformatics^® Data Mart (2007-2019). We included patients aged 18-64 with an index inpatient claim with a primary or secondary CD ICD code, and who were continuously enrolled for 1 year pre- and post-index hospitalization. The primary health outcome was CD-related rehospitalization within 1 year. Sociodemographic controls included age, gender, education, income, geography, race, and insurance type. Clinical and utilization variables included admission type, length of stay, co-morbid conditions, procedures, medications, prior hospitalizations, and ED visits. Differences across the primary outcome were tested via t-test for continuous variables and Pearson χ2 test for categorical variables. CD-related health outcomes and utilization at index were compared according to biologic use in the pre-period.

RESULTS: 13,918 patients met all inclusion criteria, of which 3,521 were rehospitalized for CD. Mean age at index differed significantly between rehospitalized and not rehospitalized patients (respectively 40.1 and 43.2, P<0.001). Lower income and black race were also associated with rehospitalization. Any biologic use in the pre-index period was strongly associated with rehospitalization (33.6% and 20.1%, P<0.001); this strong association was consistent across biologic type (TNF inhibitor vs. other). Most CD-related comorbidities and procedures were strongly associated (P<0.001) with rehospitalization. Biologic use in the pre-period was strongly associated with presence of CD-related comorbidities (e.g. fistula) and procedures (e.g. small bowel resection, colectomy, ileostomy) at index.

CONCLUSION: Clinical and utilization variables are strongly associated with CD rehospitalization, and may prove to be useful for predictive modeling. Isolating the effect of biologics on health outcomes may be confounded by unobserved disease severity. Analyses that identify the causal impact of biologics on health outcomes and resource utilization are needed.