Prevalence and Predictors of Cannabis Use Disorder among Inflammatory Bowel Disease Hospitalizations in the United States and its Effect on Length of Stay

Background: The prevalence of cannabis use has been increasing in the United States in recent years. We examined the prevalence and predictors of cannabis use disorder among IBD hospitalizations and its effect on length of stay. Methods: Using the 2017 National Inpatient Sample, cannabis use disorder was identified using ICD-10-CM code F12.xxx in adult IBD patients. Other variables of interest included age, sex, race, Crohn’s disease (vs. ulcerative colitis), region, metropolitan status, zip code household income, primary insurance, and length of stay. Multivariate logistic and Poisson regressions were used in statistical analysis. Results: Of the 17,857 IBD hospitalizations, 565 (3.1%) had cannabis use disorder. Patients with cannabis use disorder were younger (mean ± standard deviation [years]: 35.1 ± 11.8 vs. 45.6 ± 18.1), less likely female (32.7% vs. 54.1%), more likely African American (24.8% vs. 13.9%), and more likely Crohn’s disease (72.0% vs. 62.2%) (all p<0.001). There were also significant differences by region, income, and insurance. Multivariate logistic regression confirmed age (odds ratio [95% confidence interval]: 0.964 [0.957, 0.971]), female (0.444 [0.369, 0.531]), African American (1.405 [1.124, 1.750]), and Crohn’s disease (1.363 [1.126, 1.657]) as predictors of cannabis use disorder. There was no association between cannabis use disorder and length of stay, confirmed in multivariate Poisson regression. Conclusions: Young age, male sex, African American race, and Crohn’s disease were positively associated with cannabis use disorder in IBD hospitalizations. There was no effect of cannabis use disorder on length of stay.