Cannabis use and risk of Clostridioides difficile infection: Analysis of 59,824 hospitalizations.

Abstract

BACKGROUND\nThe prevalence of Clostridioides difficile Infection (CDI), the most notorious hospital acquired disease, and of excessive cannabis use (cannabis use disorder (CUD)) have both been steadily rising. Although cannabidiol, an active ingredient of cannabis, maintains gut integrity and suppresses entero-toxins from Clostridioides difficile, the relationship between CUD and CDI has not been studied.\n\n\nMETHODS\nWe selected adult records (age\u202f≥\u202f18 years) from the Nationwide Inpatient Sample 2014, and identified CUD and other clinical conditions using ICD-9-CM codes. We used propensity scores derived from a multivariate logistic model to match CUD to non-CUD in a 1:1 ratio (29,912:29,912). We estimated the relative risk for CDI using log-binomial regression models with generalized estimating equations [SAS 9.4].\n\n\nRESULTS\nAmong the matched hospitalizations (n\u202f=\u202f59,824), cannabis usage was associated with a reduced prevalence of CDI (prevalence: 455.5 [95% CI: 385.1-538.8] vs. 636.4 [95% CI: 549.9-736.5] per 100,000 hospitalizations), resulting in a 28% reduced risk of CDI (relative risk: 0.72 [95% CI: 0.58-0.88]; p\u202f=\u202f0002). Non-dependent and dependent CUD respectively had 23% and 80% reduced likelihood of CDI when compared to non-cannabis users (0.77 [95% CI: 0.60-0.95] and 0.20 [95% CI: 0.06-0.54]; p\u202f<\u202f0.05). Furthermore, dependent users had less risk of CDI compared to non-dependent users (0.26 [95% CI: 0.08-0.88]; p\u202f=\u202f0.01).\n\n\nCONCLUSIONS\nCUD was associated with a decreased risk of CDI amongst hospitalized patients. Prospective and molecular mechanistic studies are required to elucidate how cannabis and its contents impacts CDI.