Work

Early benzodiazepine exposure and short-term risk of hepatic encephalopathy treatment initiation in critically ill patients with cirrhosis: a target trial emulation using the MIMIC-IV database

Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background : Sedation is frequently required in the intensive care unit (ICU), yet benzodiazepines are traditionally avoided in cirrhosis because of their GABAergic potential to precipitate or worsen hepatic encephalopathy (HE). Whether early ICU benzodiazepine exposure actually increases short-term HE risk remains unquantified: to our knowledge, no previous study has estimated this contrast at the treatment-strategy level. Objective To evaluate whether benzodiazepine initiation within 24 h of ICU admission is associated with short-term HE worsening, operationalized as new initiation of lactulose or rifaximin between 24 h and 7 days, among critically ill patients with cirrhosis. Methods : Using the Medical Information Mart for Intensive Care IV (MIMIC-IV, v3.1) database, we emulated a target trial and applied a clone-censor-weight (CCW) approach. Exposure was benzodiazepine administration (midazolam, lorazepam, or diazepam, any route) within a grace period after ICU admission (primary: 24 h; sensitivity: 12 and 36 h). The primary outcome was HE treatment initiation, defined as new lactulose or rifaximin dispensing between 24 h and 7 days. Censoring was addressed with per-arm inverse probability of censoring weights (IPCW). Two negative-control outcomes (new proton pump inhibitor or statin initiation) quantified residual confounding.
Language: English
Subfield: Hepatology
Field: Medicine
Open Access status: green
OpenAlex ID: W7225912247
Cites: 40
Cited by:
Related to: 10