Background: Patients with substance use disorders (SUD) are at an increased risk of admission to intensive care units (ICUs) and the relationship between SUD and commonly used sedative and analgesic medications remains understudied. Objective: The objective of this study was to describe associations between substance use disorder (SUD) and medication dose requirements in patients with critical illness requiring invasive mechanical ventilation (IMV). Methods: This was a single center, retrospective study of adult patients admitted to an ICU between January 1, 2017, to October 31, 2022 who were initiated on IMV within 48 hours. The primary outcome was median total daily dose of each sedative and analgesic medication received, grouped as follows: parenteral opioids, benzodiazepines, propofol, and dexmedetomidine. Multivariable median regression was used to determine associations between SUD and median daily dose after adjustment for demographic and clinical covariables. Results: Of 1290 included patients, 358 (27.8%) had a documented SUD present on admission. In unadjusted analyses, patients with SUD received higher median daily doses of parenteral opioids (197.5 MME IQR 61.9-344.3 vs 162.2 MME IQR 60.3-324.6) and propofol (26 640 µg/kg IQR 11 659-43 518 vs 23 248 µg/kg [IQR 10 510-38 870). No differences in median daily dose were noted for parenteral benzodiazepines or dexmedetomidine. In adjusted analyses, associations between SUD and median daily dose of parenteral opioids (β, 18.1 MME; 95% CI, −13.6 to 49.4) or propofol (β, 1699.9 µg/kg; 95% CI, −2477.1 to 5882.8) were no longer present. Conclusion and relevance: Results of unadjusted analyses aligned with prior studies reporting increased dose requirements of sedation and analgesia in patients with SUD; however, associations were diminished after adjustment for clinical and demographic covariables. Our results suggest the presence of SUD alone may not necessitate empirically higher medication doses in patients with SUD receiving IMV during critical illness.
Feeney et al. (Wed,) studied this question.