Comparative effects of fentanyl, morphine, and hydromorphone on postoperative delirium and outcomes after cardiac surgery: a propensity score-matched cohort study.

Postoperative delirium (POD) significantly increases mortality and medical burden, however, evidence regarding the best opioid options after cardiac surgery remains limited.Using the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, we identified 8246 adult patients receiving > 70% of postoperative 24-h morphine milligram equivalents (MMEs) during the ICU length of stay as fentanyl, morphine, or hydromorphone. After 1:1:1 propensity score matching (PSM) (n = 701 per group), we compared POD incidence, mortality, length of stay, vasopressor- and ventilator-free days within 28 days.The matched cohort of 2103 patients showed balanced baseline characteristics (all SMD < 0.1). POD incidence was significantly lower in the morphine (10.6%) and hydromorphone (9.7%) groups than in the fentanyl group (22.1%). The hydromorphone group was associated with significantly lower rates of ICU (OR 0.07, 95% CI 0.01-0.53) and hospital mortality (OR 0.25,95% CI 0.08-0.74), as well as shorter ICU and hospital length of stay, and more vasopressor- and ventilator-free days within 28 days. Hydromorphone was associated with a significantly lower incidence of POD, lower mortality, shorter length of stay, and more vasopressor- and ventilator-free days within 28 days. Whereas morphine was also associated with a reduced risk of delirium and shorter length of stay, it was not significantly associated with a reduction in mortality.