Brief, persistent, and recurrent delirium in a surgical ICU population: Incidence, risk factors, and short-term outcomes

Aim: To identify the incidence, risk factors and short-term outcomes for brief, persistent, and recurrent courses of postoperative delirium in Intensive Care Unit (ICU) patients. Method: This secondary analysis of the prospective PREDICt cohort included adult surgical ICU patients who developed postoperative delirium (POD). Delirium was classified as brief (≤1 day), persistent (>1 day), or recurrent (reappearance after ≥48 h of recovery). Demographic, perioperative, and postoperative variables were examined. Univariable comparisons and multivariable logistic regression were performed to identify risk factors for delirium courses, with variables yielding p < 0.2 in the univariable analysis entered into the multivariable model. Results: Among the 272 patients with POD, 96 experienced brief delirium, 156 persistent delirium, and 20 recurrent delirium. Compared with brief delirium, recurrent delirium was independently associated with lower ADL scores (OR = 0.98, 95% CI: 0.96–0.99), postoperative blood transfusion within 24 h of ICU admission (OR = 4.71, 95% CI: 1.46–15.24), and benzodiazepine exposure (OR = 9.33, 95% CI: 1.91–45.48). Given the wide confidence intervals and the small number of recurrent cases, all estimates for the recurrent delirium subgroup should be considered hypothesis-generating. Persistent delirium was associated with fewer years of education (OR = 0.92, 95% CI: 0.86–0.98) and lower ADL scores (OR = 0.99, 95% CI: 0.98–0.99). Persistent and recurrent delirium were associated with worse clinical outcomes than brief delirium, including longer mechanical ventilation duration, prolonged ICU stay, and higher in-hospital mortality. Conclusion: Delirium courses differ in risk factors and prognosis. Persistent delirium was associated with lower preoperative ADL scores and fewer years of education, while recurrence was associated with lower ADL scores, postoperative blood transfusion, and benzodiazepine exposure. Recurrent delirium carried the worst outcomes, including prolonged ventilation, extended ICU stay, and higher mortality. Implication: Recognizing delirium as a dynamic process rather than a transient event underscores the need for course-based risk stratification. Early identification of high-risk patients and modification of exposures, such as transfusion and benzodiazepine use, may help prevent persistent and recurrent delirium and improve outcomes.