Postoperative Delirium in Surgical Patients in Intensive and Non-Intensive Care Units: Incidence, Risk Factors, and Prognostic Significance

BACKGROUND: Postoperative delirium (POD) is a common complication that is associated with adverse outcomes. However, differences between surgical patients in intensive care unit (ICUs) and non-ICUs are unclear. This study investigated POD incidence, prognostic significance, and risk factors in ICU and non-ICU settings. METHODS: We retrospectively analyzed data from 20,338 adults who underwent general anesthesia at Nippon Medical School Hospital between July 2021 and March 2025. Patients were classified as ICU (n=2,451) and non-ICU patients (n=17,887). POD was defined by an Intensive Care Delirium Screening Checklist score of ≥4 or prescription of antipsychotics. Logistic regression was used to identify risk factors, and Kaplan-Meier analysis assessed 180-day mortality. Subgroup analysis evaluated differences between early (≤2 days) and late (3-7 days) POD. RESULTS: The overall incidence of POD was 4.4% (889/20,338) and was higher in ICU patients (19.1%) than in non-ICU patients (2.3%; p<0.001). POD was positively associated with 180-day mortality in both groups (p<0.001). Among non-ICU patients, survival was worse among patients with late POD than among those with early POD (p<0.001), whereas onset timing had no impact in ICU patients. Risk profiles differed: advanced age, higher American Society of Anesthesiologists class, and cerebrovascular disease were associated with POD in non-ICU patients, while male sex and dialysis were significant factors for ICU patients. Excessive positive fluid balance and prolonged operative time increased POD risk in both groups, but gradients were steeper in non-ICU patients. CONCLUSIONS: POD was independently positively associated with excess mortality across settings, but the risk factors and prognostic implications varied. These findings highlight the need for ward-specific preventive strategies and tailored perioperative management.