Observational study for perioperative intensive care unit (OPICU): Planned and unplanned ICU admissions, follow-up,...

Due to limited resources, identifying high-risk surgical patients who derive clinical benefit from Intensive Care Unit (ICU) admission is essential. Our primary aim was to characterize resource allocation and evaluate the true clinical benefit of ICU admissions using modified Wang criteria. Secondary aim is to identify independent risk factors for ICU admission; validate clinical scoring systems; and compare inter-group outcomes, including Clavien-Dindo complication rates, lengths of stay, and mortality.

This multicenter, prospective, observational study was conducted across 75 ICUs. Adult inpatient surgery cases were categorized into four groups: planned admission (PAG), unplanned admission (UPAG), planned but not admitted (PBNAG), and emergent admission (EAG). Data included preoperative/intraoperative/postoperative(ICU) variables, modified Wang criteria, and mortality rates (ICU, hospital, 30/90-day).

Of 55,963 preoperatively evaluated patients, 2764 (4.9%) were planned for ICU admission; 804/2764 patients could not be admitted because of limited ICU beds, remaining of 1960 patients with a planned ICU admission, 411/1960 did not require postoperative ICU care forming the PBNAG (14.8% of 2764). Overall, 1980 patients (3.09% of all cases) were admitted to the ICU (78.2% PAG, 12.5% EAG, 9.1% UPAG). POSSUM scores effectively differentiated PBNAG (lower) and EAG (higher) from PAG. ICU benefit indicators (e.g., vasopressor need, prolonged ventilation >12 h, arrest events) were significantly higher in UPAG and EAG versus PAG (p < 0.001). Emergency surgery (OR 12.4), intraoperative complications (OR 3.1), and ASA >2 (OR 2.7) independently increased UPAG risk, whereas minimally invasive surgery and regional anesthesia reduced ICU admissions. Overall ICU and 90-day mortality rates were 5.9% and 7.5%, respectively, peaking in the EAG group (18.1% and 18.9%).

Integrating modified Wang criteria, POSSUM scores, intraoperative complications, emergency status, and ASA class >2 is effective in predicting the necessity of ICU admission.