Table 2.

Main studies about PACU length of stay

Author, year	Country	Study design	Type of procedure	Main outcomes	Objective	Final Cohort	Type of AI	Prediction Performance	External validation
Schulz EB. Br J Anaesth. 2020.	Queensland, Australia.	Observational, retrospective, single-centre study	All cases involving an anaesthetic doctor	PACU LOS	Production of case-mix and risk-adjusted post anesthesia care unit length of stay (LOS) benchmarks for integration into modern reporting tools.	67 325 cases	MinMax scaling	This predictive model was able to account for much of the variability observed in individual anaesthetists’ mean PACU LOS (Spearman’s r2 = 0.57). By subtracting the predicted PACU LOS, anaesthetists fell in a much tighter range, with 80% of anaesthetists having a mean LOSD that fell in a band of only 4.3 min, compared with a spread of 24 min for unadjusted mean LOS.	No
Cao B. Ann Palliat Med. 2021.	China	Observational, retrospective, monocentric study	Laparoscopic cholecystectomy	PACU LOS	Development of a predictive nomogram to aid in identifying which LC patients are more likely to be subjected to prolonged PACU LOS.	913 patients	LASSO regression model, C-index, calibration plot, and DCA.	This model displayed efficient calibration and moderate discrimination with a C-index of 0.662 (95% confidence interval, 0.603 to 0.721) for the training set, and 0.609 (95% confidence interval, 0.549 to 0.669) for the test set. DCA demonstrated that the prolonged PACU LOS nomogram was reliable for clinical application when an intervention was decided at the possible threshold of 7%.	No
Tully JL. J Med Syst. 2023.	California, USA	Observational, retrospective, monocentric study	Outpatient surgical procedures	PACU LOS	Development of machine learning models to predict ambulatory surgery patients at risk for prolonged PACU length of stay and then to simulate the effectiveness in reducing the need for after-hours PACU staffing.	10 928 Outpatients	Logistic regression, feedforward neural network, XGBoost regressor, balanced random forest classifier, balanced bagging classifier, and Random Forest classifier.	Female sex (P < 0.0001) and scheduled surgical case duration (P < 0.0001) were associated with prolonged PACU LOS. Based on AUC, the best performing model with SMOTE was XGBoost (AUC 0.779), whereas the worst performing model with SMOTE was logistic regression without SMOTE (AUC 0.718)	No
Gabriel RA. Anesth Analg. 2022 Jul. *	California, USA	Observational, retrospective, monocentric study	Orthopedic and ear, nose, and throat surgeries	Surgery end time and PACU LOS	Development of machine learning models that predicted the following composite outcome: surgery finished by end of operating room block time and patient was discharged by end of recovery room nursing shift.	13 447 surgical procedures	Logistic regression, Random Forest classifier, support vector classifier, simple feedforward neural network, balanced Random Forest classifier, and balanced bagging classifier. SMOTE.	It has been created a model for each start time (1 pm, 2 pm, 3 pm, or 4 pm) and showing that the ensemble learning approaches had the highest AUC scores. The balanced bagging classifier performed best with F1 score of 0.78, 0.80, 0.82, and 0.82 when predicting our outcome if cases were to start at 1 pm, 2 pm, 3 pm, or 4 pm, respectively.	No

PACU: Post Anesthesia Care Unit. LOS: length of stay. AUC: area under the curve. DCA: decision curve analysis.