Predictive factors of PACU stay after herniorraphy in infant: a classification and regression tree analysis.

Herniorraphy is a common surgical intervention in infants, particularly in those born prematurely. Prematurity and perioperative sedation have been shown to be risk factors for postoperative apnea. However, their influence upon PACU stay duration has not been evaluated. The goal of this study was to investigate predictive factors for PACU stay in infants undergoing herniorraphy. This study is a retrospective analysis of perioperative data in infants <6 months of age undergoing herniorraphy during the period November 2007-November 2009. Collected data included age, gestational age at birth, post-conceptional age, weight, weight at birth, type of anesthesia (spinal vs general), perioperative administration of opioids and paracetamol, duration of surgery, duration of PACU stay, and apnea in PACU. Data analysis used classification and regression trees (CART) with a 10-fold cross-validation. Two hundred and ninety-six patients were included in the analysis. Five parameters were found to predict the duration of PACU stay: a post-conceptional age below 45 weeks, prematurity, general anesthesia, postoperative opioid administration, and the use of intraoperative regional analgesia. CRT method allows constructing a decision tree with eight terminal nodes. The percentage of explained variability of the model and the cross-validation were 79.7% and 76.6%, respectively. Our study allows construction of an accurate predictive tree for PACU stay during herniorraphy in infants <6 months. Parameters found to influence the duration of PACU stay were related to anesthesia techniques and perinatal outcomes.