Assessment of the PI ratio for predicting the timing of endotracheal intubation: an observational study

Background The suitable timing for endotracheal intubation is currently determined mainly by clinical experience, as objective and accessible indices are lacking. Although the perfusion index (PI) correlates with anaesthetic depth, its marked baseline variability limits its reliability as a standalone predictor. This study aimed to explore the value of the standardized PI ratio (postinduction PI/baseline PI) as a new noninvasive indicator for predicting haemodynamic stability during endotracheal intubation in children. Methods This was an observational study involving 172 children aged 4–10 years who underwent daytime surgery. The PI was recorded before anaesthesia induction (T1), immediately before intubation (T2), and one minute after intubation (T3), and the PI ratio (PI₂/PI₁) was calculated. Children were assigned to the significant haemodynamic fluctuation (SHF) group or the non-SHF (NSHF) group based on whether heart rate or mean arterial pressure increased by > 20% from baseline during intubation. The predictive performance of the PI and the PI ratio for appropriate intubation timing was analysed using receiver operating characteristic (ROC) curves. Results The area under the ROC curve (AUC) for the ability of the PI alone to predict the timing of endotracheal intubation was 0.630 (95% CI 0.523–0.737) at a cut-off > 5.50, with a sensitivity of 46.5% and a specificity of 76.7%. The PI ratio demonstrated moderate predictive performance, with an AUC of 0.775 (95% CI 0.676–0.874). A PI ratio > 3.623 might predict a haemodynamically suitable state for intubation, with a sensitivity of 88.0% and a specificity of 60.0%. All intubation-related adverse events occurred only in the SHF group. Conclusion The PI ratio can help mitigate the influence of individual baseline PI variability and shows a higher AUC than PI alone, making it more suitable for predicting a stable haemodynamic response to endotracheal intubation in children. It may serve as a useful adjunct to clinical assessment, but further multicentre validation is required before routine implementation. Clinical trial registration This trial was registered at the Chinese Clinical Trial Registry ( https://www.chictr.org.cn . Registration number ChiCTR2500107129, Xu Haiping, 4 Aug 2025.)