Construction and validation of the validity of a column chart based on pulse oximetry perfusion index and its difference multiplier to predict the anesthetic effect of epidural anesthesia for cesarean delivery surgery

Abstract OBJECTIVE To assess the application of PI and dPI in epidural anesthesia for cesarean delivery, to determine the effect of anesthesia, and to provide predictive criteria for determining whether the plane of anesthesia is appropriate for surgery. METHODS 179 women undergoing cesarean section in the Department of Obstetrics of the Second Affiliated Hospital of Xinjiang Medical University were included, and after epidural anesthesia was administered, the PI values of the index finger of the hand and the thumb of the foot were measured at 5min, 7min, and 9min after administration of the drug, and when the plane of anesthesia reached T6, respectively. The model was constructed using the 5-min data and expressed using a line graph, the model performance was evaluated using ROC curves, DCA curves, and clinical impact curves, and finally, the model was tested using internal validation and external validation using the 7- and 9-min data as external data. RESULTS Patients with higher dPI (> 1.85) as well as higher PI (> 4.15) had significantly higher rates of achieving anesthesia after epidural anesthesia; the column plots were satisfactorily calibrated and discriminated between the training cohort, internal validation cohort and external validation cohort (C-index for the training cohort was 0.989, for the internal validation cohort was 0.989, for the external validation cohort was 0.885); in addition DCA as well as clinical decision curves indicated a net clinical benefit for the column-line diagram. CONCLUSION Both PI and dPI were identified as independent predictors and ultimately included in the model, which can be used to predict the effectiveness of the plane of anesthesia in women undergoing cesarean delivery after epidural anesthesia.