• John Song En Lee, Rehena Sultana, Nian Lin Reena Han, Sia Alex Tiong Heng ATH Department of Women's Anaesthesia, KK Women's and Children's Hospital, 100 Bukit Timah Road, Singapore, 229899, Singapore. , and Ban Leong Sng.
• Department of Women's Anaesthesia, KK Women's and Children's Hospital, 100 Bukit Timah Road, Singapore, 229899, Singapore.
• BMC Anesthesiol. 2018 Nov 29; 18 (1): 176.

BackgroundEpidural catheter re-siting in parturients receiving labour epidural analgesia is distressing to the parturient and places them at increased complications from a repeat procedure. The aim of this study was to develop and validate a clinical risk factor model to predict the incidence of epidural catheter re-siting in labour analgesia.MethodsThe data from parturients that received labour epidural analgesia in our centre during 2014-2015 was used to develop a predictive model for epidural catheter re-siting during labour analgesia. Multivariate logistic regression analysis was used to identify factors that were predictive of epidural catheter re-siting. The forward, backward and stepwise variable selection methods were applied to build a predictive model, which was internally validated. The final multivariate model was externally validated with the data collected from 10,170 parturients during 2012-2013 in our centre.ResultsNinety-three (0.88%) parturients in 2014-2015 required re-siting of their epidural catheter. The training data set included 7439 paturients in 2014-2015. A higher incidence of breakthrough pain (OR = 4.42), increasing age (OR = 1.07), an increased pain score post-epidural catheter insertion (OR = 1.35) and problems such as inability to obtain cerebrospinal fluid in combined spinal epidural technique (OR = 2.06) and venous puncture (OR = 1.70) were found to be significantly predictive of epidural catheter re-siting, while spontaneous onset of labour (OR = 0.31) was found to be protective. The predictive model was validated internally on a further 3189 paturients from the data of 2014-2015 and externally on 10,170 paturients from the data of 2012-2013. Predictive accuracy of the model based on C-statistic were 0.89 (0.86, 0.93) and 0.92 (0.88, 0.97) for training and internal validation data respectively. Similarly, predictive accuracy in terms of C-statistic was 0.89 (0.86, 0.92) based on 2012-2013 data.ConclusionOur predictive model of epidural re-siting in parturients receiving labour epidural analgesia could provide timely identification of high-risk paturients required epidural re-siting.

This article appears in the collection: Predictors of labour epidural breakthrough pain.

29 keywords...

hide…