Background and aims Lumbar epidural is the most effective labour analgesia but is associated with increased assisted vaginal delivery, which leads to maternal and fetal morbidities. Our primary objective is to evaluate intrapartum risk factors for ID compared to normal vaginal delivery (NVD) to mitigate these risks for better outcome.

Methods This study was approved by Singhealth Centralised Institutional Review Board. Data was obtained retrospectively from department database of obstetric epidural records at a single specialist centre - KK Women’s and Children’s Hospital between January 2012 to December 2015. Maternal data collected are sociodemographics, parity, antenatal problems, cervix dilation, labour induction and augmentation. Epidural data included technical performance, pain score, sensory and motor blockade, breakthrough pain. Fetal data included birth weight, APGAR score. Univariate analysis was performed with p<0.1 selected for multivariate logistic regression. Independent association factors were identified with p<0.05.

Results 17227 patients having labour epidural delivered vaginally from January 2012 to December 2015. NVD comprised 88% (n=15158) and ID comprised 12% (n=2069). Independent risk factors were identified: shorter maternal height, nulliparity and increased maternal age, prostin induction, no pre-epidural analgesia, longer second stage, higher fetal birth weight and having private specialist care. Epidural-related factors are breakthrough pain, high bromage score and having epidural infusion stopped at delivery.

Conclusions We can use these to identify high-risk parturients. High motor block can be avoided by using low concentrations of local anaesthetic. a PCEA regime allows greater control over breakthrough pain and reduce delays in top-ups. If possible, epidural infusion should not be switched off at delivery.