|| Checking for direct PDF access through Ovid
We have performed a randomized comparison of two low-dose epidural regimens for analgesia in labour, differing only in the manner in which initial analgesia was established. In the epidural (EPI) group, 484 women received a loading dose of 20 ml of 0.1% bupivacaine with fentanyl 2 micrograms ml-1. In the combined spinal-epidural (CSE) group, 524 women received a spinal injection of plain bupivacaine 2.5 mg with fentanyl 25 micrograms. In both groups, these initial doses were followed by 0.1% bupivacaine with fentanyl 2 micrograms ml-1 infused at a rate of 12 ml h-1, with 20-ml top-ups for breakthrough pain. The groups were compared for midwife assessment of analgesic efficacy, delivery mode, patient assessments of first stage analgesia, second stage analgesia, overall analgesia, motor block and complications. Midwives, who were not blinded to the treatment groups, assessed 61.6% of CSE as providing 'excellent' analgesia compared with 56.4% of epidurals (P = 0.02). Patients assessed overall analgesia as 'excellent' in 74.8% of CSE compared with 71.7% of epidurals (P = 0.14). Other comparisons between groups revealed no differences. These findings may have been affected by an uneven distribution of multiparous women between the groups (25% in the EPI group and 34.2% in the CSE group; P = 0.002). However, subgroup analysis of primiparous and multiparous women did not alter the results.