Epidural anaesthesia for elective caesarean section can have advantages over general anaesthesia. The anaesthesiologist can avoid endotracheal intubation as well as fetal depression following placental transfer of systemic anaesthetics. However, despite reaching an effective blockade preoperatively, intraoperative discomfort and pain may occur during epidural anaesthesia with local anaesthetics alone, necessitating supplemental systemic analgesics or even conversion to general anaesthesia [21]. Addition of epidural fentanyl has been shown to improve onset and quality of perioperative analgesia without evident side effects for mother or newborn [24]. Nevertheless, administration of epidural opioids before cord clamping is still hotly debated, some fearing maternal and or neonatal depression [6, 26]. The aim of the present study was to investigate the quality of analgesia, associated side effects and the resulting maternal and neonatal plasma opiate concentrations after a single preoperative addition of 0.1 mg fentanyl to epidural bupivacaine analgesia in comparison to epidural bupivacaine analgesia alone. METHODS Following governmental and ethics committee approval, 43 elective consenting patients for caesarean section were randomized to receive double-blind injections of either 8 ml 0.5% bupivacaine(+)0.1 mg fentanyl (B+F group, n = 22) or 8 ml 0.5% bupivacaine +2 ml saline (Bup group, n = 21) into an epidural catheter. In both groups additional injections of bupivacaine were given to achieve sensory blockade up to T4. Systolic blood pressure, heart and respiratory rates were measured regularly. Quality of intraoperative pain relief was assessed at delivery, uterine eventration, and during uterine and abdominal closure using a visual analogue scale (VAS). The duration of postoperative analgesia was compared between groups, as well as the incidence of nausea, itching or sedation. Similarly, Apgar scores and umbilical arterial and venous blood gas analyses were compared. Fentanyl concentrations were determined in maternal venous blood sampled before and 20 and 40 min after epidural injection and at birth, and in umbilical venous and arterial blood sampled after delivery. Radioimmunoassay analysis was performed from plasma specimens centrifuged and frozen at -20 degrees C [19]. The statistical level of significance was defined as P < 0.05. RESULTS Groups were comparable regarding age, weight and time of gestation. Total bupivacaine doses and injection to delivery times were similar in both groups. Figure 1 shows that there were 40% more pain-free (VAS = 0) patients in the B+F group during uterine eventration and wound closure (P < 0.05). Mean postoperative duration of analgesia was significantly longer in the B+F group (382 vs 236 min). The rate of nausea and mild itching was significantly higher in the B+F group. Respiratory depression was never detected in patients or newborns. Small group differences in blood pressure or respiratory rate were inconstant and clinically irrelevant, as were differences in umbilical venous pCO2. One hundred and twenty-five blood samples were analysed for fentanyl concentrations. The mean fentanyl concentration before epidural injection was not zero, but 0.25 ng/mg (range 0.02-0.32). Maternal concentrations at 20 and 40 min after injection were 0.55 ng/ml (0.12-1.14) and 0.52 ng/ ml (0.26-1.04) (Fig. 3). At delivery, mean maternal fentanyl concentration was 0.58 ng/ml (0.14-1.18); mean umbilical arterial and venous concentrations were 0.51 ng/ml (0.04-1.8) and 0.41 ng/ml (0.18- 1.2), respectively. Rare results of fentanyl concentrations > 1.0 ng/ml correlated neither with sedation, maternal respiratory rate and side effects, nor with Apgar scores and umbilical blood gas values. No Apgar score at 5 min was below 9, and no umbilical pH was below 7.20. (ABSTRACT TRUNCATED)