Materials and Methods

Institutional review board approval was obtained; 270 women were enrolled into a double-masked randomized trial that compared 3 commonly used approaches to labor analgesia ( Figure ). Nulliparous women at >36 weeks’ gestational age who were in spontaneous labor and requested epidural anesthesia were included. Women with contraindications to regional spinal anesthesia, and cervical dilation >5 cm, nonvertex fetal presentation, and without a category I fetal tracing, multiple gestations, and fetal malformations were excluded.

Flow diagram of patient enrollment

Haydon. Outcomes and satisfaction with patient-controlled epidural analgesia. Am J Obstet Gynecol 2011.

A combined spinal epidural approach was used with an 18G Tuohy needle followed by dural puncture with a 27G needle through a needle approach. Subjects in each group initially were given an intrathecal bolus of bupivacaine 0.5% (0.4 mL) and fentanyl 50 μg/mL (0.4 mL). Before the initial bolus, the subject’s verbal pain score (least pain, 0; most pain, 10) was recorded. The epidural catheter was then inserted, and a preprogrammed pump with 1 of the 3 dosing regimens (Painsmart Curlin Medical, Huntington Beach, CA) was connected. At this point, patients were assigned randomly to 1 of 3 groups, and the pump was initiated in all patients. The subject, nurse, anesthesiologist, and obstetrician were each blinded to the group assignment. Identical pumps that were identified uniquely were brought to the bedside by the research nurse who was not involved in the subject’s care. The pumps were each preprogrammed with the appropriate protocol. Each patient was given a patient-controlled button for demand dosing and provided instructions on its use. The use of the PCEA bolus button was acknowledged by the pump and recorded the subject’s request, regardless of patient group assignment. This included the patients in group I (the CEI-only group) who had a PCEA button to push, despite it being a sham bolus. Group I (CEI-only) was started immediately on a continuous pump infusion of bupivacaine 0.1% and fentanyl 2 μg/mL at 10 mL/hr. Group II (CEI + PCEA; CEI at 10 mL/hr plus PCEA 10 mL, 20-minute lockout) was given a continuous pump infusion of bupivacaine 0.1% and fentanyl 2μg/mL at 10 mL/hr plus the opportunity to administer a bolus (10 mL) of bupivacaine 0.1% and fentanyl 2 μg/mL every 20 minutes. Group III (PCEA-only; PCEA 10 mL, 20-minute lockout) was given a bolus (10 mL) of bupivacaine 0.1% and fentanyl 2μg/mL every 20 minutes at their request with no continuous infusion. Subjects in each group were allowed up to 2 total boluses of 10 mL bupivacaine 0.25% by the anesthesiologist in the event the patient requested further analgesia. If pain control was not adequate after these boluses by the anesthesiologist, the epidural catheter was replaced. Each hour, the nursing staff documented the subject’s verbal pain score. Additionally, an hourly modified Bromage score was recorded (no motor block, 0; unable to raise extended leg, 1; just able to flex knees, 2; foot movement only, 3).

After delivery, total anesthetic dosing and the number of bolus requests, anesthesia interventions, and obstetric and neonatal outcomes were evaluated; patient satisfaction questionnaires were administered. Subjects were asked to rate their pain experience on a scale of 0-100 (absent or none, 0; present or severe, 100). The scale of zero for the most satisfied and 100 the least satisfied was used to correlate with the patient’s response to hourly pain scores. The questionnaires included responses regarding pain relief with the PCEA button and overall satisfaction. Responses were also collected regarding a feeling of self-control, which was described to the subject as the ability to exert self-control in ameliorating pain. The primary outcome was the dosage of local anesthetic that was used. Secondary outcomes included obstetric outcomes and maternal satisfaction.

Statistical analysis was performed with statistical computer software (version 12.0; SPSS Inc, Chicago, IL). Analysis of variance was used for continuous variables. The Kruskal-Wallis test was used for nonparametric parameters, and chi-square test was applied to test categoric parameters. A total of 270 subjects were assigned randomly into 1 of 3 treatment arms in a 1:1:1 ratio with the use of a computer-generated random number table. The randomization assignments were placed in sealed envelopes. Based on dosing criteria for bupivacaine that was shown by Ferrente et al with a power calculation to achieve 80% power with an α = .05, 79 patients in each group were required to detect a 20% difference in total anesthetic used among groups. We enrolled 90 in each group to account for the patients who would not complete the study because of technical difficulties. A probability value of < .05 was considered significant.