We appreciate the interest in our work and comments expressed by Drs. Rosman and Connelly et al. with regard to our recent article. 1Rosman expresses concern that the collective effects of subtle, nonsignificant differences between the combined spinal–epidural (CSE) and epidural groups could lead to an observation bias in the rate of cervical dilation. The influence of the factors he cites (earlier rupture of membranes and more aggressive use of oxytocin) is controversial but has been found to be unimportant in a large randomized trial from our institution. 2Nonetheless, a multivariate analysis of initial cervical dilation rate versus analgesic group, controlling for use of oxytocin before analgesia, artificial or spontaneous rupture of membranes, and rupture of membranes before analgesia, with or without first-order interactions, still found group assignment to be a highly significant determinant (main effects model, P = 0.0024). Moreover, the difference between groups in examination frequency was not significant clinically (approximately 20 min) or statistically. The time after analgesia to the next examination was also not significant (CSE group, 1.6 ± 1.3 h, vs. epidural group, 1.6 ± 1.3 h, P = 0.92). Therefore, we do not believe the examination frequency could have significantly altered the observed rates of dilation. Rosman’s concern regarding the length of ruptured membranes before analgesia is also unlikely to have influenced our observations. The inclusion of only those patients in whom ruptures occurred before analgesia (as opposed to all patients, even if ruptures occurred after analgesia, as reported in the original article) shows that in the CSE group ruptures occurred closer to the time of analgesia initiation than in the epidural group (5.2 ± 4.2 h vs. 6.8 ± 2.6 h, P = 0.04). In summary, although it is not practical to standardize every aspect of labor management given the uncertainties of nulliparous labor, we believe that there were no important differences that could have materially altered our results.
In addition, because a dural puncture is part of the CSE technique, we assume that Rosman is referring to the potential for headaches as a complication after regional techniques; of interest, use of the CSE technique has been suggested to prevent dural puncture with the larger epidural needle because it allows for confirmation of the dural space with a smaller needle. Although we cannot make any strong conclusions about potential complications of these two techniques because our study was not designed for that purpose and remains underpowered to make robust conclusions of that nature, we did not observe any differences in postdural puncture headache, fetal bradycardia, maternal hypotension, nausea, pruritus, or excessive blockade.
Connelly et al. suggest that our findings could be the result of epidural analgesia slowing the progress of labor, rather than CSE analgesia enhancing it. Although this is a possibility, we do not believe it to be likely nor supported by their previous work. 3Our epidural group experienced a mean cervical dilation rate of 1.3 ± 0.7 cm/h for the first stage of labor in nulliparous women, a rate considered to be normal by Friedman, 4the American College of Obstetricians and Gynecologists, 5and major, recent obstetric texts. 6Of note, although the timing of cervical dilation relative to analgesia was neither specified nor standardized in the work by Dunn et al. , 3it appears that the mean cervical dilation in both their intrathecal and epidural groups was even slower than in our epidural group.