Department of Anesthesiology, Women’s Hospital, University of Michigan Health System, Ann Arbor
January 17, 2003.
Address correspondence and reprint requests to Linda S. Polley, MD, Department of Anesthesiology, 1500 East Medical Center Dr., F3900 Mott Children’s Hospital, Ann Arbor, MI 48109-0211. Address e-mail to email@example.com.
Epidural bupivacaine provides excellent analgesia for labor and delivery and remains the most widely used local anesthetic in obstetric anesthesia. However, disadvantages include cardiac and central nervous system toxicity with accidental IV injection of large doses and the potential for motor blockade, particularly with the use of higher concentrations. These concerns have prompted the search for alternative drugs. Ropivacaine is an amino amide local anesthetic that is structurally similar to bupivacaine. Originally developed in the 1950s (1), it was introduced for commercial use in 1996 in hopes that it would have lesser potential for cardiovascular toxicity and the side effect of motor blockade. Ropivacaine has been studied to determine its suitability for labor analgesia and has undergone numerous clinical comparisons to bupivacaine. Studies using the minimum local analgesic design (MLAC) have found that ropivacaine is approximately 40% less potent than bupivacaine (2–4). This is important because equipotent doses must be compared to draw meaningful conclusions regarding analgesic efficacy and side-effect profiles. Early studies comparing ropivacaine with bupivacaine used equal doses of both, whereas some recent investigations have used larger doses of ropivacaine to adjust for potency differences. In this current issue of the journal, Halpern and Walsh (5) present a meta-analysis of randomized, controlled trials comparing ropivacaine with bupivacaine to determine whether or not there is any difference in obstetric or neonatal outcome with the incidence of spontaneous vaginal delivery as the primary outcome. The authors correctly excluded the MLAC studies from this meta-analysis, which although are prospective, randomized designs, are not usually designed to assess any effect on obstetric outcome. They have concluded that there is no statistically significant difference between the two drugs in the incidence of any of the measured obstetric or neonatal outcomes. Because many of the component studies used supramaximal concentrations of local anesthetics, the meta-analysis could not determine if there was any difference in the incidence of motor blockade.
In 1998, Writer et al. (6) published a prospective meta-analysis of six randomized, double-blinded studies of 403 laboring women who received epidural analgesia with ropivacaine or bupivacaine 0.25% wt/vol. The outcomes of interest included Apgar scores, neurological and adaptive capacity scores (NACS), degree of motor blockade, and mode of delivery. Results showed similar analgesia and drug consumption in the two groups. Early NACS did not differ between the groups, but there was a significant difference between the groups at 24 h, with fewer infants with NACS <35 in the ropivacaine group. Spontaneous vaginal deliveries occurred more frequently in the ropivacaine group (58% versus 49%;P < 0.05), instrumental deliveries less frequently (27% versus 40%;P < 0.01), and the frequency of cesarean delivery was similar between groups. The authors also reported that the intensity of motor blockade was less with ropivacaine, as assessed by a modified Bromage scale. There are several important limitations of this investigation. The component studies used large concentrations of local anesthetics no longer in routine clinical use. These supramaximal concentrations are at the top of the concentration-response curve and correspond to the upper flatter portion of the curve where differences between drugs are obscured. Both of the primary outcomes are flawed. The NACS has been shown to be an unreliable method of neonatal evaluation (7). Although delivery outcomes were reported, there was no standardization of obstetric practice regarding indications for instrumental delivery or cesarean delivery. The methods do not describe cervical dilation at the time of entry into the study. It has been clearly demonstrated that the degree of dilation affects local anesthetic requirements (8). Last, confidence intervals are lacking for the reported results.
The previous meta-analysis by Writer et al. (6) demonstrated significant advantages for ropivacaine in the frequency of spontaneous vaginal deliveries. The studies in the Writer et al. (6) meta-analysis used higher concentrations (0.25% wt/vol) than routinely used today. To investigate this further, it is useful to break down the latest Halpern and Walsh (5) meta-analysis into studies before and since 1998 and estimate the odds ratio (OR) and 95% confidence interval (CI) for spontaneous vaginal delivery. There is a significant advantage for ropivacaine (OR, 1.47; 95% CI, 1.00–2.18;P = 0.0499) in the earlier reports before 1998, which is then lost with the later studies (OR, 1.01; 95% CI, 0.81–1.27;P = 0.90).
Many of the concerns with the Writer et al. (6) investigation have been successfully addressed in the Halpern and Walsh (5) meta-analysis. The authors have analyzed 23 randomized trials and 2074 parturients. Whereas some of the studies use high concentrations of local anesthetic, the more recent studies used more clinically relevant concentrations. The primary outcome is the incidence of spontaneous vaginal delivery, and umbilical artery pH is used to assess fetal status rather than neurological and adaptive capacity scores. CIs are calculated and reported, which is essential to study result interpretation (9).
The authors are to be congratulated on the design and execution of the study. However, there seem to be some calculation errors in the primary outcome of spontaneous vaginal delivery, as in Figure 1 in the paper. It can be seen that there are 20 studies plotted on the Forest plot but with the summary data from 21 studies tabulated! On checking, it would seem that the 10th entry in the tabulated section is in error and should be deleted to keep the column totals correct. There also seem to be errors in data abstraction from approximately five of the studies, with important errors in the Gautier, Campbell, and Fernandez studies. These data have been corrected and are presented in the same sequence in Figure 1. Using a random effects design, the corrected DerSimonian-Laird pooled OR is 1.11 (95% CI, 0.92–1.35;P = 0.28). This estimate indeed provides more weight to the authors’ conclusions when compared with the ratio of 1.17 (95% CI, 0.96–1.44;P = 0.12) originally reported.
It is of interest to crudely look at the trends in this meta-analysis. As expected, there have been significant (P < 0.0001) reductions in the concentrations of local anesthetics used for epidural analgesia (Fig. 2). At the same time, there has been a trend, although not significant (P = 0.17), to an increase in the proportion of spontaneous vaginal deliveries over the same period (Fig. 3). This effect has been more pronounced for bupivacaine (F = 4.22;P = 0.055) than with ropivacaine (F = 0.31;P = 0.58) (Fig. 4). This latter point is most likely explained by the differences in motor-blocking potencies of the drugs, which is suggested by the authors. The epidural motor-blocking potency ratio of ropivacaine-to-bupivacaine has been estimated by Lacassie et al. (4) to be 0.66 using the MLAC design.
Finally, the specific limitations associated with meta-analysis bear mention. Pooled meta-analyses are most useful when high quality studies of similar methodology are analyzed and the technique essentially increases statistical power. Of concern in this study are the inclusion of older studies using large local anesthetic concentrations no longer clinically relevant and the inclusion of both single bolus, multiple bolus, and continuous infusion studies that may have pharmacokinetic differences. The majority of studies compare equal concentrations of ropivacaine and bupivacaine, with two studies adjusting for potency differences.
The fundamental question is whether ropivacaine offers significant advantage over bupivacaine for labor analgesia. As detailed in the article, numerous studies have demonstrated comparable analgesia with both drugs using patient-controlled epidural infusions throughout labor (10–12). Because dilute concentrations of local anesthetics are used for labor analgesia and bupivacaine is rarely used for anesthesia for cesarean delivery, toxicity concerns are not as important as for other anesthetic procedures, such as brachial plexus blockade where higher concentrations are in routine use. The remaining question regarding ropivacaine in obstetrics concerns the incidence of motor blockade compared with bupivacaine. This could not be effectively answered in this study because of the statistical heterogeneity in the component study results. More research is required comparing ropivacaine and bupivacaine at clinically relevant and equipotent doses to determine if there is any real difference in the incidence of motor blockade. However, it is unlikely that any difference between the drugs has substantial clinical relevance, because there is no difference in the incidence of instrumental or cesarean delivery.
Regardless of the possible limitations, the authors are to be congratulated on their rigorous approach to evaluating and combining the results of conflicting studies in an area of current clinical interest. This meta-analysis is clearly the largest and most thorough effort to date to evaluate the important clinical outcomes of the two local anesthetics in obstetric anesthesia. We agree with the authors’ primary conclusion that both ropivacaine and bupivacaine provide excellent labor analgesia with no significant differences between the two drugs in the incidence of spontaneous vaginal delivery or other measured obstetric and neonatal outcomes. It is clear from this meta-analysis that attention to relative potency issues leads to the reduction and optimization of dosing of local anesthetics and allows for meaningful comparison.
1. Columb MO. Clinical trials: up-down sequential allocation. In: Reynolds F, ed. Obstetric anaesthesia: millenium update. London: Springer-Verlag, 2000: 162–3.
2. Polley LS, Columb MO, Naughton NN, et al. Relative analgesic potencies of ropivacaine and bupivacaine for epidural analgesia in labor: implications for therapeutic indexes. Anesthesiology 1999; 90: 944–50.
3. Capogna G, Celleno D, Fusco P, et al. Relative potencies of bupivacaine and ropivacaine for analgesia in labour. Br J Anaesth 1999; 82: 371–3.
4. Lacassie HJ, Columb MO, Lacassie HP, et al. The relative motor blocking potencies of epidural bupivacaine and ropivacaine in labor. Anesth Analg 2002; 95: 204–8.
5. Halpern SH, Walsh VB. Epidural ropivacaine vs. bupivacaine for labor: a meta-analysis. Anesth Analg 2003; 96: 1473–9.
6. Writer WD, Stienstra R, Eddleston JM, et al. Neonatal outcome and mode of delivery after epidural analgesia for labour with ropivacaine and bupivacaine: a prospective meta-analysis. Br J Anaesth 1998; 81: 713–7.
7. Halpern SH, Littleford JA, Brockhurst NJ, et al. The neurological and adaptive capacity score is not a reliable method of newborn evaluation. Anesthesiology 2001; 94: 958–62.
8. Capogna G, Celleno D, Lyons G, et al. Minimum local analgesic concentration of extradural bupivacaine increases with progression of labour. Br J Anaesth 1998; 80: 11–3.
9. Altman DG. Clinical trials and meta-analyses. In: Altman DG, Machin D, Bryant TN, Gardner MJ, eds. Statistics with confidence: confidence intervals and statistical guidelines. London: BMJ Books, 2000: 120–38.
10. Owen MD, Thomas JA, Smith T, et al. Ropivacaine 0.075% and bupivacaine 0.075% with fentanyl 2 μg/mL are equivalent for labor epidural analgesia. Anesth Analg 2002; 94: 179–83.
11. Campbell DC, Zwack RM, Crone LL, et al. Ambulatory labor epidural analgesia: bupivacaine vs. ropivacaine. Anesth Analg 2000; 90: 1384–9.
12. Meister GC, D’Angelo R, Owen M, et al. A comparison of epidural analgesia with 0.125% ropivacaine with fentanyl versus 0.125% bupivacaine with fentanyl during labor. Anesth Analg 2000; 90: 632–7.