A comparative random-allocation graded dose-response study of epidural bupivacaine and ropivacaine in laboring nulliparous women
Abstract Number: 249
Abstract Type: Original Research
Previous studies that have compared the potency of epidural bupivacaine and ropivacaine for labor analgesia have mostly utilized up-down methodology (UDM) to compare ED50 (1,2). Few data are available for more clinically relevant values of ED90 or ED95. Furthermore, UDM provides no information about the comparative slopes of the dose-response curves. We performed a random-allocation graded dose-response study of epidural bupivacaine and ropivacaine. Full dose-response curves were defined and potency at mid and higher points on the curves were compared.
IRB approval and written informed consent were obtained. In a randomized, double-blinded trial, 300 nulliparous patients in painful labor were given an epidural dose of bupivacaine (5, 10, 15, 20 30 or 40 mg) or ropivacaine (7, 15, 20, 30, 45 or 60 mg) in 20 ml saline. Visual analog scale pain scores were recorded for 30 min. Response was defined by the percentage decrease in pain score from baseline at 30 min and dose-response data were analyzed using nonlinear regression.
Sigmoidal Emax model dose-response curves (figure) were fitted to the datasets for bupivacaine (R2 = 0.53) and ropivacaine (R2 = 0.59). The curves had similar steepness (Hill coefficient 2.02 (95% CI: 1.55 - 2.50) vs 2.25 (1.70 - 2.79), P = 0.55). The D50 (dose that reduces pain score to 50% of baseline at 30 min) of ropivacaine was greater than that of bupivacaine (15.3 (95% CI: 13.7 - 17.1) mg vs 11.3 (10.0 - 12.7) mg, P = 0.0003) but D90 was similar 40.6 (32.4 - 51.1) mg vs 33.4 (26.2 - 42.7) mg, P = 0.29). The potency ratio at D50 for ropivacaine:bupivacaine was 0.75 (95% CI: 0.65 - 0.88).
Ropivacaine is less potent than bupivacaine but their dose-response characteristics are otherwise similar. The difference in potency is not statistically significant at D90 doses. Although our use of random-allocation graded dose-response methodology with nonlinear regression analysis is more demanding of time and resources than UDM, it has many advantages. Complete dose-response curves can be characterized and compared and pain score data measured on a graded scale are fully utilized. This contrasts with other methods such as UDM and probit or logit regression that require transformation of response data to dichotomous values ("success" or "failure"). Such transformations are inefficient and result in loss of power.
1. Anesthesiology 1999; 90: 944-50.
2. Br J Anaesth 1999; 82