To the Editor:
Gambling et al.  compared postcesarean outcomes of epidural morphine 3 mg plus epidural butorphanol 3 mg versus epidural morphine 3 mg alone. Patients receiving butorphanol had less analgesia, an equivalent need for treatment of pruritus, and equivalent satisfaction compared with controls. In an earlier study , we had compared postcesarean outcomes of epidural morphine 4 mg plus epidural butorphanol 3 mg versus epidural morphine 4 mg alone. Patients receiving butorphanol in our study had significantly greater analgesia, a significantly lower incidence of treatment for pruritus, and significantly greater overall satisfaction compared with controls. We carefully reviewed both studies to determine why our results differed Table 1.
With 10-cm visual analog scales (VASs), numerical pain score values imply the following: 1) VAS > 3.0 represents inadequate analgesia; 2) VAS = 1-3 represents good analgesia; and 3) VAS < 1 represents extraordinary analgesia ("pain prevention") . Over the expected range of adequate-to-superb VAS analgesia scores (0-3 cm), a difference of 0.5 cm is clinically important. Specifically, postoperative pain scores < 1 cm generally require continuous epidural administration of local anesthetics [3,4]. VAS sensitivity depends on administering an anchored scale in a standardized manner, evaluating patients on multiple occasions, and obtaining a uniform distribution of scores .
With respect to cumulative pain assessment, our study group was unique in achieving "pain prevention" (mean VAS = 0.7 cm) with an extremely narrow distribution of low scores (SD = 1.0 cm) that was both statistically and clinically significant. Statistical analysis of VAS scores involved pairwise comparisons of each study group with the control group using one-way analysis of variance. Furthermore, chi squared analysis of 265 evaluations among the two groups demonstrated that butorphanol significantly reduced the incidence of inadequate analgesia to a level (4%) threefold less than control (11%). Our post hoc power analysis demonstrated that 265 evaluations among two groups produced a statistically significant difference (P < 0.05) in mean VAS pain scores of less than 0.5 cm.
In contrast, the study design of Gambling et al.  presumed that 102 evaluations among two groups produced a power of 0.8 to detect a difference (P < 0.005) in mean VAS scores of 2.0 cm. Nearly a third of their pain score evaluations claimed inadequate analgesia (VAS pain score > 3). The duration of analgesia after epidural morphine administration does not reliably extend to 24 h [2,6], so one third of their patient assessments occurred at a time (24 h) when epidural opioid analgesia was dissipating. Overall, the lack of statistically significant differences in the study by Gambling et al.  appears to be due to inadequate numbers of patient evaluations (no VAS satisfaction scores were presented), unusually high VAS pain scores [2,6], and statistical methods that were biased toward unreasonable limits of detection.
Bernard Wittels, MD, PhD
Alicia Toledano, ScD
Department of Anesthesia and Critical Care, The University of Chicago, Chicago, IL 60637
1. Gambling DR, Howell P, Huber C, Kozak S. Epidural butorphanol does not reduce side effects from epidural morphine after cesarean birth. Anesth Analg 1994;78:1099-104.
2. Wittels B, Glosten B, Faure EAM, et al. Opioid antagonist adjuncts to epidural morphine for postcesarean analgesia: maternal outcomes. Anesth Analg 1993;77:925-32.
3. Armitage EN. Postoperative pain--prevention or relief. Br J Anaesth 1989;63:136-7.
4. Cohen S, Amar D, Pantuck CB, et al. Postcesarean delivery epidural patient-controlled analgesia: fentanyl or sufentanil? Anesthesiology 1993;78:486-91.
5. Paige D, Cioffi AM. Pain assessment and measurement. In: Sinatra RS, Hord AH, Ginsberg B, Preble LM, eds. Acute pain: mechanisms and management. St. Louis: Mosby-Year Book, 1992:609-12.
6. Abboud TK, Afrasiabi A, Davidson J, et al. Prophylactic oral naltrexone with epidural morphine: effect on adverse reactions and ventilatory responses to carbon dioxide. Anesthesiology 1990;72:233-7.