This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Shafer, S. L. Articles by Nekhendzy, V. Search for Related Content PubMed PubMed Citation Articles by Shafer, S. L. Articles by Nekhendzy, V. Related Collections Pain Medicine Pain Pharmacology

---

EDITORIAL

Anesthesia Matters: Statistical Anomaly or New Paradigm?

Steven L. Shafer, MD^*, and Vladimir Nekhendzy, MD

*Editor-in-Chief, Anesthesia & Analgesia. From the Department of Anesthesiology, Columbia University, New York, New York; and Clinical Associate Professor of Anesthesia and Otolaryngology, Department of Anesthesia, Stanford University, Stanford, California.

Address correspondence to Steven L. Shafer, MD, Department of Anesthesiology, Columbia University Medical Center, 622 W. 168th St, PH 5-505, New York, NY 10032-3725. Address email to sshafer{at}columbia.edu.

Every so often we are startled to discover something new about a topic we thought we understood completely. Last month one of us (SLS) learned about the "Goldbach conjecture," which your children probably already know. The Goldbach conjecture states that every even number is the sum of two primes. It is unproven, but there are no counterexamples below 10¹⁸. Having played with even and prime numbers since elementary school, how could anyone miss something so obvious?

In this month’s issue of Anesthesia & Analgesia, Cheng et al. suggest we have missed something similarly familiar: maintenance of anesthesia with propofol is associated with less postoperative pain than maintenance of anesthesia with isoflurane.1 Is this possible? After nearly three decades of research specifically comparing the recovery characteristics of propofol and isoflurane, could such an obvious finding be missed?

The short answer is "yes." There are many examples in medicine of important clinical findings being missed for decades. After nearly 40 years of apparently safe spinal administration, anesthesiologists discovered that lidocaine was potentially neurotoxic in the intrathecal space.2,3 However, postoperative pain is fundamentally different from lidocaine spinal toxicity. Postoperative pain is not a rare event. It is a frequently measured outcome. Haven’t the many clinical trials comparing isoflurane and propofol recovery already measured postoperative pain and addressed this question?

The short answer is "yes, and no." As of today (October 18, 2007), PubMed lists 644 clinical trials with the search terms "isoflurane propofol." Our quick review identified about a dozen that measured postoperative pain and opioid consumption. Nearly all of these reported no difference. This is not unexpected. Pain is highly variable. Unless studies are powered to examine pain as an outcome, and incorporate repeated assessments of postoperative pain over time to separate intersubject from intrasubject variability, differences in postoperative pain may easily be missed.

The few previous studies that found differences do not agree. Boccara et al. compared postoperative pain and analgesic requirements in patients receiving propofol for maintenance with patients receiving isoflurane.4 Patients receiving propofol had increased pain and opioid requirements for the first 6 hours compared with patients receiving isoflurane, exactly the opposite of the findings of Cheng et al., but equally surprising. In patients with sleep apnea undergoing uvulopalatopharyngoplasty, Hendolin et al. found that propofol significantly reduced pain in the second hour compared with patients receiving isoflurane,5 corroborating the results of the present study.

Here we must disclose our conflict of interest. We are both avid practitioners of propofol-based anesthesia, nearly to the exclusion of inhaled anesthetics. We practice this way, in part, because our N = 2 consensus is that patients awakening from propofol have less pain than patients awakening from inhaled anesthetics. We would like these findings to be true, but what we would like counts for nothing. Bayesian inference and common sense demand that extraordinary findings be supported by extraordinary evidence.6 A single study is inadequate to reach the conclusion that propofol is associated with less postoperative pain than isoflurane, despite a P value less than 0.01.1 Clinical investigators need to repeat this study to see whether the difference in postoperative pain is a reproducible finding.

Unexpected results merit scrutiny, which helps guide future studies. Cheng et al. performed only two Visual Analog Scale assessments after the first hour: one at 2 hours, and one at 24 hours. Postoperative pain scores were identical at 2 hours, but diverged again at 24 hours. Is this a fluke, or does it represent the pharmacology of propofol and isoflurane? Can the pharmacodynamic difference between the propofol and isoflurane patient groups be explained by a hyperalgesic effect of isoflurane,7 a hitherto undocumented analgesic effect of propofol, or both? The present study was limited to nonsmokers. Is this difference in postoperative pain also seen in smokers? The protocol was designed to keep patients at the same anesthetic depth by providing strict guidelines for anesthetic titration. The fentanyl doses were comparable among the groups. Although the protocol required titration to a Bispectral Index of 50, the actual Bispectral Index values were not recorded. Modest differences might affect the outcome.

Are we ready to embrace the findings of Cheng et al.’s study and add a postoperative analgesic benefit of propofol to a long list of its virtues, such as a significant reduction of postoperative nausea and vomiting8 and cellular and organ-protective effects caused by scavenging of free radicals and protection against peroxidative injury?9–11 Or are the present findings a statistical anomaly? We can only tell with prospective studies. If future studies are negative, they will demonstrate that this is an anomaly. If future studies are positive, they will presage a paradigm change for improving postoperative pain management. Either way, Anesthesia & Analgesia invites submission of adequate and well controlled studies that address this question. Should a paradigm shift result, we hope you will have read about it first, and hopefully second, third, and fourth, in the pages of this journal.

	Footnotes

 
Accepted for publication October 19, 2007.

	REFERENCES

Top REFERENCES

 

1. Cheng, SS, Yeh J, Flood P. Anesthesia matters: patients anesthetized with propofol have less postoperative pain than those anesthetized with isoflurane. Anesth Analg 2008;106:264–9[Abstract/Free Full Text]
2. Rigler ML, Drasner K, Krejcie TC, Yelich SJ, Scholnick FT, DeFontes J, Bohner D. Cauda equina syndrome after continuous spinal anesthesia. Anesth Analg 1991;72:275–81[Abstract/Free Full Text]
3. Drasner K. Lidocaine spinal anesthesia: a vanishing therapeutic index? Anesthesiology 1997;87:469–72[Web of Science][Medline]
4. Boccara G, Mann C, Pouzeratte Y, Bellavoir A, Rouvier A, Colson P. Improved postoperative analgesia with isoflurane than with propofol anaesthesia. Can J Anaesth 1998;45:839–42[Web of Science][Medline]
5. Hendolin H, Kansanen M, Koski E, Nuutinen J. Propofol-nitrous oxide versus thiopentone-isoflurane-nitrous oxide anaesthesia for uvulopalatopharyngoplasty in patients with sleep apnea. Acta Anaesthesiol Scand 1994;38:694–8[Web of Science][Medline]
6. Shafer SL. Did our brains fall out? Anesth Analg 2007;104:247–8[Free Full Text]
7. Flood P, Sonner JM, Gong D, Coates KM. Isoflurane hyperalgesia is modulated by nicotinic inhibition. Anesthesiology 2002;97:192–8[Web of Science][Medline]
8. Visser K, Hassink EA, Bonsel GJ, Moen J, Kalkman CJ. Randomized controlled trial of total intravenous anesthesia with propofol versus inhalation anesthesia with isoflurane-nitrous oxide: postoperative nausea with vomiting and economic analysis. Anesthesiology 2001;95:616–26[Web of Science][Medline]
9. Kevin LG, Novalija E, Stowe DF. Reactive oxygen species as mediators of cardiac injury and protection: the relevance to anesthesia practice. Anesth Analg 2005;101:1275–87[Abstract/Free Full Text]
10. Zhang SH, Wang SY, Yao SL. Antioxidative effect of propofol during cardiopulmonary bypass in adults. Acta Pharmacol Sin 2004;25:334–40[Web of Science][Medline]
11. Rohm KD, Suttner SW, Boldt J, Schollhorn TA, Piper SN. Insignificant effect of desflurane-fentanyl-thiopental on hepatocellular integrity–a comparison with total intravenous anaesthesia using propofol-remifentanil. Eur J Anaesthesiol 2005;22:209–14[Web of Science][Medline]

This article has been cited by other articles:

				ERRATUM Anesth. Analg., December 1, 2008; 107(6): 2037 - 2037. [Full Text] [PDF]

				S. L. Shafer Full Disclosure Matters! Anesth. Analg., March 1, 2008; 106(3): 1017 - 1017. [Full Text] [PDF]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Shafer, S. L. Articles by Nekhendzy, V. Search for Related Content PubMed PubMed Citation Articles by Shafer, S. L. Articles by Nekhendzy, V. Related Collections Pain Medicine Pain Pharmacology