Whether nitrous oxide (N2 O) increases the incidence of vomiting has been a contentious issue for at least 25 years. Fortunately, so many investigations have addressed the question [1-27] that a straightforward and simple analysis of their findings yields a conclusive answer.
Standard procedures (electronic search, follow-up on references, serendipity) were used to collect reports of the incidence of emesis in patients that received N2 O versus anesthetics or analgesics without N2 O. Each paper was classified as having found or not found more frequent vomiting in N2 O patients, and the statistical significance of each such finding was classified as P < 0.05 or not < 0.05. Studies for which the incidence of vomiting could not be distinguished from the incidence of nausea (e.g., ) were not included because the incidence of vomiting could not be determined, and one study was eliminated because "nausea requiring therapeutic intervention" was treated with antiemetics, which presumably altered the subsequent distribution of emesis .
The primary analysis consisted of calculating the binomial probability that chance could account for the number of studies in which emesis was more frequent in patients who received N2 O. Assuming that N2 O has no effect on vomiting (P = 0.5 that a study will show an increased incidence), this is like calculating the probability that an observed number of heads or tails would occur on any number of flips of a fair coin (ties were counted as evidence in favor of the null hypothesis).
For investigations that failed to find a statistically significant association between N2 O and emesis, the presence or absence of the minimum acceptable statistical power (1 - beta = 0.80) to detect a 20% difference in incidence was determined.
Twenty-seven studies were found in which the incidence of emesis in patients who received N2 O with or without other anesthetics or analgesics could be compared to the incidence of emesis in patients who received anesthetics or analgesics without N2 O [1-27]. In 24 of those 27 investigations, patients who received N2 O had an absolutely higher incidence of vomiting (Table 1, P < 0.00005).
Patients given N2 O did not have an absolutely higher incidence of emesis in three studies. In the first of those, a higher incidence of emesis was found in patients who received cyclopropane . The second study was a tie, with 2 of 18 patients in a ketamine-only group vomiting and 2 of 18 patients in a ketamine-N (2) O group vomiting . In the third study , 11 of 21 halothane-N2 O patients vomited (52.4%) versus 10 of 19 halothane-only patients (52.6%).
Of the 24 studies that evidenced more frequent emesis among patients who received N2 O, 12 found the association to be statistically significant. Among the investigations that failed to find a statistically significant association between N2 O and emesis, only two had statistical power >0.80 to detect a 20% effect size. One of those two included significantly more patients (P < 0.03) with a history of vomiting after anesthesia in its non-N2 O groups but still found an absolutely higher incidence of vomiting among its N2 O patients . The other study  stated a directional hypothesis: "We studied the hypothesis that administration of nitrous oxide causes … vomiting." Although a directional hypothesis warrants a one-tailed statistical test, the authors chose to simultaneously test the hypothesis that N2 O has antiemetic properties, which increased their P value to 0.059 . Had they performed a one-way Fisher's exact test, their observed association between N2 O and emesis would have been statistically significant at P < 0.036.
The liability inherent in the methodology used here is that relevant studies may not have been obtained for examination. However, few such investigations are likely to have been missed, and the minimum number of studies required to dissipate the statistical significance of the current finding is 10--each of which would need to be an investigation in which patients receiving N2 O did not have a higher incidence of emesis.
Every statistical analysis makes a trade-off between power and robusticity. A powerful analysis is one that has a low probability of generating a false-negative inference (type II error) as a consequence of artifacts inherent in the analysis itself (primarily assumptions required by the analysis). In distinction, a robust analysis is one that has a low probability of causing a false-positive inference (type I error). The analysis used here has low power and high robusticity. In less formal terms, it is a crude instrument in the sense that it is likely to fail to detect a true but weak (small effect size) relationship. However, when a robust analysis does detect a relationship, as is the case here (P < 0.00005), one can be confident that such a finding is not an artifact of assumptions required by that analysis.
Meta-analysis, in contrast, is relatively powerful but lacks robusticity. That is, because meta-analyses require more potentially invalidating assumptions and decisions, they are more likely to generate false-positive inferences. Nevertheless, for the purpose of addressing subanalyses, such as which groups of patients or procedures are most susceptible to a drug or intervention that has been shown to have an effect, meta-analysis is indispensable. The above notwithstanding, for the purpose of addressing the primary question of whether a drug or intervention has an effect, the robusticity of a binomial probability cannot be surpassed.
The data imply that N2 O, perhaps the most ubiquitously used drug in anesthesiology, produces a greater incidence of emesis, perhaps the most common postoperative morbidity, than other currently used anesthetics (P < 0.00005). This relationship is undoubtedly variable according to context and circumstances. Those contexts and circumstances warrant investigation without hinderance from the prevailing premise that N2 O has not been shown to increase the incidence of emesis ([30,31] cf. ). That is, we need to move from the question "Does N2 O cause vomiting?" to "When, why, in whom, and under what circumstances does N2 O cause vomiting?"
1. Mehta V, Ratra CK, Badola RP, Bhargava KP. Role of different anaesthetic techniques in the incidence of early postanesthetic sickness. Br J Anaesth 1969;41:689-94.
2. McKie BD. Postoperative vomiting: the effects of premedication, anaesthetic and oxytocic drugs. Med J Aust 1969;June 14:1236-8.
3. Jones PL, Rosen M, Mushin WW, Jones EV. Methoxyflurane and nitrous oxide as obstetric analgesics. I. A comparison by continuous administration. BMJ 1969;3:255-9.
4. Jastak JT, Goretta C. Ketamine HCL as a continuous-drip anesthetic for outpatients. Anesth Analg Curr Res 1973;52:341-4.
5. Jastak JT, Paravecchio R. An analysis of 1331 sedations using inhalation, intravenous, or other techniques. J Am Dent Assoc 1975;91:1242-9.
6. Muir VM, Leonard M, Haddaway E. Morbidity following dental extraction. Anaesthesia 1976;31:171-80.
7. Alexander GD, Skupski JN, Brown EM. The role of nitrous oxide in postoperative nausea and vomiting [abstract]. Anesth Analg 1984;63:175.
8. Lonie DS, Harper NJN. Nitrous oxide and vomiting. The effect of nitrous oxide on the incidence of vomiting following gynecological laparoscopy. Anaesthesia 1986;41:703-7.
9. Martin J, Williams D, Weis FR Jr. Comparison of three anesthetic techniques on emetic symptoms using sufentanil for outpatient surgery. J Am Assoc Nurse Anesth 1987;55:245-9.
10. Gibbons P, Davidson P, Adler E. Nitrous oxide does not affect postoperative vomiting in pediatric eye surgery [abstract]. Anesthesiology 1987;67(Suppl):A540.
11. Korttila K, Hovorka J, Erkola O. Nitrous oxide does not increase the incidence of nausea and vomiting after isoflurane anesthesia. Anesth Analg 1987;66:761-5.
12. Melnick BM, Johnson LS. Effects of eliminating nitrous oxide in outpatient anesthesia. Anesthesiology 1987;67:982-4.
13. Muir JJ, Warner MA, Offord KJ, et al. Role of nitrous oxide and other factors in nausea and vomiting. A randomized and blinded perspective study. Anesthesiology 1987;66:513-8.
14. Bloomfield E, Hilberman M, Brown P, et al. Postoperative nausea and vomiting. A comparison of sufentanil, nitrous oxide, and isoflurane. Cleve Clin J Med 1988;55:549-52.
15. Sengupta P, Plantevin OM. Nitrous oxide and day-case laparoscopy. Effects on nausea, vomiting, and return to normal activity. Br J Anaesth 1988;60:570-3.
16. Hovorka J, Korttila K, Erkola O. Nitrous oxide does not increase nausea and vomiting after gynecological laparoscopy. Can J Anaesth 1989;36:145-8.
17. Eger EI II, Lampe GH, Wauk LZ, et al. Clinical pharmacology of nitrous oxide: an argument for its continued use. Anesth Analg 1990;71:575-85.
18. Felts JA, Poler SM, Spitznagel EL. Nitrous oxide, nausea, and vomiting after outpatient gynecologic surgery. J Clin Anesth 1990;2:168-71.
19. Ranta P, Nuutinen L, Laitenen J. The role of nitrous oxide in postoperative nausea and recovery in patients undergoing upper abdominal surgery. Acta Anaesth Scand 1991;35:339-41.
20. Smiley BAS, Paradise NF. Does the duration of N2
O administration affect postoperative nausea and vomiting? Nurse Anesth 1991;2:13-8.
21. Watcha MF, Simeon RM, White PF, Stevens JL. Effect of propofol on the incidence of postoperative vomiting after strabismus surgery in pediatric outpatients. Anesthesiology 1991;75:204-9.
22. Lim BL, Low TC. Total intravenous anaesthesia vs. inhalational anaesthesia for dental day surgery. Anaesth Intensive Care 1992;20:475-8.
23. Wilson IG, Fell D. Nitrous oxide and postoperative vomiting in children undergoing myringotomy as a day case. Paediatr Anaesth 1993;3:283-5.
24. Reimer EJ, Montgomery CJ, Becan JC, et al. Propofol anaesthesia reduces early postoperative emesis after paediatric strabismus in surgery. Can J Anaesth 1993;40:927-33.
25. Pandit UA, Malviya S, Lewis IH. Vomiting after outpatient tonsillectomy and adenoidectomy in children: the role of nitrous oxide. Anesth Analg 1995;80:230-3.
26. Splinter W, Roberts DJ, Rhine EJ, et al. Nitrous oxide does not increase vomiting in children after myringotomy. Can J Anaesth 1995;42:274-6.
27. Alexander G, Bellefleur J, Brown M. The role of nitrous oxide in postoperative nausea and vomiting with desflurane and isoflurane [abstract]. Anesthesiology 1995;83:A27.
28. Pandit U, Pryn S, Randel G, et al. Nitrous oxide does not increase postoperative nausea/vomiting in pediatric out-patients undergoing tonsillectomy-adenoidectomy [abstract]. Anesthesiology 1990;73(Suppl):A1245.
29. Sukhani R, Lurie J, Jabamoni R. Propofol for ambulatory gynecologic laparoscopy: does omission of nitrous oxide alter postoperative emetic sequelae and recovery? Anesth Analg 1994;78:831-5.
30. Watcha MF, White PF. Postoperative nausea and vomiting. Anesthesiology 1992;77:662-84.
31. Watcha MF, White PF. In reply to Hartung [letter]. Anesthesiology 1993;78:405-6.
32. Hartung J. Nitrous oxide--it's enough to make you vomit [letter]. Anesthesiology 1993;78:403-5.