Henzi, Iris MD*; Walder, Bernhard MD†; Tramèr, Martin R. MD, DPhil*
Although dexamethasone has antiemetic properties in patients undergoing highly emetogenic chemotherapy (1), the mechanism of its antiemetic action is not well understood. A commonly held theory is that corticosteroids exert their antiemetic activity via prostaglandin antagonism (2). Others have suggested that the usefulness of dexamethasone in the control of chemotherapy-related emesis may be caused by the release of endorphins, resulting in mood elevation, a sense of well-being, and appetite stimulation (3).
More recently, dexamethasone has been added to a 5-HT3 receptor antagonist for use in chemotherapy (4,5). There are several reasons why this combination should be especially effective in controlling emetic symptoms. First, corticosteroids may reduce levels of 5-hydroxytryptophan in neural tissue by depleting its precursor tryptophan (6). Second, the antiinflammatory properties of corticosteroids may prevent the release of serotonin in the gut (7). And third, dexamethasone may potentiate the main effect of other antiemetics by sensitizing the pharmacological receptor (8). Thus, the combination of dexamethasone with a 5-HT3 receptor antagonist seems to be a logical choice for the control of chemotherapy-induced nausea and vomiting.
The role of dexamethasone in the surgical setting is less well understood. The first clinical trial that suggested that dexamethasone may prevent postoperative nausea and vomiting (PONV) was published in 1993 (9). Subsequent studies indicated that dexamethasone, alone or in combination with a 5-HT3 receptor antagonist, may indeed be an interesting alternative for the control of emetic symptoms in the postoperative period. The aim of this quantitative systematic review was to define the antiemetic efficacy and safety of dexamethasone in the prevention of PONV.
Systematic Search and Inclusion Criteria
We did a systematic search for full reports of randomized, controlled trials that tested the effect of prophylactic dexamethasone compared with any comparator (active or placebo) on PONV after general anesthesia. Relevant trials had to report end points of interest in dichotomous form (i.e., presence or absence of the end point with both dexamethasone and control). We searched MEDLINE (from 1966), COCHRANE Library (issue 2, 1999), and EMBASE (from 1982) databases without restriction to the English language and used different search strategies with the free text terms “dexamethasone;” “nausea,” “vomiting,” or “emesis”; “randomized” or “randomised”; “surgery,” “surgical,” or “postoperative”; and combinations of these words. The last electronic search was performed in April 1999. Additional trials were identified from reference lists of retrieved reports and by manually searching locally available anesthesia journals. Abstracts, letters, review articles, and animal data were not considered.
All authors independently read all reports that could possibly meet the inclusion criteria, and scored them for inclusion and methodological validity using the three-item, five-point Oxford scale (10). This score takes into account randomization, blinding, and description of withdrawals and dropouts. We met to reach a consensus by discussion. The minimal score of an included randomized controlled trial was one, the maximal score was five.
We took information about patients, surgery, dose, and route of administration of dexamethasone and comparators, study end points, and adverse effects from each included report. We extracted the cumulative incidence of PONV within 6 h and 24 h after surgery. Incidences of PONV during the two time periods (i.e., 0–6 h and 0–24 h) were used as indicators of “early” and “late” antiemetic efficacy, respectively. These two periods were chosen to gather as much information as possible on the antiemetic efficacy of dexamethasone. The early observation period represents the maximal time an ambulatory patient would stay in the recovery room before leaving the hospital. The late observation period would report on the cumulative incidence of PONV in the postoperative period. Antiemetic efficacy on “delayed” PONV (for instance, 6 to 24 h) were not analyzed, because such data were inconsistently reported. When several incidences of events were reported at different times, we analyzed the cumulative values nearest to the 6th and 24th postoperative hours. Two different PONV events, nausea and vomiting (including retching), both early and late, were extracted in dichotomous form. These events were treated separately.
We defined antiemetic efficacy as prevention of a PONV event with dexamethasone or control (11,12). We made calculations for individual trials, and by combining dexamethasone and control arms of independent trials. We used both relative benefit and number needed to treat as estimates of antiemetic efficacy. As an estimate of the statistical significance of a difference between dexamethasone and control, we calculated relative “benefits” as relative risks with 95% confidence intervals (CI) (13). For combined data, a fixed effect model that considers within-study variation (14) was used when data from no more than two trials were combined or when there was no significant heterogeneity (i.e., P > 0.1). In all other situations, we used a random effects model (15). As an estimate of the clinical relevance of a treatment effect, we calculated numbers needed to treat (i.e., the reciprocal of the absolute risk reduction) (16) for both individual trials and combined data. For combined data we used the weighted means of the experimental and control event rates, respectively. A positive number needed to treat indicated how many patients had to be exposed to dexamethasone to prevent one particular PONV event in one of them, who would have had this event had they all received the control intervention. A negative number needed to treat suggested superiority of control. Thus, the number needed to treat has the advantage of applicability to critical practice, and shows the effect required to achieve a particular therapeutic target. A 95% CI around the number needed to treat point estimate was obtained by taking the reciprocals of the values defining the 95% CI for the absolute risk reduction (17). In text and tables, the actual upper and lower limits of the 95% CI around the number needed to treat, regardless of whether they were positive or negative, are reported (18).
When the 95% CI around the relative benefit did not include one, we assumed a statistically significant difference between dexamethasone and control. In this case, we would expect the 95% CI around the number needed to treat to range from a positive limit to a negative limit, indicating that the confidence interval includes zero, and thus infinity. To estimate the additional risk of drug-related adverse effects, relative risk and number needed to harm as for number needed to treat (19) were calculated with 95% CI.
Excluded and Included Trials
We considered 19 trials for analysis. Two were subsequently excluded; in one (20), PONV was not a study end point, and in another (21), the combination of dexamethasone 150 μg/kg with a small-dose of ondansetron (50 μg/kg) was compared with a larger dose of ondansetron (150 μg/kg).
Seventeen trials with data from 1,946 patients (1,961 patients had initially been randomized) were analyzed (9,22–37). Of those, 598 received dexamethasone, 582 received ondansetron, granisetron, droperidol, metoclopramide, or perphenazine, 423 received a placebo, and 343 received a combination of dexamethasone with a 5-HT3 receptor antagonist (ondansetron or granisetron). The average number of patients per trial was 108 (range, 49 to 270), and per group was 55 (range, 22 to 135). The median validity score was 3 (range, 2 to 5). Ten trials were in adults, and seven were in children. Sixteen different dexamethasone regimens were tested, oral and IV, fixed doses (full milligrams), and variable doses (micrograms per kilogram body weight). In all trials, dexamethasone was given as a single prophylactic dose, either orally as a premedication or IV at induction. The most frequently used regimens of dexamethasone were 8 or 10 mg IV in adults, and 1 or 1.5 mg/kg IV in children. In most pediatric trials, the incidence of vomiting was the only end point.
Dexamethasone Versus Placebo
Dexamethasone was compared with placebo in four trials in adults and in three trials in children (Table 1, A and B, and Figure 1). In adults, dexamethasone 8 and 10 mg, orally or IV, was tested, in children, dexamethasone 0.5 mg/kg, 1 mg/kg, and 1.5 mg/kg IV. All results were statistically significant in favor of dexamethasone, except early nausea with dexamethasone 8 mg IV in a small trial with 25 treated adults (Table 1 A) and late vomiting with dexamethasone 0.5 mg/kg IV in a trial with a very low control event rate (the incidence of vomiting with placebo was only 9%) in children (Table 1 B). When data were combined to increase power, the number-needed-to treat to prevent early and late vomiting with any dose of dexamethasone compared with placebo in adults and children was 7.1 (95% CI 4.5 to 18) and 3.8 (95% CI 2.9 to 5). Two adult trials analyzed dexamethasone’s antinausea effect (Table 1 B); the number needed to treat was 4.3 (95% CI 2.3 to 26).
Dexamethasone Versus Other Antiemetics
In adults, dexamethasone 8 mg was compared with ondansetron 4 mg, granisetron 3 mg, or droperidol 0.02 mg/kg, in one trial each (Table 1, C and D, and Figure 1). In children, dexamethasone 150 μg/kg was compared with perphenazine 70 μg/kg in one trial (Table 1 C). Two results were statistically significant, both in favor of the comparator antiemetic. The combined data from the comparisons of dexamethasone with ondansetron and granisetron, respectively, suggested superiority of the 5-HT3 receptor antagonists for the prevention of early vomiting (Table 1 C); the number needed to treat was −5.9 (95% CI −3.5 to −20). In one large trial in children (Table 1 C), perphenazine 70 μg/kg was significantly more effective in preventing early vomiting than dexamethasone 150 μg/kg; the number needed to treat was −4.4 (95%CI −3.0 to −8.5).
Concomitant Use of Dexamethasone with Other Antiemetics
Dexamethasone alone was compared with the combination of dexamethasone with a 5-HT3 receptor antagonist (ondansetron or granisetron) in eight trials (Table 2, A and B, and Figure 2), with the combination of dexamethasone with droperidol in one trial (Table 2 e), and with the combination of dexamethasone with metoclopramide in one trial (Table 2 f). Only the concomitant use of dexamethasone with a 5-HT3 receptor antagonist showed a statistically significant improvement. The combined data from adults and children suggested a long-term benefit with dexamethasone 8 mg plus ondansetron 4 mg, or granisetron 40 μg/kg, or granisetron 3 mg, respectively, compared with the respective 5-HT3 receptor alone; the number needed to treat to prevent late nausea (adult data only) and vomiting (data from adults and children) was 7.8 (95% CI 4.1 to 66), and 7.7 (95% CI 4.8 to 19) (Table 2 B).
Dexamethasone Added to a 5-HT3 Receptor Antagonist Versus Placebo
Dexamethasone 8 mg added to granisetron 20 or 40 μg/kg was compared with placebo in two trials only, both in adults (Table 2, C and D, and Figure 2). Event rates with the combination therapy were very low, between 2% and 5% for both early and late outcomes. The number needed to treat point estimate to prevent early nausea and vomiting with the combination therapy compared with placebo was approximately 4 and to prevent late nausea and vomiting was 3.7 and 5.5.
Incidence of PONV with Placebo, Dexamethasone, 5-HT3 Receptor Antagonists, and Combination Therapy
To get qualitatively more insight into the potential usefulness of the combination of dexamethasone with a 5-HT3 receptor antagonist, we plotted graphically the incidence of nausea and vomiting with placebo, with dexamethasone, with any 5-HT3 receptor antagonist, and with the combination of dexamethasone with a 5-HT3 receptor antagonist from any trial that included one of these treatment arms (Figures 3). and 4 This was done with data from both direct comparisons (23–28,31,32,34,36) and independent trials (9,22,23,28–30,33,35,37); there was no attempt to analyze these data quantitatively.
With placebo, the average incidence of early nausea and vomiting was 33% (95% CI 22% to 44%), and 34% (95% CI 28% to 40%), respectively (Figure 3). With the combination therapy, the average incidence of early nausea was 3.9% (95% CI 1% to 7%) and of early vomiting was 1.4% (95% CI 0.2% to 3%). Average incidences of both dexamethasone and 5-HT3 receptor antagonist alone were between those of placebo and combination therapy.
For late outcomes, there was a similar distribution of PONV incidences with the different therapies (Figure 4). With placebo, the average incidence of late nausea was 45% (95% CI 36% to 54%) and of late vomiting was 48% (95% CI 42% to 54%). With the combination therapy, these incidences were 25% (95% CI 18% to 32%), and 17% (95% CI 13% to 21%).
Adverse effects were most frequently reported for the association of dexamethasone with a 5-HT3 receptor antagonist. Adverse effects were headache (23–28), dizziness (23–27), drowsiness and sedation (23–27), constipation (24–27), and muscle pain (24–27). There was no statistically significant difference between dexamethasone, 5-HT3 receptor antagonists, and placebo for these adverse effects. In one trial, pneumonia was reported in 1 of 41 children, and secondary hemorrhage in 1 of 41 children treated with dexamethasone (22).
In one trial in children undergoing adenotonsillectomy (22) and in one trial in adults undergoing molar extraction (9), surgical edema and inflammation was an end point. In both trials, edema and inflammation in the postoperative period was significantly less severe with dexamethasone compared with placebo, and delay until first oral intake was significantly shortened (9,22). Adults undergoing extraction of third molars reported significantly less intense pain with dexamethasone compared with placebo (9). Two patients (0.8% of all day case patients) needed unplanned hospital admission because of intractable PONV; both had received a placebo (35).
Although much attention has been paid to the prevention of PONV during the last decades, the optimal antiemetic regimen for both adults and children in the surgical setting has still not been established. The optimal antiemetic regimen would decrease the incidence of nausea and vomiting without increasing the risk of unacceptable adverse effects.
There is evidence from systematic review that interventions now used to prevent PONV either do not work very well (38–40) or are not antiemetic (41). Furthermore, all these interventions increase the risk of adverse reactions. Thus, the value of prophylaxis of PONV may be questioned. There are three main results from this systematic review on the antiemetic efficacy and safety of prophylactic dexamethasone in the surgical setting.
First, dexamethasone showed antiemetic efficacy compared with placebo. Efficacy in children and adults were similar. We therefore combined data to further increase power. Late efficacy of dexamethasone was most pronounced. Approximately four patients, adults or children, need to be treated with one prophylactic dose of dexamethasone for one not to vomit within 24 hours, who would have done so had they all received a placebo (Table 1 B). For antinausea efficacy, the message was less clear; there were two trials with only a limited number of adult patients reporting nausea data (23,28). The number needed to treat point estimate, however, suggested that dexamethasone’s late effect on nausea was similar to its late effect on vomiting. Reasons for the increased late efficacy are unclear. Dexamethasone has a biological half-life of 36 to 72 hours (42). Thus, the late antiemetic efficacy may be a result of favorable pharmacokinetics. In chemotherapy, there is some evidence that delayed emesis (i.e., beyond 24 hours) is better controlled with dexamethasone compared with classic antiemetics (43,44). A dose-response relationship for dexamethasone could not be established. This may mainly be because of the narrow ranges of doses used both in adults and in children. Thus, we do not know if smaller doses would still be effective, or if it was worthwhile to test larger doses. This should define the research agenda.
Second, there was evidence of an increased antiemetic effect when dexamethasone was added to a 5-HT3 receptor antagonist. With 5-HT3 receptor antagonists alone, the risk of PONV was decreased compared with placebo (Figures 3 and 4). This is in agreement with the quantitative analysis of systematically searched placebo-controlled ondansetron trials in the surgical setting (40). When dexamethasone was given concomitantly with a 5-HT3 receptor antagonist, the absolute risk of PONV was even lower (Figures 3 and 4). We do not know if the combination of dexamethasone with a 5-HT3 receptor antagonist leads to improved control of PONV symptoms via an additive or a synergistic effect. Only one trial attempted to address this issue (21). These authors found a better antiemetic efficacy with the combination of dexamethasone with a small dose of ondansetron compared with a three times higher dose of ondansetron alone. However, there is evidence that the small dose of ondansetron used in this trial (50 μg/kg) is as effective as the larger dose (150 μg/kg) (40,45). Thus, the trial of Splinter and Rhine (21) may support the hypothesis of an additive (but not a synergistic) antiemetic effect when dexamethasone and ondansetron are combined.
Third, adverse effects were rarely reported, and they were mainly related to 5-HT3 receptor antagonists (i.e., headache, constipation). The problem is that only a limited number of patients was tested in these trials (598 received dexamethasone alone and 343 received dexamethasone combined with another antiemetic), and patients were preselected. For instance, patients with a history of gastrointestinal disease, diabetes, corticosteroid therapy, immunosuppression, recent tuberculosis, Cushing’s syndrome, cataract, or hypertension were not included in these trials. Also, in none of the trials was functional capacity and potential inhibition of the hypothalamic-pituitary-adrenal axis tested. This should define the research agenda for adverse effects of dexamethasone in the surgical setting. In patients without risk factors who received dexamethasone 20 mg per day for five days for the control of chemotherapy-induced emesis, there was no evidence of immunosuppression or dysfunction of the hypothalamic-pituitary-adrenal axis (46). However, we still do not know if a single bolus dose of dexamethasone 8 or 10 mg (the most frequently used doses in these trials) is safe in patients at risk of corticosteroid-related adverse effects, nor do we know if a single dose of dexamethasone would suppress adrenal function in otherwise healthy patients during surgical stress, if suppression of adrenal function should happen in these patients, or if this was clinically relevant (for instance, if it increased the risk of wound dehiscence or infection).
A final issue relates to the definition of the most appropriate end points in PONV trials. In the surgical setting, a truly effective antiemetic regimen would be expected to increase a patient’s comfort, to shorten a stay in the recovery room (and thus enable early discharge after day case surgery), and to prevent unplanned hospital admission caused by intractable PONV. It has been claimed that nausea and vomiting are surrogate end points in this context (47). However, unless antiemetic regimens for the prevention of PONV are truly effective, we cannot expect them to have an impact on patient satisfaction or duration of hospital stay. The combination of dexamethasone with a 5-HT3 receptor antagonist is likely to be the most effective prophylactic antiemetic intervention currently available for the control of PONV. It seems to be worthwhile to test the impact of this combination therapy on so-called true end points in valid trials.
In conclusion, in the surgical setting, a single prophylactic dose of dexamethasone is antiemetic compared with placebo without evidence of any clinically relevant toxicity in otherwise healthy patients. Late (i.e., up to 24 hours) efficacy seems to be most pronounced. It is likely that the best prophylaxis of PONV currently available is achieved by combining dexamethasone with a 5-HT3 receptor antagonist. Optimal doses of this combination need to be identified.
We thank Daniel Haake from the Documentation Service of the Swiss Academy of Medical Sciences (DOKDI) for his help in searching electronic databases.
1. Kris M, Gralla R, Tyson L, et al. Controlling delayed vomiting: double-blind randomized trial comparing placebo, alone, and metoclopramide plus dexamethasone in patients receiving cisplatin. J Clin Oncol 1989; 7:108–14.
2. Rich W, Abdulhayoglu G, Di Saia PJ. Methylprednisolone as antiemetic during cancer chemotherapy: a pilot study. Gynecol Oncol 1980; 9:193–8.
3. Harris AL. Cytotoxic-therapy-induced vomiting is mediated via enkephalin pathways. Lancet 1982; 1:714–6.
4. Smith D, Newlands E, Rustin G, et al. Comparison of ondansetron and ondansetron plus dexamethasone as antiemetic prophylaxis during cisplatin-containing chemotherapy. Lancet 1991; 338:487–90.
5. Perez E. Use of dexamethasone with 5-HT3-receptor antagonists for chemotherapy-induced nausea and vomiting. Cancer J 1998; 2:72–7.
6. Young S. Mechanism of decline in rat brain 5-hydroxytryptamine after induction of liver tryptophan pyrrolase by hydrocortisone: roles of tryptophane catabolism and kynurenine synthesis. Br J Pharmacol 1981; 74:695.
7. Frederikson M, Hursti T, Fürst C, et al. Nausea in cancer chemotherapy is inversely related to urinary cortisol excretion. Br J Cancer 1992; 65:779–80.
8. Sagar S. The current role of anti-emetic drugs in oncology: a recent revolution in patient symptom control. Cancer Treat Rev 1991; 18:95–135.
9. Baxendale BR, Vater M, Lavery KM. Dexamethasone reduces pain and swelling following extraction of third molar teeth. Anaesthesia 1993; 48:961–4.
10. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996; 17:1–12.
11. Korttila K. The study of postoperative nausea and vomiting. Br J Anaesth 1992; 69:S20–3.
12. Rose J, Watcha M. Postoperative nausea and vomiting in paediatric patients. Br J Anaesth 1999; 83:104–17.
13. Morris J, Gardner M. Calculating confidence intervals for relative risk, odds ratios, and standardised ratios and rates. In: Gardner MJ, Altman DG, eds. Statistics with confidence: Confidence intervals and statistical guidelines. London: BMJ, 1995: 50–63.
14. Yusuf S, Peto R, Lewis J, et al. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovasc Dis 1985; 27:335–71.
15. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177–88.
16. Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. N Engl J Med 1988; 318:1728–33.
17. Cook RJ, Sackett DL. The number needed to treat: a clinically useful measure of treatment effect. BMJ 1995; 310:452–4.
18. Altman D. Confidence intervals for the number needed to treat. BMJ 1998; 317:1309–12.
19. McQuay HJ, Moore RA. Using numerical results from systematic reviews in clinical practice. Ann Intern Med 1997; 126:712–20.
20. Catlin FI, Grimes WJ. The effect of steroid therapy on recovery from tonsillectomy in children. Arch Otolaryngol Head Neck Surg 1991; 117:649–2.
21. Splinter WM, Rhine ED. Low-dose ondansetron with dexamethasone more effectively decreases vomiting after strabismus surgery in children than does high-dose ondansetron. Anesthesiology 1998; 88:72–5.
22. April MM, Callan ND, Nowak DM, Hausdorff MA. The effect of intravenous dexamethasone in pediatric adenotonsillectomy. Arch Otolaryngol Head Neck Surg 1996; 122:117–20.
23. Fujii Y, Tanaka H, Toyooka H. Granisetron-dexamethasone combination reduces postoperative nausea and vomiting. Can J Anaesth 1995; 42:387–90.
24. Fujii Y, Tanaka H, Toyooka H. Granisetron and dexamethasone provide more improved prevention of postoperative emesis than granisetron alone in children. Can J Anaesth 1996; 43:1229–32.
25. Fujii Y, Tanaka H, Toyooka H. The effects of dexamethasone on antiemetics in female patients undergoing gynecologic surgery. Anesth Analg 1997; 85:913–7.
26. Fujii Y, Tanaka H, Toyooka H. Prophylactic antiemetic therapy with granisetron-dexamethasone combination in women undergoing breast surgery. Acta Anaesthesiol Scand 1998; 42:1038–42.
27. Fujii Y, Toyooka H, Tanaka H. Prophylactic antiemetic therapy with a combination of granisetron and dexamethasone in patients undergoing middle ear surgery. Br J Anaesth 1998; 81:754–6.
28. Lopez-Olaondo L, Carrascosa F, Pueyo FJ, et al. Combination of ondansetron and dexamethasone in the prophylaxis of postoperative nausea and vomiting. Br J Anaesth 1996; 76:835–40.
29. Liu K, Hsu CC, Chia YY. Effect of dexamethasone on postoperative emesis and pain. Br J Anaesth 1998; 80:85–6.
30. Ohlms LA, Wilder RT, Weston B. Use of intraoperative corticosteroids in pediatric tonsillectomy. Arch Otolaryngol Head Neck Surg 1995; 121:737–42.
31. McKenzie R, Tantisira B, Karambelkar DJ, et al. Comparison of ondansetron with ondansetron plus dexamethasone in the prevention of postoperative nausea and vomiting. Anesth Analg 1994; 79:961–4.
32. McKenzie R, Riley TJ, Tantisira B, Hamilton DL. Effect of propofol for induction and ondansetron with or without dexamethasone for the prevention of nausea and vomiting after major gynecologic surgery. J Clin Anesth 1997; 9:15–20.
33. Pappas ALS, Sukhani R, Hotaling AJ, et al. The effect of preoperative dexamethasone on the immediate and delayed postoperative morbidity in children undergoing adenotonsillectomy. Anesth Analg 1998; 87:57–61.
34. Rothenberg DM, McCarthy RJ, Peng CC, Normoyle DA. Nausea and vomiting after dexamethasone versus droperidol following outpatient laparoscopy with a propofol-based general anesthetic. Acta Anaesthesiol Scand 1998; 42:637–42.
35. Splinter WM, Roberts DJ. Dexamethasone decreases vomiting by children after tonsillectomy. Anesth Analg 1996; 83:913–6.
36. Splinter W, Roberts DJ. Prophylaxis for vomiting by children after tonsillectomy: dexamethasone versus perphenazine. Anesth Analg 1997; 85:534–7.
37. Tom L, Templeton J, Thompson M, Marsh RR. Dexamethasone in adenotonsillectomy. Int J Pediatr Otorhinolaryngol 1996; 37:115–20.
38. Tramèr M, Moore A, McQuay H. Omitting nitrous oxide in general anaesthesia: meta-analysis of intraoperative awareness and postoperative emesis in randomized controlled trials. Br J Anaesth 1996; 76:186–93.
39. Tramèr M, Moore A, McQuay H. Propofol anaesthesia and postoperative nausea and vomiting: quantitative systematic review of randomized controlled studies. Br J Anaesth 1997; 78:247–55.
40. Tramèr MR, Reynolds DJM, Moore RA, McQuay HJ. Efficacy, dose-response, and safety of ondansetron in prevention of postoperative nausea and vomiting: a quantitative systematic review of randomized placebo-controlled trials. Anesthesiology 1997; 87:1277–89.
41. Henzi I, Walder B, Tramèr MR. Metoclopramide in the prevention of postoperative nausea and vomiting: a quantitative systematic review of randomised placebo-controlled trials. Br J Anaesth. In press.
42. Haynes R. Adrenocorticotropic hormone: adrenocortical steroids and their synthetic analogs–inhibitors of the synthesis and actions of adrenocortical hormones. In: Goodman Gilman A, Gilman LS, Rall TW, Murad F, eds. The pharmacological basis of therapeutics. 8th ed. New York: Pergamon Press, 1990: 1447–8.
43. Tavorath R, Hesketh PJ. Drug treatment of chemotherapy-induced delayed emesis. Drugs 1996; 52:639–48.
44. Jones AL, Hill AS, Soukop M, et al. Comparison of dexamethasone and ondansetron in the prophylaxis of emesis induced by moderately emetogenic chemotherapy. Lancet 1991; 338:483–7.
45. Watcha MF, Bras PJ, Cieslak GD, Pennant JH. The dose-response relationship of ondansetron in preventing postoperative emesis in pediatric patients undergoing ambulatory surgery. Anesthesiology 1995; 82:47–52.
46. Del Priore G, Gurski K, Warshal D. Adrenal function following high-dose steroids in ovarian cancer patients. Gynecol Oncol 1995; 59:102–4.
47. Fisher D. Surrogate outcomes: meaningful not! Anesthesiology 1999; 90:355–6.