Droperidol compared to 5-HT3-antagonists for prophylaxis of postoperative nausea and vomiting - A meta-analysis of controlled randomised studies

Objective: Randomised controlled trials comparing prophylactic droperidol w ith 5-HT3-receptor antagonists (dolasetron, granisetron, ondansetron, or tr opisetron) to prevent postoperative nausea and vomiting (PONV) were include d in a meta-analysis to estimate the relative efficacy of both treatments. Materials and methods: Studies were systematically searched using Medline, EMBASE, the Cochrane-Library, and the manufacturers' databases. The search was completed by manual screening of the reference lists and current issues of locally available anaesthesia journals. The main end point was defined as complete absence of nausea, retching, and vomiting within 6 hours ("earl y PONV") and within 48 hours ("late PONV") after treatment with either drop eridol or a 5-HT3-receptor antagonist. The pooled odds-ratios (OR) and the numbers-needed-to-treat (NNT) with their corresponding 95%-confidence inter vals (given in parentheses) were calculated using a random effects model. Results: A total of 62 studies with 7075 patients (3743 receiving droperido l, 3332 receiving a 5-HT3-antagonist) were analysed. If all 5-HT3-antagonis ts are grouped together, these drugs were significantly more effective than droperidol to prevent both "early" and "late" PONV (OR: 1.51 (1.19-1.93) a nd 1.56 (1.28-1.90) respectively). The corresponding NNTs (numbers of patie nts who must be treated with a 5-HT3-antagonist to prevent one patient from PONV that otherwise would have suffered from PONV though treatment oder by being treated with droperidol) were 15 (9.5-36) for "early PONV" and 11.1 (7.7-18.9) for "late PONV" These results are based mainly on studies using ondansetron and granisetron that comprise more than 90% of all available studies. Dolasetron (2 studie s) and Tropisetron (3 studies) are not yet sufficiently proven to be superi or to droperidol. Comparing the four different 5-HT3-antagonist, granisetron seems to be the most effective drug ("early PONV" OR: 3.15 (2.19-4.54); "late PONV" OR: 2.9 6 (2.18-4.03)). However, this result must be interpreted with caution, as 1 9 out of the 20 available studies were published by the same group. The res ults of the present analysis lead to the speculation that there might be a systematical error biasing the data concerning granisetron. Conclusion: The 5-HT3-antagonists are significantly more effective than dro peridol. However, this difference is not of high clinical relevance, consid ering the high NNT. Thus, droperidol still has an important place in the pr ophylaxis of postoperative nausea and vomiting.