Prevention or treatment of postoperative vomiting using ondansetron? A mathematical assessment

To assess the effectiveness of ondansetron versus placebo when used as prophylaxis of postoperative vomiting compared to its use for the treatment of established postoperative vomiting.

All prospectively randomized controlled trials, identified using the MEDLINE database and which had been included in two previously published meta-analyses.

The effectiveness of ondansetron (1, 4, and 8 mg) used for prevention (28 studies) or treatment (two studies) of postoperative vomiting was compared using odds ratios and number needed to treat (NNT) scores. NNT values were extrapolated to a similar incidence of vomiting in the placebo groups. The effectiveness equivalence between both meta-analyses was compared by assessing the degree of coincidence between the respective 95% confidence intervals (CI). The 1-mg dose was superior in treatment than in prophylaxis, odds ratios (95% CI): 2.73 (1.81-4.11) versus 1.34 (1.04-1.74), respectively (p < 0.05), although this difference did not reach statistical significance when the NNT values were compared. For the 4-mg dose, a near congruence was observed in the odds ratios; 3.34 (2.22-5.03) versus 2.61 (2.24-3.05) as well as in the NNT values: 3.90 (2.97-5.78) versus 3.40 (2.93-4.07). For the 8-mg dose, there were equivalent NNT values: 4.10 (3.08-6.23) versus 4.08 (3.28-5.43), and with respect to the odds ratios: 3.18 (2.11-4.81) versus 2.42 (1.95-2.99). The differences were not statistically significant.

For the usual doses recommended for postoperative emesis, there was equivalent effectiveness of ondansetron whether administered as prophylaxis or as a treatment of established vomiting.