Nine RCTs were identified (n=1,232, range 60 to 240). Mean validity score was 5.4 (range 3 to 7). Two hundred patients received other antiemetics and were excluded from the analysis. The authors give the total number of patients as 1,152.
Incidence of pruritus was not significantly reduced with 5-HT3 receptor antagonists compared to placebo (RR 0.94, 95% CI 0.81 to 1.09, I2=65%; five RCTs), but their use significantly reduced incidence of severe pruritus (RR 0.79, 95% CI 0.65 to 0.97, NNT=13; four RCTs) and need for treatment of pruritus (RR 0.80, 95% CI 0.64 to 0.96, NNT=15; six RCTs). One RCT assessed treatment success for established pruritis and reported that ondansetron significantly reduced the pruritis severity score (RR 0.30, 95% CI 0.16 to 0.59, NNT=3).
Use of 5-HT3 receptor antagonists significantly reduced incidence of postoperative nausea (RR 0.75, 95% CI 0.58 to 0.96, NNT=9, I2=23%; four RCTs), vomiting (RR 0.49, 95% CI 0.30 to 0.81, NNT=12, I2=33%; four RCTs), need for postoperative rescue antiemetic treatment (RR 0.38, 95% CI 0.21 to 0.68, NNT=7; three RCTs) and incidence of severe postoperative nausea and vomiting (RR 0.55, 95% CI 0.33 to 0.76, NNT=9; four RCTs) compared to placebo.
Subgroup analyses indicated that only treatment with ondansetron resulted in statistically significant results (as reported in the review). There was no evidence for dose responsiveness for ondansetron for prophylaxis of pruritis or severity of pruritis. There was no significant effect for 5-HT3 receptor antagonists compared to placebo on incidence of postoperative pruritus when the analysis was limited to trials that used intrathecal morphine without lipophilic drugs.
No significant differences between 5-HT3 receptor antagonists and placebo were reported for side effects; these included headache, cardiac arrhythmias and extrapyramidal side effects.