Meta-analysis of the safety of 5-HT3 antagonists with dexamethasone or droperidol for prevention of PONV.

BACKGROUND

Antiemetic guidelines recommend a combination of serotonin (5-HT3) with a second agent such as droperidol or dexamethasone. Physicians have been reluctant to employ these guidelines due to concerns over the black-box warning of droperidol and safety concerns with a steroid.

OBJECTIVE

To assess the safety profiles of 5-HT3 receptor antagonist (5-HT3RA) monotherapy and combination therapy with a steroid or droperidol for prophylaxis of postoperative nausea and vomiting (PONV).

METHODS

A MEDLINE search of English-language reports of randomized controlled trials (RCTs) was conducted (1966-September 2005) using the key terms 5-HT3, granisetron, ondansetron, dolasetron, tropisetron, PONV, postoperative, vomiting, emesis, and nausea. RCTs with treatment arms comparing 5-HT3RA monotherapy (granisetron, ondansetron, dolasetron, or tropisetron) with dexamethasone or droperidol or 5-HT3RA combinations and providing incidence data on adverse events were identified and reviewed. Within-study odds ratios with 95% confidence intervals were calculated to determine the incidence rates of all adverse events in RCTs using 5-HT3RA monotherapy and combination therapies. Overall effect sizes for frequently reported adverse events were estimated by pooling ORs using fixed- and random-effect models.

RESULTS

Pooled ORs (OR(pooled)) for adverse events with 5-HT3RA/dexamethasone versus 5-HT3RA for PONV prophylaxis were not significant for any reported adverse events or the overall incidence of adverse events; 5-HT3RA/droperidol versus 5-HT3RA was significant only for decreased headache incidence (fixed model: OR(pooled) 0.35; 95% CI 0.18 to 0.69). The OR(pooled) for 5-HT3RA/dexamethasone versus dexamethasone was not significant for any reported adverse events except headaches (fixed model OR(pooled) 1.75; 95% CI 1.01 to 3.03), none of which was serious. OR(pooled) for 5-HT3RA/droperidol versus droperidol was not significant for any reported adverse events. Avascular necrosis, occult infection, and delayed wound healing were not observed with either combination therapy. Cardiac abnormalities were observed with 5-HT3RA/droperidol therapy.

CONCLUSIONS

This meta-analysis indicates that either therapy has a safety profile similar to that of dexamethasone, droperidol, or 5-HT3RA.