A factorial study of ondansetron, metoclopramide, and dexamethasone for emesis prophylaxis after adenotonsillectomy in children

Background: We conducted a factorial study of emesis prophylaxis with ondansetron (OND), metoclopramide (MET), and dexamethasone (DEX). Methods: After informed parental consent, 240 children having adenotonsillectomy were randomized to one of 15 combinations of OND (0-60 μg·kg^-1), MET (0-400 μg·kg^-1), and/or DEX (0-500 μg·kg^-1). Using multivariable logistic regression, models were generated for the probability of emesis before discharge, after discharge and overall for 24 h. Results: Odds of emesis increased by a factor of three to four for children older than 7 years. Before discharge, odds of emesis decreased by factors of 0.29 for each 15 μg·kg^-1 of OND and 0.37 for each 100 μg·kg^-1 of MET. After discharge, odds of emesis decreased by a factor of 0.67 for each 125 μg·kg^-1 of DEX and increased by a factor of 3.5 for emesis before discharge. Over 24 h, odds of emesis decreased with OND, MET, and DEX (ORs as above). A negative interaction between OND and MET was seen before discharge and over 24 h, reducing the efficacy of their combination. Conclusions: We present novel study design and methods of analysis which are uniquely suited to studies of multiple interventions. Factorial design was a powerful tool, allowing simultaneous determination of dose-response relationships for three drugs and identifying a previously unreported negative interaction between OND and MET.