An international multicenter protocol to assess the single and combined benefits of antiemetic interventions in a controlled clinical trial of a 2×2×2×2×2×2 factorial design (IMPACT)

Abstract

For various diseases clinicians have to combine different drugs or interventions when a single drug or intervention does not lead to satisfactory results. However, quantifying the relative benefit of certain drugs or interventions when given alone and in combination under controlled conditions requires a complex factorial design. This paper describes such a method applied to a large multicenter trial for the prevention of postoperative nausea and vomiting (PONV), which may be of great interest for other specialties. Approximately 28 million operations are performed annually in the United States, mainly under general anesthesia with volatile anesthetics. Unfortunately, one-third of these patients suffer from PONV. This prompted extensive research of antiemetic and anesthetic drugs, but none of the interventions appeared to satisfactorily prevent PONV. Scuderi et al. were the first to almost eliminate PONV by combining various antiemetic interventions; however, the relative benefit of each intervention remained unclear. Accordingly, we have designed a large randomized controlled trial studying six different antiemetic interventions—three involving use of various antiemetic drugs and three involving choice of anesthetic drugs—to answer the following main questions: (1) What is the relative benefit of each of the antiemetic intervention? (2) Are certain combinations of antiemetic interventions more effective than others? Using a complete factorial design this leads to 2³ = 8 antiemetic combinations, which multiply with the 2³ = 8 combinations of anesthetic drugs, leading to a total of 2⁶ = 64 possible combinations. The six factors are the antiemetics ondansetron (versus control), dexamethasone (versus control), droperidol (versus control), and the intravenous anesthetic propofol (versus volatile anesthetics), air (versus nitrous oxide), and remifentanil (versus fentanyl). The primary outcome is freedom from PONV within the first 24 hours after anesthesia. Eligible patients are adults scheduled for elective surgery under general anesthesia with an increased risk for PONV so that the expected incidence in the control group (with none of the six antiemetic interventions) is approximately 60%. In order to allow analyses for up to three factor interactions, a sample size was estimated to be in the range of approximately 5000 patients. To the best of our knowledge this is the first randomized controlled trial of a six-way factorial design that may serve as an example for numerous other medical specialties.