When and how should we cluster and cross over: methodological and ethical issues

Excerpt

Spence et al. describe the need for more randomized controlled trials to provide rigorous evidence of the real-world effectiveness of treatments and treatment strategies in anesthesia practice. This call is supported by evidence citing a large degree of individual practice variability among cardiac anesthesiologists regarding benzodiazepine use within the standard of care.1 They describe the cluster randomized crossover design as ideal for this purpose. In the proposed approach, they “seek to evaluate the impact of two different approaches to cardiac anesthesia, one where nearly all patients receive intraoperative benzodiazepines unless there are contraindications (routine benzodiazepine arm), and the other where nearly all patients receive no intraoperative benzodiazepines unless there are contraindications (benzodiazepine restricted arm)…during 12, four-week crossover periods”.1 The interventions are implemented as policies such that all eligible patients in the hospital during each period receive the allocated interventions by default without any patient recruitment or consent. While we do not dispute the need for rigorous evidence to inform policy and practice, we have concerns regarding the methodological and ethical issues raised in trials such as the B-Free trial. …