Cross-checking to reduce adverse events resulting from medical errors in the emergency department: study protocol of the CHARMED cluster randomized study

BACKGROUND: Medical errors and preventable adverse events are a major cause of concern, especially in the emergency department (ED) where its prevalence has been reported to be roughly of 5–10 % of visits. Due to a short length of stay, emergency patients are often managed by a sole physician – in contrast with other specialties where they can benefit from multiples handover, ward rounds and staff meetings. As some studies report that the rate and severity of errors may decrease when there is more than one physician involved in the management in different settings, we sought to assess the impact of regular systematic cross-checkings between physicians in the ED. DESIGN: The CHARMED (Cross-checking to reduce adverse events resulting from medical errors in the emergency department) study is a multicenter cluster randomized study that aim to evaluate the reduction of the rate of severe medical errors with implementation of systematic cross checkings between emergency physician, compared to a control period with usual care. This study will evaluate the effect of this intervention on the rate of severe medical errors (i.e. preventable adverse events or near miss) using a previously described two-level chart abstraction. We made the hypothesis that implementing frequent and systematic cross checking will reduce the rate of severe medical errors from 10 to 6 % - 1584 patients will be included, 140 for each period in each center. DISCUSSION: The CHARMED study will be the largest study that analyse unselected ED charts for medical errors. This could provide evidence that frequent systematic cross-checking will reduce the incidence of severe medical errors. TRIAL REGISTRATION: Clinical Trials, NCT02356926 ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12873-015-0046-1) contains supplementary material, which is available to authorized users.