Emulating an RCT in observational research in chest trauma: The target trial approach.

Abstract

While the concept of a "target trial"-optimizing the quality of observational studies by attempting to emulate the ideal world conditions of a randomized controlled trial-was first expounded over a decade ago, the take up of this concept in the design and analysis of trials in trauma is lacking. The target trial approach avoids common errors in observational research to increase its scientific validity as well as potentially enable causal questions to be answered without the expense and intricacies of a randomized controlled trial. This review article briefly introduces the reader to the concepts and utility of a "target trial" approach before providing demonstrations of its application in the subject area of chest trauma. Four articles published in the last 5 years-two case control and two cohort studies-are chosen and considered in terms of their causal question; study population; inclusion and exclusion criteria; designation of time 0; clarity of the follow-up period; study outcomes; methods to minimize confounding; results; overall issues regarding study time; and the presence of avoidable errors such as introduction of immortal time bias or information bias. Two of the studies had an unclear causal question; none of the studies designated a time 0; the follow-up period was unclear for all but one of the studies; and one study had a serious issue with information bias resulting from differential misclassification. Failure to emulate a "target trial" framework may lead to serious methodologic issues in observational research. Expansion of the awareness of this approach in trauma literature will improve the quality of our observational research and potentially translate into significant benefits for our patients.