Causal inference in multi-cohort studies using the target trial approach

Longitudinal cohort studies have the potential to examine causal effects of complex health exposures on longer-term outcomes. Utilizing data from multiple cohorts has the potential to add further benefit, by improving precision of estimates and allowing examination of effect heterogeneity and replicability. However, the interpretation of findings can be complicated by unavoidable biases that may be compounded when pooling data from multiple cohorts, and/or may contribute to discrepant findings across cohorts. Here we extend the 'target trial' framework, already well established as a powerful tool for causal inference in single-cohort studies, to address the specific challenges that can arise in the multi-cohort setting. Using a case study, we demonstrate how this approach enables clear definition of the target estimand and systematic consideration of sources of bias with respect to the target trial as the reference point, as opposed to comparing one study to another. This allows identification of potential biases within each cohort so that analyses can be designed to reduce these and examination of differential sources of bias to inform interpretation of findings. The target trial framework has potential to strengthen causal inference in multi-cohort studies through improved analysis design and clarity in the interpretation of findings.