Quality AssuranceOngoing
Predicting clinical trial replicability using elements of trial design, conduct and reporting
Background
Clinical research needs to produce trustworthy and reliable results to translate into improved patient care. The robustness of findings from clinical research is usually assessed over time by systematic reviews of randomized controlled trials. This process, however, suffers from a considerable time lag and the accuracy of conclusions may be adversely affected by factors such as the risk of bias of included trials, selective reporting, publication bias or questionable research practices. To date, it is not known whether clinical trial reports contain information that allows predicting how likely it is that the findings are robust and will replicate over time.
Objectives
We aim to identify elements of clinical trial design, conduct and reporting that predict the likelihood that findings from a clinical trial will be replicable.
Methodology
We will assess the replicability of clinical trials by comparing the findings of original clinical trials to pooled effect estimates from meta-analyses. Study characteristics that might influence replicability will be extracted from the original clinical trial publications. We will then determine whether any of the study characteristics examined predict the likelihood of clinical trial replicability.
Expected Results / Implications
Knowing which elements of clinical trial design, conduct and reporting are associated with the likelihood of replication for clinical trials may help practitioners to interpret new clinical trial results and may guide the development of reporting guidelines and screening tools.
