For prospective meta-analyses (PMA), eligible studies are identified and PMA hypotheses, selection criteria and analysis methods are pre-specified before results of any of the studies are known. This reduces publication and selection bias and provides a unique opportunity for outcome standardisation. We conducted a world first PMA of four randomised controlled trials (EPOCH) investigating interventions to prevent childhood obesity.
To quantitatively analyse the effects of prospective planning on variations across trials, outcome harmonisation and the power to detect an intervention effect, and to derive recommendations for future PMA.
We analysed registration records, variable lists and key characteristics from the four trials included in the EPOCH PMA. We compared the outcomes trialists planned to collect prior to inclusion in the PMA to the adjusted outcomes trialists collected after inclusion. We analysed the proportion of matching outcome definitions across trials, numbers of included outcomes and how collaborating increased the power to detect an effect.
The included trials varied in intervention design and participants, which improved external validity and the ability to perform subgroup analyses. Prospective planning led to greater outcome harmonisation with 18% of outcome categories included in all trials prior to PMA-inclusion versus 91% after PMA-inclusion, and a 54% increase in number of collected outcome categories. However, trials often used different outcome definitions (e.g. measuring physical activity by hours of outside play versus using an activity monitor), so only 7% of outcomes were defined identically across all trials. While individual trials had limited power to detect the observed intervention effect, pooling the data substantially increased power.
Prospective planning led to greater outcome harmonisation and greater power to detect intervention effects, while maintaining variation in trial designs and studied populations, which heightens external validity. Recommendations for future PMA include more detailed harmonisation of outcome variables, and carefully defined pre-specification of analyses to avoid unnecessary over-collection of data.
Patient and healthcare consumer involvement:
None. Relevance: PMA can minimise biases to help improve health guidelines.