Systematic review protocol registration and reporting guidelines: development, implementation, assessment and utility

It is essential that systematic reviews are methodologically robust to minimise bias and well reported so users can have confidence in the findings. Limitations in the quality of published systematic reviews prompted my research into methodological aspects with the aim of improving the robustness of systematic review evidence.

I undertook an international Delphi consultation, with 200 experts agreeing a minimum data set for systematic review protocol registration. PROSPERO was designed and implemented based on the 22 required and 18 optional items identified as key protocol registration elements. An evaluation of the utility of the register at one year showed registration was feasible, with growing international engagement and positive feedback from the survey of users. I was subsequently involved in the consensus development of reporting guidelines for systematic review protocols leading to publication of the 17-item PRISMA-P checklist.

PROSPERO became a resource for methodological research, and I undertook an examination of outcome reporting bias, previously only possible in Cochrane reviews. In comparing the details in 96 published reviews with their PROSPERO records, 32% had discrepancies in their primary outcome and 39% did not specify a primary outcome. Having a favourable result or positive conclusion did not increase the risk of a discrepancy in outcome reporting.

Registration records can be the only publicly available source of planned methods, leading to my methodological study comparing registration data with protocol reporting guideline requirements. In a random sample of 439 PROSPERO records for reviews of health interventions, 53% (14,469/227,279) of the elements compared were classified as reported. This indicates that PROSPERO records are not a substitute for public access to a full protocol.

The research landscape has changed rapidly over the last decade and there is a need to revisit and clarify or re-purpose the roles of the systematic review protocol and protocol registration.