Peter McCulloch, Fei Shan, Nicole Bilbro & Allison Hirst
A seminar reporting a rapid systematic review and a sampling review and analysis of current literature with a discussion of how IDEAL should reduce research waste and an attempt to evaluate it’s potential impact.
- High level comparative literature overview of research quality, quantity and sources of bias in current research on surgery, therapeutic devices and pharma
- Types of Research Waste (REWARD classification) and typical faults of research on complex interventions
- Asking the wrong question: focusing on outcomes before demonstrating stability
- Poor design: Failure to achieve consensus over intervention definition, delivery fidelity, indications; failure to conduct qualitative research on attitudes and values
- Failure to publish: Failed trials due to poor preparatory studies
- Poor reporting: Failure to use standard definitions, report modifications to intervention or indication, study learning curves or report delivery fidelity
- Mechanisms by which IDEAL should reduce research waste
- Increasing transparency and clarity (d)
- Reducing repetition due to misunderstanding or ignorance (b,d)
- Increasing speed of recognition of intervention stability (a,b)
- Avoidance of waste due to invalid studies arising from quality and learning effects (b,c)
- Increased speed and probability of consensus around an RCT (b,c)
- Reduced waste from RCT failure due to lack of consensus or equipoise (b,c)
- Theoretical quantification of overall benefit from using IDEAL – sampling study of current literature