Improving Clinical Trials Reporting

ongoing

This proposal will rigorously evaluate three approaches to reducing research waste and publication bias in clinical trials. The findings will provide a firm evidence base to inform efforts to improve the robustness of scientific evidence across the natural and social sciences. Non-publication of results has been estimated to directly waste $85 billion per year in medical research funding. Gaps left by unreported clinical trial results systematically distort the scientific evidence base because negative findings are especially likely to remain unreported. Health systems and insurers are often unable to determine how safe, effective and cost-effective treatments really are, leading to costly procurement missteps and suboptimal and sometimes dangerous patient care. For example, governments worldwide spent over $18 billion on the drug Tamiflu before the previously hidden results of several unsuccessful clinical trials emerged. Today, research waste safeguards for clinical trials are stronger and more advanced than in any other area of research. Multiple approaches to tackling the twin problems of research waste and publication bias have been pioneered in the clinical trials space and are now being adopted in other branches of science. However, it remains unclear which of these measures are effective because their impact has rarely been studied in a systematic and rigorous manner. This undermines efforts to improve the robustness of research across the natural and social sciences. This proposal will fill an important knowledge gap by providing robust evidence on which approaches to reducing research waste and curbing publication bias are effective. Breaking new methodological ground, it will answer three questions: Can stronger funder safeguards reduce research waste? Can sending reminder messages to research institutions reduce research waste? What reminder messages to principal investigators reduce research waste most?