TY - JOUR T1 - 2 Time for a real manifesto JF - BMJ Evidence-Based Medicine JO - BMJ EBM SP - A1 LP - A2 DO - 10.1136/bmjebm-2019-EBMLive.2 VL - 24 IS - Suppl 1 AU - Stephen Bradley Y1 - 2019/07/01 UR - http://ebm.bmj.com/content/24/Suppl_1/A1.2.abstract N2 - Objectives The scandal of poor medical research is rooted in failings of individuals and institutions. Improving the situation requires us to formulate political solutions and lobby hard for their implementation.The EBM manifesto is a worthy statement of principles, but it is nebulous and lacks specific objectives. The dire scale of the crisis it eloquently describes threatens to overwhelm the impetus to action with a counsel of despair. The problems are well understood, now we must identify and ruthlessly promote a limited number of achievable and high impact solutions.Method Five Demands1. Fix the Research Evaluation Framework REF rewards academics who demonstrate ‘impact’, commonly interpreted as trials with positive outcomes that alter practice. REF should be re-engineered to incentivize rigorous and transparent research that matters to patients. Institutional returns should include metrics of patient participation and research integrity, including research registration, data sharing and reporting. Returns for papers with unwarranted deviation from protocols will be prohibited.2. Overhaul Research Funding Contractual compliance measures are required for government health funding, with payments made contingent on fulfilment of agreed methodologies and reporting of outcomes. In cases of breaches of such terms, provision must be made for funding ‘clawback’.Results 3. Establish a National Health Research Data Repository All research data should be submitted on a national repository, including ‘lab books’ documenting experimental data. Depositing additional documentary evidence such as videos of lab procedures will become an expectation of good practice. Submission of publically viewable, blank research databases and examples of pseudonymised patient data that will be made available for peer (and/or) public review will be a mandatory condition of grant applications. Data which will not be publically accessible must be identified and justified along with the terms under which the data will be made available for peer scrutiny.4. Standardise & Integrate Journal Submissions Journals should adopt a uniform format in which papers will be accepted for peer review, whilst the submission platforms should be integrated to eliminate duplication. Achieving this will engage grassroots researchers and could offset additional workload resulting from enhanced reporting and transparency in research.5. Establish a Health Research Ombudsman Experience suggests that even with rules, compliance is poor. An ombudsman is required to police fulfilment of obligations including data sharing, trial reporting and the measures outlined above. The ombudsman will respond to concerns from patients and professionals and will have oversight of public funding of health research. The ombudsman will also enlist peer reviewers to conduct audits of ongoing research, akin to the role of Specialist Inspectors in the CQC.Conclusions The profound problems in health research are widely acknowledged. The underlying causes are dysfunctional incentives, opacity and misalignment of research with patient priorities. There is no prospect that these issues will be resolved if we cannot translate the principles in the EBM manifesto into a set of practical measures. The five actions highlighted are not a panacea but they are practical first steps in overturning the scandal of poor medical research. ER -