Editors and reviewers often cannot spot fake or fatally flawed clinical trials using only summary tables. Audits that required anonymized individual participant data (IPD) found roughly a quarter of trials were untrustworthy, versus ~1% detected from summaries. Making IPD submission and audit a precondition for publishing randomized trials would expose errors and fraud before they enter the literature.
— This would change journal standards and strengthen the evidence base behind clinical guidelines, reimbursement, and public health policy.
2023.07.18
100% relevant
John Carlisle’s 2017–2020 audit of 500+ RCT submissions to Anaesthesia: 26% 'zombie' trials detected when IPD were available; ~1% detected without IPD.
← Back to All Ideas