ESPE Yearbook of Paediatric Endocrinology

BMJ 2025; 389:e082068. PMID: 40268307 https://www.bmj.com/content/389/bmj-2024-082068.long

In Brief: This literature-based analysis performed forward citation searching of randomised controlled trials in humans that were retracted for any reason. They identified 1330 retracted trials and 847 systematic reviews that included any retracted trials in a total of 3902 meta-analyses. The authors replicated those meta-analyses but excluding any retracted trials; this changed the direction of the pooled effect in 8.4% (95% CI 6.8% to 10.1%) of meta-analyses, and the statistical significance in 16.0% (14.2% to 17.9%).

Comment: Unfortunately there are increasing numbers of research papers, including randomised controlled trials, that are retracted after publication due to questionable issues in their processes or data quality. Fortunately, you can search for retracted papers in the online Retraction Watch database (https://retractiondatabase.org/), which was established with major philanthropic funding. However, retractions often occur after the published data are included in systematic reviews and subsequently contribute to the evidence base that supports clinical guidance and public health policies. Most of these are not re-evaluated until we are aware that new evidence is available.

The authors caution that incorporation of “contaminated” evidence into clinical practice guidelines may lead to incorrect conclusions and potentially harm patients. A worrying finding was of 157 clinical guidelines that relied on evidence from 69 systematic reviews that were significantly influenced by meta-analyses containing retracted data. Each retracted trial could contaminate, on average up to 13 meta-analyses from 3 systematic reviews, and each systematic review would further contaminateat least 3 clinical guidelines. The authors present some examples of affected guidelines on management of unexplained infertility, postoperative nausea and vomiting and the use of ACE inhibitors.

Addressing this issue will require the attention and active input of health professionals, editors, publishers, organisations and scientific societies to monitor and assess the impact of retracted papers on their evidence and clinical guidelines.