The use of interim data and Data Monitoring Committee recommendations in randomized controlled trial reports: frequency, implications and potential sources of bias

Research output: Contribution to journalArticle

Authors

Abstract

Background: Interim analysis of accumulating trial data is important to protect participant safety during randomized controlled trials (RCTs). Data Monitoring Committees (DMCs) often undertake such analyses, but their widening role may lead to extended use of interim analysis or recommendations that could potentially bias trial results. Methods: Systematic search of eight major publications: Annals of Internal Medicine, BMJ, Circulation, CID, JAMA, JCO, Lancet and NEJM, including all randomised controlled trials ( RCTs) between June 2000 and May 2005 to identify RCTs that reported use of interim analysis, with or without DMC involvement. Recommendations made by the DMC or based on interim analysis were identified and potential sources of bias assessed. Independent double data extraction was performed on all included trials. Results: We identified 1772 RCTs, of which 470 (27%; 470/1772) reported the use of a DMC and a further 116 (7%; 116/1772) trials reported some form of interim analysis without explicit mention of a DMC. There were 28 trials ( 24 with a formal DMC), randomizing a total of 79396 participants, identified as recommending changes to the trial that may have lead to biased results. In most of these, some form of sample size re-estimation was recommended with four trials also reporting changes to trial endpoints. The review relied on information reported in the primary publications and methods papers relating to the trials, higher rates of use may have occurred but not been reported. Conclusion: The reported use of interim analysis and DMCs in clinical trials has been increasing in recent years. It is reassuring that in most cases recommendations were made in the interest of participant safety. However, in practice, recommendations that may lead to potentially biased trial results are being made.

Details

Original languageEnglish
Article number12
JournalBMC Medical Research Methodology
Volume8
Issue number1
Publication statusPublished - 20 Mar 2008