By the same authors

From the same journal

From the same journal

Potential for technical errors and subverted allocation can be reduced if certain guidelines are followed: Examples from a web-based survey

Research output: Contribution to journalArticle

Published copy (DOI)



Publication details

JournalJournal of Clinical Epidemiology
DatePublished - Mar 2009
Issue number3
Number of pages9
Pages (from-to)261-269
Original languageEnglish


Objective: To elicit researchers' experiences and knowledge of how the randomization process can be undermined.

Study Design and Setting: Web-based survey conducted in February 2006 using a convenience sample of individuals who are, or have been, involved in some aspect of randomized controlled trials.

Results: Thirty responses were received that described incidences of manipulation. Seven reasons were identified for manipulation: interest of participants, demonstrating treatment efficacy, treatment preference, lack of knowledge, pressure from participants, pressure from trial workers, and practical or technical concerns. In many cases when manipulation was discovered, it was rarely mentioned in the trial publication. Twenty-three responses that described technical errors were received. Technical errors were reported for both the generation and implementation stages of the randomization process.

Conclusions: This study provides further evidence on trial subversion and highlighted that the potential for technical errors can be reduced, and in most cases eliminated, if certain guidelines are followed. Recommendations are as follows: use simple randomization where possible, use third party allocation, test computer randomization programs prior to participant recruitment and ensure that individuals are aware of the procedures needed to be performed if the treatment allocations cannot be accessed using the intended methods. (C) 2008 Elsevier Inc. All rights reserved.

    Research areas

  • Randomization, Subversion, Technical errors, Manipulation, Bias, Online survey, RANDOMIZED-TRIALS, MINIMIZATION, CONCEALMENT, QUALITY, BIAS

Discover related content

Find related publications, people, projects, datasets and more using interactive charts.

View graph of relations