Intended for healthcare professionals

CCBYNC Open access
Research

Sharing of clinical trial data among trialists: a cross sectional survey

BMJ 2012; 345 doi: https://doi.org/10.1136/bmj.e7570 (Published 20 November 2012) Cite this as: BMJ 2012;345:e7570

Re: Sharing of clinical trial data among trialists: a cross sectional survey

Rathi and colleagues1 have provided more insight into clinical trialists’ opinions and experiences of sharing of clinical trial data. Although the results as they stand are interesting on their own, we would like to raise two issues related to these findings.

First, the study by Rathi and colleages1 as well as previous studies referred to within the article, are limited to sharing data. We would like to take this one step further and plead for also sharing questionnaires, intervention manuals, analysis scripts (e.g. syntax files), output files, and other materials that are used to generate study reports. Accurate replication requires these additional materials. Since successful replications strengthen our evidence that a given theory, model, or assumption holds, whereas unsuccessful replications are our only means to falsify theories, models, or assumptions,2 this is essential for scientific progress. Publication of analysis scripts and output files is necessary as researchers make a lot of important choices that are not disclosed in research articles.3

Second, the concerns regarding data sharing that Rathi et al.’s respondents indicated can mostly be refuted.4 For example, one concern was the ability to publish own research: when a researcher wants to publish several articles about one dataset, publishing the dataset before all articles are published can be risky, as others might conduct analyses you planned in later articles and publish those earlier. A solution is not to publish the entire dataset, but only those variables described in published articles. This is related to another identified reason: sufficient academic or scientific recognition. In the rat race of ‘publish or perish’, it might seem as if one helps others by publishing materials (data, questionnaires, intervention manuals, etc.). However, materials are only published after acceptance of the article; so any race has already been won—or lost.

References
1. Rathi V, Dzara K, Gross CP, et al. Sharing of clinical trial data among trialists: a cross sectional survey. BMJ 2012;345:e7570.
2. Ritchie SJ, Wiseman R, French CC. Failing the future: three unsuccessful attempts to replicate Bem’s “retroactive facilitation of recall” effect. PLOS ONE 2012;7:e33423.
3. Simmons JP, Nelson LD, Simonsohn U. False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science 2011;22:1359–66.
4. Peters G-JY, Abraham C, Crutzen R. Full disclosure: doing behavioural science necessitates sharing. The European Health Psychologist 2012;14:77-84.

Competing interests: No competing interests

02 December 2012
Rik Crutzen
Senior researcher
Gjalt-Jorn Y. Peters, Charles Abraham
Maastricht University/CAPHRI
P.O. Box 616, 6200 MD Maastricht, The Netherlands