Non-publication of large randomized clinical trials: cross sectional analysis

BMJ 2013; 347 doi: http://dx.doi.org/10.1136/bmj.f6104 (Published 29 October 2013)
Cite this as: BMJ 2013;347:f6104

Recent rapid responses

Rapid responses are electronic letters to the editor. They enable our users to debate issues raised in articles published on bmj.com. Although a selection of rapid responses will be included as edited readers' letters in the weekly print issue of the BMJ, their first appearance online means that they are published articles. If you need the url (web address) of an individual response, perhaps for citation purposes, simply click on the response headline and copy the url from the browser window.

Displaying 1-5 out of 5 published

Author’s Reply:
Thank you, Dr. Altus, for your insights into the classification of study funding.

We would also like to thank Drs. van Lent and Out for pointing out the distinction between industry-sponsored and industry-supported trials. We acknowledge that there are multiple approaches currently used to categorize trials by funding source, and we were pleased to see the authors’ recent publication attempting to standardize this practice.

We did not search for trial data in industry-sponsored registries. We chose publication in a peer-reviewed journal as our primary outcome because this form of publication makes the results widely accessible, helps ensure that methodological details are provided to assess internal and external validity, and encourages the presentation of results within the appropriate context. Among unpublished trials, we chose to look for results within the ClinicalTrials.gov results database because this is a standardized, permanent, and independent repository of trial data. Further, the 2007 FDA Amendments Act established ClinicalTrials.gov as the largest public database in which applicable trials are required by law to report results. While we agree that other sources of trial data exist, our goal in searching ClinicalTrials.gov was to determine the extent to which the site’s results database serves as a source of unpublished trial data.

Christopher W. Jones, MD
Timothy F. Platts-Mills, MD

Competing interests: None declared

Christopher W. Jones, Physician

Timothy F. Platts-Mills

Cooper Medical School of Rowan University, One Cooper Plaza, Suite 152, Camden, NJ, 08103

Click to like:

I agree with author that the trials results should be published. For this there have to be inbuilt checks in the trials and the funding agencies. The installments of money should be disbursed after the interim reports of the trials are approved by the experts in that area by the funding agency periodically, failing which the trial should be stopped mid-way. There are other ways of penalizing both the scientists of the trial or the funding agency, e.g. banishing the future work or funding or at least debar them for some time.

Competing interests: None declared

Neeru Gupta, Scientist E

Indian Council of Medical Research, Ansari Nagar, New Delhi-110029

Click to like:

The article by Jones and colleagues is very important for the scientific community as well as for the consumers’ and patients’ community (1). For the patients that gave their informed consent to enter a randomised clinical trial it is difficult to accept that a very large percentage (29%) of results of clinical trials are not published and thus they will not be available for clinical practice and further research. All patients look at the results of clinical research with interest hoping that they will be applicable to improve the diagnosis and therapy of their disease. For these reasons a large number of them accept to participate in clinical research. The non-transparency and non-publication of results has been considered a cause of great concern and deep thinking for researchers and patients' representatives (2).

Recently we have collected data on the availability of clinical trial registries information in webpages of research hospitals and patients associations. Results showed that in most cases the tools used to increase transparency of clinical research are neglected (3). At the moment, in the framework of activities of the Fondazione Nerina e Mario Mattioli for research on ovarian cancer, we are matching the publication of protocols on the Clinicaltrials.Gov registry with the availability of results in the Clinicaltrial.Gov section or the related publication on PubMed. Preliminary findings showed a similar picture as presented by Jones and colleagues: more than 80% of the results of ovarian cancer clinical trials are not reported in Clinicaltrial.Gov section, and more than 42% remained unpublished.

It's important that clinical results are constantly documented and discussed in a transparent way and publication of results and the use of registers of clinical trials should be strongly encouraged. It is also important that these data are disseminated among citizens, patients and their representatives, in order to work together for the improvement of clinical research credibility.

References
1. Jones CW, Handler L, Crowell KE, Keil LG, Weaver MA, Platts-Mills TF. Non-publication of large randomized clinical trials: cross sectional analysis. BMJ 2013; 347: f6104.
2. Liberati A. Need to realign patient-oriented and commercial and academic research. Lancet 2011; 378: 1777-1778.
3. Mosconi P, Roberto A. Open-access clinical trial registries: the Italian scenario. Trials 2012; 13: 194-198.

Competing interests: None declared

Paola Mosconi, Researcher

Anna Roberto

IRCCS-Istituto Mario Negri and Fondazione Nerina e Mario Mattioli, Via G. La Masa 19, Milano

Click to like:

Thank you, Dr. Outs, for distinguishing between industry-sponsored trials and industry-supported trials. I agree that this distinction makes a difference, and as such is helpful for analyses like yours. However, in a concern based on the dictum “follow-the-money”, I do not find such a difference.

Pharmaceutical companies paid for studies, whether the companies had formal responsibility as in industry-sponsored trials, or did not have formal responsibility as in industry-supported trials. Financial backers of an activity ought to acknowledge and not disavow the shortcomings of the activity. Otherwise, the consequence could become one of shifting responsibility (passing the blame) because “we only paid for it”.

Competing interests: None declared

Michael S. Altus, Freelance medical writer-editor

Intensive Care Communications, Inc.®, 3626 Fords Lane, Baltimore MD 21215 USA

Click to like:

We very much agree with Christopher W Jones and colleagues that non-publication of trial results is unethical towards research subjects who are, as a result, unnecessarily exposed to risks and burdens of trial participation.

In their article, Jones et al. report that non-publication was more common among trials that received industry funding (150/468, 32%) than those that did not (21/117, 18%), P=0.003. Of the 150 unpublished industry funded studies, 38 (25%) had results available on ClinicalTrials.gov, while none of the 11 unpublished trials funded by the National Institutes of Health or the 20 unpublished trials with other funding sources provided results in the results database.

In determining the potential role of the funding source on non-publication, the authors did not discriminate between industry-supported or industry-sponsored trials. Sponsorship has been defined in ICH-GCP guidelines (1) and constitute formal responsibility for the conduct and reporting of the trial. In trials that are just financially supported by industry, that formal responsibility still lies primarily with the non-profit organization. We have previously recommended to make this distinction (2) and have shown its relevance in publication bias (3).

Furthermore, Jones et al. have only included ClinicalTrials.gov to find results of unpublished trials. They may have missed results of trials that were posted in industry results registries (clinicaltrials.ifpma.org, registries of individual pharmaceutical companies). Therefore, they may have underestimated the availability of results of particularly industry-sponsored trials in results databases.

References
1. International conference on harmonisation of technical requirements for registration of pharmaceuticals for human use. ICH Harmonised Tripartite Guideline - Guideline for Good Clinical Practice E6. 1996.
2. van Lent M, Overbeke J, Out HJ. Recommendations for a uniform assessment of publication bias related to funding source. BMC medical research methodology. Sep 30 2013;13(1):120.
3. van Lent M, Overbeke J, Out HJ. Role of Editorial and Peer Review Processes in Publication Bias: Analysis of Drug Trials Submitted to 8 Medical Journals. Presented at: Seventh International Congress on Peer Review and Biomedical Publication 2013; Chicago. http://www.peerreviewcongress.org/2013/Plenary-Session-Abstracts-9-9.pdf

Competing interests: Henk Jan Out is a paid employee from Teva Pharmaceuticals next to his professorship at the university.

Marlies van Lent, Researcher

Henk Jan Out

Radboud University Medical Centre , PO Box 9101, 6500 HB, Nijmegen, The Netherlands

Click to like:

THIS WEEK'S POLL