Restoring invisible and abandoned trials: a call for people to publish the findings

On 15 July 2013, I signalled intent to work with a team of scientists to republish study 329 of paroxetine in children and adolescents, in accordance with the RIAT guidelines. Shortly afterwards GSK set up a process for researchers to ‘submit research proposals and request anonymised data from clinical studies’. Although they only included studies conducted since 2007, they offered the opportunity to ‘enquire about the availability of data from our clinical studies that are not listed on the site before they submit a research proposal’. I made such a query in relation to study 329 on 4 August. GSK’s website designates the status of my query as ‘under review’. It is not seem plausible that it would take 3 weeks to decide whether data is available.

In the absence of any apparent means to find anything more through the GSK website, I post this rapid response as an open letter to GSK seeking more information.

Following the publication of the Restoring invisible and abandoned trials (RIAT) paper [1], GlaxoSmithKline (GSK) was notified by the RIAT authors by email on June 14 that study 329 was amongst the studies requiring correction. Study 329 was published in 2001 in the Journal of the American Academy of Child and Adolescent Psychiatry as ‘Efficacy of paroxetine in the treatment of adolescent major depression: a randomized, controlled trial’[2]. This paper seriously misrepresented the outcomes of study 329, as I and others have extensively documented.

GSK was asked to signal intent to publish a corrected version by sending an electronic response to the RIAT article within 30 days. Since they have not declared their intention to restore the record within this timeframe, I write to declare my intent to work with a team of scientists to republish study 329 of paroxetine in children and adolescents, in accordance with the RIAT guidelines. We plan to use the Clinical Study Report available on the GSK website [3], and other documents relating to Study 329 that are publically available through litigation [4] to prepare a report of outcomes for 93 adolescents treated with paroxetine, 95 with imipramine, and 87 with placebo.

1. Doshi P, Dickersin K, Healy D, Vedula SS, Jefferson T. Restoring invisible and abandoned trials: a call for people to publish the findings. BMJ. 2013 Jun 13;346:f2865
2. Keller MB, Ryan ND, Strober M, Klein RG, Kutcher SP, Birmaher B, et al. Efficacy of paroxetine in the treatment of adolescent major depression: a randomized, controlled trial. J Am Acad Child Adolesc Psychiatry 2001;40:762-72.
3. http://www.gsk.com/media/resource-centre/paroxetine/paroxetine-paediatri...
4.http://www.healthyskepticism.org/global/news/int/hsin2010-01

Dear Dr Godlee,

I applaud the RIAT concept developed by Doshi, Dickersin, Healy, Vedula and Jefferson (1)

I am a former scientist and information specialist who has carried out mediated literature searches for health professionals and taught information research skills to tertiary health science students and academics. It is clear, and has been for some years, that much of the published medical evidence base, but not all of course, is flawed (only partly because of unavailability of important unpublished clinical trial data) (2). In the light of this fact I believe that all health library professionals need to look to their sense of ethics and question the implications of their literature searching activities.

My main interest in RIAT stems from the well known issue of the massive consumption of antidepressants in many countries, and subsequent harms to many, something which has been debated recently in this journal(3). Therefore, I particularly noted the paper by Erick H. Turner and colleagues cited by Doshi et al (4). It examined the published and unpublished trial data submitted to the US FDA as required for the approval process for various antidepressants. For example, Pfizer’s Zoloft (sertraline) was approved as a treatment for depression in 1991, but of the five relevant clinical trials studied by Turner et al, only two were published and only one of the published trials (5) showed any evidence of a therapeutic effect. The data in the other published trial (6) was considered by the FDA to be negative or doubtful, but was eventually published as if it was positive. I hope that Dr Turner or one of his colleagues will act on RIAT. Interestingly, Doshi et al note that they have unpublished Pfizer sertraline trial documents (from study 206, dated 1983) in their possession. The sertraline studies examined by Turner were numbered 101, 103, 104, 310 and 315, and didn’t include study number 206.

Sincerely,
P. Jane Wilson

1. Doshi P et al Restoring invisible and abandoned trials: a call for people to publish the findings BMJ 2013;346:f2865

2. Scott IA, Glasziou PP. Improving the effectiveness of clinical medicine: the need for better science. Med J Aust. 2012 Mar 19;196(5):304-8

3. Spence D. Are antidepressants overprescribed? Yes. BMJ. 2013 Jan 22;346:f191

4. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med. 2008 Jan 17;358(3):252-60.

5. Reimherr FW, Chouinard G, Cohn CK, Cole JO, Itil TM, LaPierre YD, Masco HL, Mendels J. Antidepressant efficacy of sertraline: a double-blind, placebo- and amitriptyline-controlled, multicenter comparison study in outpatients with major depression. J Clin Psychiatry. 1990 Dec;51 Suppl B:18-27.

6. Fabre LF, Abuzzahab FS, Amin M, Claghorn JL, Mendels J, Petrie WM, Dubé S,Small JG. Sertraline safety and efficacy in major depression: a double-blind fixed-dose comparison with placebo. Biol Psychiatry. 1995 Nov 1;38(9):592-602

Dear readers,

On June 19, we received an email from Matthew Cahill, Associate Vice President of Merck Global Scientific and Medical Publications, who pointed out that results from Merck's rofecoxib (Vioxx) study 078, which appears in our list of clinical study reports in our possession (http://www.bmj.com/content/346/bmj.f2865#T1), have been published in the following two articles:

1. Thal LJ, Ferris SH, Kirby L, Block GA, Lines CR, Yuen E, et al. A randomized, double-blind, study of rofecoxib in patients with mild cognitive impairment. Neuropsychopharmacology. 2005 Jun;30(6):1204–15. http://www.ncbi.nlm.nih.gov/pubmed/15742005

2. Aisen PS, Thal LJ, Ferris SH, Assaid C, Nessly ML, Giuliani MJ, et al. Rofecoxib in patients with mild cognitive impairment: further analyses of data from a randomized, double-blind, trial. Curr Alzheimer Res. 2008 Feb;5(1):73–82. http://www.ncbi.nlm.nih.gov/pubmed/18288935

Therefore we would like to alert readers to these two publications.

We are also aware of analyses that indicate the Thal et al. 2005 primary publication of study 078 was misreported and ghostwritten. On misreporting, see Madigan et al [1]. On ghostwriting, see Ross et al. [2].

Please note that Table 1 of our paper is a list of clinical study reports in our possession. While a great number of these studies remain unpublished many years after their completion, and others are documented to be misreported in journal publications, Table 1 is not a definitive list of abandoned studies. Determining abandonment takes time and effort, and we are offering the clinical study reports in our possession to authors committed to restoring abandoned trials to correct the scientific record where necessary.

Sincerely,
Peter Doshi, PhD
Kay Dickersin, MA, PhD
David Healy, MD
Swaroop Vedula, MD, PhD
Tom Jefferson, MD

References:

[1] Madigan D, Sigelman DW, Mayer JW, Furberg CD, Avorn J. Under-reporting of cardiovascular events in the rofecoxib Alzheimer disease studies. American Heart Journal. 2012 Aug;164(2):186–93.

[2] Ross JS, Hill KP, Egilman DS, Krumholz HM. Guest authorship and ghostwriting in publications related to rofecoxib: a case study of industry documents from rofecoxib litigation. JAMA. 2008 Apr 16;299(15):1800–12.