A comparison of the scientific quality of publicly and privately funded randomized controlled drug trials


  • Copyright: The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non-exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be published in BMJ editions and any other BMJPGL products and sublicences such use and exploit all subsidiary rights, as set out in our licence [http://resources.bmj.com/bmj/authors/checklists-forms/licence-for-publication].

  • Guarantor: Jim Thornton.

  • Conflicts of interest: At the time of the study, R. J., S. Y. and A. M. were medical students whose courses were publicly funded and all were recipients both of a government-funded student loan and of an NHS-funded fifth year student tuition fee bursary. J. T.'s main salary is publicly funded. He has also received fees for speaking and consulting, and reimbursements for attending symposia, from the pharmaceutical industry.

Jim Thornton
Professor of Obstetrics and Gynaecology
Division of Obstetrics and Gynaecology
Maternity Department
City Hospital
Hucknall Road
Nottingham NG5 1PB
E-mail: jim.thornton@nottingham.ac.uk


Background  There is disagreement but few objective data on the relative quality of publicly or privately funded research. Cochrane reviews of randomized trials provide a good comparison opportunity because there is widespread agreement on how trial quality should be assessed and the Cochrane reviewers routinely do this.

Aims  To compare the quality of publicly or privately funded randomized controlled trials.

Methods  A total of 105 trials included in two Cochrane reviews were studied. Their quality assessments were abstracted from the relevant review and information about their funding source was collected from the original trial publications.

Main results  Funding information was obtained for 87 trials. Of these, trials funded by pharmaceutical companies were larger (median sample size 126 vs. 45, P < 0.001), more likely to have avoided ascertainment bias 11/14 vs. 15/41 (P = 0.05). Non-significant trends in avoiding entry bias 19/19 vs. 35/37 and performance bias 13/22 vs. 14/48 also favoured the commercial trials. Commercial trials also had higher recorded attrition rates (median 6% vs. 1%, P = 0.007), but this difference was entirely caused by more non-commercial trials reporting a zero attrition rate.

Discussion  The apparently lower attrition rate in the non-commercial trials should be interpreted with caution. Zero attrition in clinical trials with follow-up of many months is somewhat implausible.

Main conclusion  Commercially funded randomized trials tend to be of higher methodological quality than government-funded ones.