Interview conducted by April Cashin-Garbutt, BA Hons (Cantab)
Please can you give a brief introduction to drug and device studies?
Our review included analyses of drug and device studies conducted in humans. The drugs or devices could be compared to placebo or sham treatments, or other effective treatments. The drugs and devices examined for this review are used for a wide variety of clinical conditions, including cardiovascular and psychiatric diseases.
Please could you give some examples of the drug and device studies that you reviewed?
The papers included in our review examined over 1700 different studies. These included randomized controlled trials of psychotropic drugs, drugs to treat acute pain and migraine, statins, surgical interventions, and treatments for cancers including breast, colo-rectal and non-small-cell lung cancer.
Do these studies influence the recommendations that doctors make about drugs and devices?
Yes, doctors rely on the results of clinical trials to determine what works and what does not work. In addition, drug and device trials are particularly influential when they are included in systematic reviews and clinical practice guidelines that provide guidance on treatment decisions.
How many drug and device studies are sponsored by pharmaceutical industry? What other sources of sponsorship are available?
I can’t tell you the specific number, but an increasing number of drug trials are sponsored by the pharmaceutical industry. According to the most recent reference we could find, during the period 2003 to 2008, the pharmaceutical industry spent more on clinical research than the National Institutes of Health in the United States.
Although government funding is the other main source of funding, the pharmaceutical industry is the main sponsor of both pre- and post-marketing studies. This presents a conflict of interest because if the sponsoring company’s product is found to be ineffective or harmful, the company is at a financial risk.
Is this conflict of interest managed in any way?
Most conflict of interest management strategies focus on managing the conflicts of individual investigators (for example, putting limits on consulting agreements) and not the sponsorship of entire studies.
Could this management be improved?
Journals are increasingly asking that industry sponsored studies include a statement that the sponsor did not play a role in the design, conduct or dissemination of the research. However, this is not done consistently.
Please can you outline what was previously known about the research outcomes of industry sponsored clinical trials?
Previous systematic reviews have shown that pharmaceutical industry sponsored drug trials are more likely to have results and conclusions that favour the sponsor’s product compared to trials funded by others.
How did your work differ from previous systematic reviews?
This review approximately doubles the number of studies that were included in previous reviews and expands the review to include device studies.
In addition, we examined whether industry sponsored studies have different risks of bias compared to studies with other sponsors.
What did your review find?
We found that published industry sponsored studies reported greater benefits and fewer harmful side effects for the sponsor’s product than those reporting on trials that were not sponsored by industry. The reports of industry sponsored studies also presented more favourable overall conclusions, and the results and conclusions sections in these papers were less likely to agree with each other. In addition, when two drugs were compared head-to-head in studies sponsored by different companies, the drug that compared favourably in terms of efficacy or harm, was most often the drug manufactured by the sponsor of that study.
In the studies that assessed the risks of bias in the drug and device trials, the association of industry sponsorship and research outcomes favourable to the sponsor could not be explained by the industry studies having different risks of bias compared to the non-industry sponsored studies.
What impact do you think your work will have?
We hope that our review has an impact on how decision makers, systematic reviewers and guideline developers evaluate the validity of drug and device trials. There are many ways that industry sponsors can influence the outcomes of studies, including how the question is asked, as well as how the study is designed, conducted and disseminated.
Traditional tools for assessing risks of bias do not capture all of these ways to influence a drug study, so decision makers need to take sponsorship into account as a potential source of bias. In addition, the findings of our review support the growing efforts to make the full results of drug trials available to the public, as well as the need for more public funding for drug studies.
Are there any plans in place to restrict the number of industry sponsored drug and device studies?
To my knowledge, no, but it is not just a issue of restricting the sponsorship, but making sure that the studies that are sponsored have public health relevance and are not simply marketing tools to promote newer products. This is where public funding, or the pooling of public and private funding, could have an impact. If public funders were involved, then the questions asked and the comparison groups tested could be more relevant for public health.
How do you think public funding for drug studies could be increased?
In Italy, there was a model to collect fees from drug companies based on the amount of money the companies spent on marketing. These fees were pulled together into a public fund to sponsor the types of needed drug studies that individual companies would typically not fund (for example, comparisons of new products to generics). In addition, insurers who pay for drugs, could pool funds with public agencies to sponsor drug studies.
What do you think the future holds for drug and device trials?
Increasing scepticism towards industry-sponsored studies will support the push towards more transparency regarding how the studies are designed and conducted, as well as the full reporting of results.
Would you like to make any further comments?
For more information in a podcast on our review, please see cochrane.org/podcasts.
Where can readers find more information?
They can find our research paper here: http://onlinelibrary.wiley.com/doi/10.1002/14651858.MR000033.pub2/abstract
About Professor Lisa Bero
Lisa A. Bero, PhD, Professor and Vice Chair, Research, Department of Clinical Pharmacy, School of Pharmacy and Institute for Health Policy Studies, School of Medicine, University of California, San Francisco (UCSF), is a pharmacologist with primary interests in how clinical and basic sciences are translated into clinical practice and health policy.
She chaired the UCSF Chancellor’s Advisory Committee on Conflicts of Interest for 11 years. She has developed and validated methods for assessing bias in research and scientific publication and measures influences on the quality of research, including university-industry relations.
Dr. Bero has also conducted analyses to examine the dissemination and policy implications of research evidence. She has published numerous peer-reviewed scientific articles related to her research as well as co-authored The Cigarette Papers (UC Press, 1996).
Her international activities include advisor to the World Health Organization Department of Essential Medicines and Pharmaceutical Policies, member of the WHO Essential Medicines Committee and member of the Pan American Health Organization Advisory Committee on Health Research.
She is Director of the San Francisco Branch of the United States Cochrane Center. She was an elected member of the Cochrane Collaboration Steering Group for 12 years and serves on several national and international committees related to conflicts of interest and research, such as the Institute of Medicine Committee on Conflict of Interest in Medical Research, Education and Practice.