Clinical Context

Research is critical to establish the efficacy and safety of medications, and randomized controlled trials represent the reference standard for interventional research. However, these studies are complex and expensive. Many such trials require financial support from the maker of the drug being studied, which introduces a substantial risk for bias.

Lundh and colleagues evaluated this risk in a meta-analysis published in the December 12, 2012, issue of the Cochrane Database of Systematic Reviews.^[1] A total of 48 studies were included in the systematic review, and research sponsored by industry was more likely than nonsponsored research to have favorable outcomes for benefits, harms, and overall conclusions. Only 2 studies reported on device research only, and there was no significant higher rate of positive results in drug maker-supported studies of medical devices.

In general, industry-sponsored research did not include differences in methodology compared with nonsponsored research. In fact, industry-sponsored research had a lower risk for bias from blinding. The bias in industry-sponsored research was more likely to be found in interpretation of the study data; such research was slightly less likely to feature agreement between data and the conclusions of research.

This meta-analysis failed to account for compensation of the study authors themselves, and the current research by Ahn and colleagues fills this gap.

Study Synopsis and Perspective

Drug effectiveness trials authored by principal investigators who have financial ties to drug makers are more likely than other trials to report favorable results, according to a study published online January 17 in the BMJ.^[2]

Previous studies have shown that industry-funded trials are more likely to be positive than those without industry funding, but these new data suggest that the likelihood of positive results increases when a principal investigator has industry ties, even after controlling for funding sources.

Specifically, trials run by primary investigators with industry ties were more than 3 times more likely to report positive results than were trials led by investigators without such ties. This result was obtained after adjustment for funding source, sample size, study phase, and design and outcome measures (odds ratio, 3.37; 95% confidence interval, 1.43-7.9), according to the study by Rosa Ahn, an MD/MPH student at Oregon Health and Science University in Portland, and colleagues.

The authors reviewed 195 randomized controlled trials published in 2013. They focused on drug effectiveness trials because of their direct effect on practices and healthcare costs.

More Than Half Had Financial Ties

Overall, 58% of principal investigators had financial ties to the drug industry in the form of travel expenses, honorariums, consultant fees, or stock ownership.

Among positive studies, the prevalence of a financial relationship between the principal investigators and the funder was 76% (103/136). Among negative studies, the prevalence was 49% (29/59).

US authors were more likely to have financial ties than were authors from other countries (70% vs 49%; P <.001).

Of the categories of ties that principal investigators reported, the most common was advisor/consultant fees, reported by 39% of principal investigators, followed by speakers' fees and unspecified ties (both 20%) and honorariums and employee relationships (13% each).

"These findings may be suggestive of bias in the evidence base," the authors write.

One strategy for reducing the potential bias is to require publishing of datasets, the authors write. The International Committee of Medical Journal Editors recently proposed that publication of datasets should be a requirement for journal publication, as reported by Medscape Medical News.^[3]

"This requirement is increasingly common in other fields of inquiry such as economics," Ahn and colleagues note.

They also suggest it may be time to bring back independent statistical analysis of major randomized controlled trials, a practice largely replaced by transparency measures.

The authors say that among the reasons for the association between financial ties and positive outcomes may be that negative industry-funded studies may be less likely to get published than the positive ones. Also, single-site randomized controlled trials may not draw the interest of journals.

It could also be that the relationship affects the design of the study and its analytical approach, they say.

The authors acknowledge that the study is observational and cannot prove causation; however, they say, "more thought needs to be given to the roles that investigators, policy makers, and journal editors can play in ensuring the credibility of the evidence base."

Company's True Role Key

In an accompanying editorial,^[4] Andreas Lundh, MD, from the Center for Evidence-based Medicine at Odense University Hospital in Denmark and University of Southern Denmark in Odense, and Lisa Bero, PhD, professor in the Charles Perkins Centre and Faculty of Pharmacy at the University of Sydney in Sydney, Australia, said it is important to know the company's true role in the study.

Some industry-linked trials are authored by opinion leaders "who act as window dressing" to add scientific credibility, but such trials are usually influenced by companies with little author input, they say.

A possible reason for more positive results, they say, is that companies may have a role in the design, such as the population used and comparators. Using an inferior dose of a comparison drug or changing the way outcomes are coded could favor the company's drug, for instance, they write.

"Although journals increasingly require authors to disclose what role a sponsor had in the design, conduct, or publication of a trial, to our knowledge no journals ask authors whether their financial relationships influenced the design, conduct, or publication of a trial," the editorialists write.

Journal editors could also reject work by authors who will not share data and penalize those who do not fully disclose financial relationships.

"Trials with industry funding or authors with financial ties should be interpreted with caution until all relevant information is fully disclosed and easily accessible," they conclude.

This project was not directly supported by any research funds. One coauthor has received support from the National Institutes of Health and Veterans Affairs Health Services Research and Development Service. Another coauthor's work on this article was supported by a Cancer Center Support Grant from the National Cancer Institute to Memorial Sloan Kettering Cancer Center. Dr Lundh is a member of the Danish organization Physicians without Sponsor, which seeks to promote independence from commercial influence in clinical work, research, and education. The other authors and editorialist have disclosed no relevant financial relationships.

BMJ. Published online January 17, 2017.

Clinical Implications

• A previous systematic review by Lundh and colleagues found that research sponsored by industry was more likely than nonsponsored research to have favorable outcomes for benefits, harms, and overall conclusions. Sponsored research was more likely to draw conclusions unsupported by study data.
• The current study by Ahn and colleagues suggests that studies in which the first author receives some form of compensation from the drug maker are more likely to include positive results.
• Implications for the Healthcare Team: A torrent of clinical data is available on a daily basis; therefore, keeping track of all of these data and practicing some healthy skepticism toward the results are daunting tasks. The healthcare team may help by using solid medical evidence to construct algorithms that promote evidence-based care at the local level.

CME Test