What does “collaboration” in medical research mean exactly? Most clinical drug trials today are collaborations between industry funders and academic researchers, but it’s not clear how much control these funders have over the study design, analysis, and reporting. While researchers acknowledge that industry funding is a valuable resource, some are concerned that collaborating with industry may sacrifice academic freedom and lower the quality of research.
A new analysis in The BMJ helps answer this question by examining the links between study design, funding, and results. Researchers analyzed the roles of industry funders and academic authors in 200 recent industry-funded trials in prominent medical journals, to provide a detailed illustration of what each partner was contributing. They found that employees of industry funders had a large role in study design, data analysis, and reporting:
Having control over study design, including the outcomes, control groups, and comparators, can make a big difference in the trial’s eventual results. For example, using a surrogate endpoint for the trial, such as “progression-free survival” in cancer patients, makes it easier to show success, but these outcomes are not intrinsically meaningful to patients as are “hard” outcomes like overall survival or quality of life.
Similarly, funders may choose a “comparator” drug for the study (a currently used drug compared to the new drug) that is not as good as other drugs commonly used – this makes it look like the new drug is much better. In some trials, there is no placebo arm at all, making it impossible to tell whether the trial drug is doing better than a placebo would have.
The authorship of the manuscript is also critical, because whoever writes the content has the opportunity to frame the results of the study. For example, if a study is designed to show that a new drug is better than an old drug and the results are not statistically significant, that means the trial was negative. However, often authors “spin” the results by saying that the new drug is “just as safe and effective” as the old drug, even though that wasn’t the point of the trial. Medical writers used this tactic in writing the results of a drug trial called RELEVANCE, which Dr. Vinay Prasad noted on his podcast Plenary Session. Another popular tactic is ignoring the failure of the primary endpoint in favor of a secondary endpoint, or post-crossover group, as some did in presenting the CABANA trial.
In his criticism of the RELEVANCE trial, Prasad makes deliberate note of the impact of medical writers. “There are a number of sentences in this article that I would wonder if they would still be there if the article was written by investigators and not written by someone who wants to portray this failed trial in as favorable light as possible,” he said.
Study design matters. Authorship matters. Given that most drug trials are funded by industry, and that these collaborations give industry large role in study design, analysis, and writing, we should be thoughtful in looking at the results of these trials. In the long run, we need more non-conflicted funding sources for trials if we want to raise the quality of medical research for new drugs and devices.