Association between principal investigators’ financial ties to pharma and published trial outcomes
“Medicine is a mess”. This is the view of Ben Goldacre, physician and author of Bad Pharma, in a recent article in the Times newspaper. It comes after the publication of a study by Ahn et al. in the British Medical Journal examining the association between principal investigators’ financial ties to pharma and published trial outcomes.
The scientific and medical communities use results from randomised clinical trials (RCTs) to assess the efficacy and safety of new and currently-used drugs. There are some that suggest the involvement of pharma in RCTs can bias the design, interpretation and publication of results. To investigate this relationship further, Ahn et al. analysed the association between the industry financial ties of principal investigators and study outcome by looking at a random sample of 195 efficacy-related RCTs published in 2013. Their analysis found that 68% of the RCTs reported financial ties between principal investigators and industry, and this rose to 76% among studies with positive outcomes. The authors suggest that publication bias, ties between principal investigators and pharma potentially influencing study design and analysis, and a lack of independent statistical analysis could all be contributing factors. They conclude that “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.”
However, as reported by Ben Goldacre, industry is not the only culprit. Academics and industry alike have been implicated in the non-reporting of clinical trial results and, as in the study above, those that are published tend to have more favourable outcomes. Indeed, everyone involved in the process of drug development, including pharma, academia, regulators, journal editors and health services, has a role to play. Goldacre concludes that all of those involved need to stand up and take ownership of the problem and work out a solution. This is imperative if clinicians are to have confidence in what they read and use the published data to make informed choices on the best treatment options for their patients.
——————————————————–
Summary by Jo Chapman, PhD from Aspire Scientific
Categories
The number of positive trials is higher than you would expect by random chance. That’s not publication bias. It’s because the industry doesn’t test drugs at random, but instead undergoes a long process of R&D first!
Tom – I certainly agree that your explanation is very plausible. However, I also think the non-publication of negative results (which can occur for a huge variety of reasons involving a multitude of stakeholders including publishers), must also come into play. The cited study was performed using data from 2013… and I am pretty sure that even over the past few years – because of greater transparency requirements and awareness – the proportion of positive industry sponsored trials would have decreased.