4

I used a commercial tool created by a small company to perform quantitative analysis of medical images for a study that I did. I was very conservative in my statistical analysis and consulted with a university-employed biostatistician throughout the process. The results were negative. I presented the results at a conference and have submitted the paper to a journal where it is currently under review.

The problem is that the founder of the company whose tool I used came up to me after the presentation and started to interpret my results differently, putting them in a more positive and hopeful light. He is an engineer and is, of course, highly knowledgeable about the tool we used since he developed it. However, he wants to help us re-analyze the data since he 'saw some interesting results' in my presentation. I feel uncomfortable about this since I believe there is a conflict of interest at play if we were to work directly with him. I think he has a clear economic incentive to promote his analysis tool, and I can see how the results of our study might influence the perception of his product.

Am I overthinking this situation? What level of cooperation with this person would be acceptable from a COI perspective?

  • 2
    Even if helps analyzing the data, that does not mean that you need to publish the new findings if they do not make sense! – The Guy Jun 22 '17 at 12:34
4

I don't see a problem. Many people create a tool, use it to analyze something, then publish the results. Many companies that produce products publish papers that try to put their products in a good light. It is the job of both the authors and the reviewers to do all this in a rigorous fashion, but it doesn't mean such papers are disallowed.

1

It appears there are two stages of the analysis. First, that of individual images, which is done using the commercial software tool. Second, the statistical analysis and interpretation of a sample of the first. So, in which of these two is said person proposing a re-analysis? Clearly, their realm is only the first type of analysis (though they may have some well-founded ideas about the second part). If they offer to re-analyse just the first part, I don't think there is any problem, as long as you apply the same statistical scrutiny in the second phase.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.