Skills Lab: How to Read a Phase 3 Clinical Trial Paper

This transcript has been edited for clarity. 
Hello, everyone. This is Dr Bishal Gyawali from Queen’s University, Kingston, Canada, and I’m happy to be back with you. This is Skills Lab: How to Interpret Clinical Research With Dr Bishal Gyawali. Today we’ll discuss how to read a phase 3 clinical trial publication. 
Let’s say you have the PDF in your hand of a new drug trial that has been published, which is a phase 3 randomized trial. How do you go about reading that? There are a couple of points that I want to highlight that you should look at carefully. 
Before we start delving into the paper, we need to look at the following: (a) who the sponsor or the funder of the trial is; (b) what the role of the sponsor or the funder was; (c) who are the authors; and (d) are there any conflicts of interest among the authors?
You might be surprised that I’m asking this question of who wrote the paper, because the obvious answer would be, of course, that the authors wrote the paper. As you will see in many new cancer drug trials, the authors actually don’t write the paper. I’ll give you some examples of that.
First, you need to check who funded the trial. Usually for new cancer drugs, almost 100% of the time, it’s funded by the industry that manufactures the drug. Depending on which journal you are looking into, there are specific places where you can find this information. For example, in The New England Journal of Medicine, you’ll find this information at the end of the abstract.
Second, you should look at the conflicts of interest and the affiliations of the authors. In many of these papers, you will see that there are some authors from the company itself in the authorship byline. 
Third, as I mentioned, we want to look at the role of the funder. In some of the trials, you will see a statement saying that the funder had no role in the study design, data collection, data analysis, data interpretation, or writing of the report, which is perfect, and that is what the role of the funder should ideally be.
In many of the new cancer drug trials, you’ll see that the funder was involved in the design, collection, analysis, and interpretation of the data, as well as the writing of the manuscript. I have no problem with the funder being involved in designing the trial and collection of the data, but the analysis should be done independently of the funder, who has a big financial stake in the outcomes.
The analysis may be done by the funder, but at the very least, the interpretation should not be done by the funder. We see in most cases that the interpretation is also being done by the funder. 
Most important is the writing of the manuscript. We would assume that authors write the manuscript, but you should look carefully into the trial publication. In most of these cases, you’ll see that even the manuscript was not written by the authors, but by professional medical writers who were funded by the industry. 
Why am I being so critical about this? When you write a manuscript and you have billions of dollars at stake, then you will try to write the manuscript in a way that will make your drug look much better than what it actually is. This is what we call a spin. 
If you see these red flags in the manuscript, that the manuscript was written by professional medical writers funded by the company, then you need to very carefully look into whether there are several spins in the trial, and you should be very careful about interpreting the trial results based on how the publication is interpreting it. Do not rely on the publication’s interpretation. We have to interpret it ourselves, and that’s what we are going to do in this course. 
Just to give you an example of the conflict of interest of the authors, which can also be important in how the results are framed, even if the paper is not written by professional writers and is written by the authors, if the authors have a huge conflict of interest, then that can also lead to bias and spin.
In one of our studies where we looked at highly paid oncologists in the United States, we saw that among the 1% of medical oncologists in the US who received industry payments of more than $100,000 per year, the median value of payment they received was $150,000 per year. The highest earners would earn more than their annual salary through conflict of interest, through payments from the industry. 
Now, when such a huge amount is at stake, when people are earning more from consulting jobs in the industry than from their actual work, then it’s not difficult to understand why people might want to present the results in a much more favorable light than what the data show.
That’s some background information that we need to look into even before we start delving into the paper. Thank you.