Thread by @JohnArnoldFndtn

I came across a study on Twitter that made the claim that a pharma patient support program reduces healthcare costs. I like lower healthcare costs so I click through. Instead of coming across intriguing research, I found the worst of industry sponsored propaganda. 1/12

The claim is that AbbVie’s program that reduces out-of-pocket costs for patients taking the highest revenue drug in the US, Humira, leads to lower healthcare costs. If true, it’s important for health policy. Such a policy would also increase AbbVie’s profits. 2/12

Allow me to be a cynic for a minute. If I were a pharma exec and wanted to get junk science published to support a health policy that benefited my company, here are the steps I would take: 3/12

Pay researchers consulting fees for years to make sure they are partial to your interests.
Check.

Make sure to embed the company’s employees into the research team.
Check.

Have direct influence on the design, study conduct, and interpretation of the data.
Check. 6/12

Most important, make sure to use retrospective data that’s already been analyzed before hiring outside researchers rather than a prospective RCT.
Check. 7/12

Have a professional firm, not the external researchers, write the research report to ensure it emphasizes the aspects beneficial to the company.
Check. 8/12

Make sure to have final approval of the manuscript to be able to bury the study should it somehow find a negative result.
Check. 9/12

I hesitate to even publish a link to this junk but here it is. 10/12
jmcp.org/doi/pdf/10.185…

In summary, pay external researchers already friendly to the company to work with company employees to use an already analyzed data set using company-specified methods. Have a for-profit company write the report. Maintain option to bury the study if don’t like result. 11/12

Use the “study” to advocate for a policy change that would benefit company.
This isn’t science. This isn’t how evidence-based policy is supposed to work. This is special interests hijacking the system. 12/12

The only positive I see is that the standards on disclosure of conflicts of interest seem to be working. Although buried at the end, they are there. Years ago many of these would not have been disclosed at all.