OHDSI Home | Forums | Wiki | Github

URGENT CALL TO ACTION: Assessing safety GLP-1 use in obesity


Yesterday, JAMA published a research letter, which is extremely relevant for us: Sodhi et al, JAMA 2023. The authors looked at the risk of GLP-1 agonists (Ozempic and Saxenda) for weight management before it became indicated for that. They reported increased risks of pancreatitis, bowel obstruction, and gastroparesis, some of them dramatic! If this is true, it would be a big deal, and it already has picked up a ton of media attention.

However, they didn’t exactly abide by the LEGEND principles of how to do this robustly. They used just one database (PharMetrics), half a handful of handpicked covariates to adjust confounding with, and their cohort definitions are less than precisely defined. Only one comparator and no negative controls.

We should do something about this and close the evidence gap. It’s important for patients, providers, and the broader community to be well-informed of the associated risks of a class of drugs which have gained huge popularity. And if the findings are flawed, it’s our responsibility to correct the narrative and protect public trust.

But we have VERY LITTLE time. Within 4 weeks we can write a letter to the Editors. Other folks are probably doing the same thing as we speak.

I am calling the community for help with DATA who can run a package and can do it in two weeks! We (the OHDSI Center at Northeastern) need patients with those prescription drugs and the above outcomes, which probably are more likely hospital issues. I am planning to send out the package by EoD Tuesday. We will start revieiwng the results on Tuesday October 24, right after the symposium. :slight_smile:

Please let me know.

tagging @Patrick_Ryan @slseager78 @atifadam @PriyaDesai @Andrew @Paul_Nagy @hripcsa @msuchard @scottduvall @rkwpnw and all other data partners.


Tagging Colorado’s OHDSI analyst @mgkahn!

1 Like

IQVIA will run the package especially against Pharmetrics