Adults recruited through Prolific who reported at least three months of weekly semaglutide or tirzepatide use submitted a photograph of their medication, which two blinded coders classified as a branded formulation or a compounded product. Among 322 participants, self-report and photo coding agreed for 82.6% (kappa 0.78), and 83 (25.8%) were coded as using compounded GLP-1 receptor agonists, two-thirds of whom had named a branded product.
Why it is interesting: It puts a number on the share of GLP-1 users outside claims and prescribing data, and on how far self-report misclassifies them.