What 49 Online GLP-1 Prescribers Required

A secret-shopper study of 49 US websites found that getting a GLP-1 prescription online was usually easy, while direct clinical contact was often absent. Researchers reported that 92% of the sites prescribed a GLP-1 medicine to a fictional patient who met treatment criteria, and 69% went on to mail medication. The result documents how the sampled sites handled one eligible profile. It does not show that every telehealth service works this way, nor does it measure whether patients were helped or harmed^[1][2].

The paper, Online Prescribing of GLP-1 Receptor Agonists, was published in JAMA in July 2026. It examined access to medicines in the same class as semaglutide and tirzepatide. The finding is about the prescribing process, not a comparison of those medicines. It is not evidence that either one should be started or stopped without individual clinical review^[1].

How the secret-shopper study worked

The Yale team created a profile for a 27-year-old man who qualified for GLP-1 treatment, then attempted to obtain prescriptions through 49 online sellers. A secret-shopper design lets researchers record what a customer encounters rather than relying on a site's marketing claims. Here, the researchers recorded prescription offers and shipments. They also tracked clinician engagement^[1][2].

The Yale account says some requests placed on a Monday evening were approved soon afterward, with medication delivered by Tuesday or Wednesday. It also reports that most sampled companies sold compounded versions. The paywalled STAT story independently exposes only its opening account: the study covered nearly 50 telehealth sites and found online access extraordinarily fast and easy. We do not use the inaccessible portion of that story for additional figures^[2][4].

What the 92% and 69% figures mean

Forty-five of 49 sites is approximately 92%, while 34 of 49 is approximately 69%. The first figure describes a prescribing decision for the study profile. The second indicates that medication was actually mailed. Neither percentage is a treatment-success rate. No participant took a drug. The study measured neither weight change nor adverse events, and it did not track adherence or long-term follow-up^[2].

The clinician-contact finding may be the more important signal. Yale reported that the vast majority of sites did not require communication with a clinician, and that the shopper often obtained a prescription without speaking with one to verify the submitted information. A form can collect a medical history, but this study raises a narrower question: whether the process allowed meaningful clarification before prescribing^[2].

Why clinical contact matters

GLP-1 medicines are prescription drugs, and suitability cannot be reduced to a weight or body mass index entry. A prescriber may need to clarify medical history and current medicines. Contraindications, prior treatment, and the patient's goals can matter too. The study did not test whether every omitted conversation would have changed a decision. Many sampled pathways nevertheless reached approval without real-time clinician-patient communication^[2].

The FDA now lists the absence of screening and a prescription by a licensed doctor as a telehealth warning sign. The agency also flags services that do not make a licensed doctor available for questions after medication arrives. Those warnings do not mean that telehealth prescribing is inherently unsafe. They describe safeguards consumers should expect whether care happens online or in a clinic^[3].

The compounded-drug issue is separate

The Yale report says most sampled companies sold compounded GLP-1 products. Compounding and clinical oversight are related but distinct issues. A platform could conduct a careful assessment and still offer a compounded product, or prescribe an FDA-approved product after an inadequate assessment. The study's prescribing result should not be treated as a laboratory analysis of what was shipped^[2].

FDA states that compounded drugs are not FDA approved. The agency does not review them for safety and effectiveness or for quality before marketing. Its current GLP-1 page reports concerns involving shipping storage and fraudulent labels. It separately discusses dosing errors and adverse-event reports. That record supports caution about unapproved products, but it does not prove that every compounded prescription in the Yale sample was defective^[3]. Readers looking for molecule-specific background can use our semaglutide evidence overview and tirzepatide evidence overview.

What we do not yet know

This was an observational audit of a selected set of websites using one fictional, treatment-eligible profile. The results cannot estimate experiences across different ages and health histories. Insurance status and contraindications could also change the pathway. We do not know how often a clinician reviewed a form carefully without contacting the shopper, or how care changed after delivery. The study cannot compare telehealth with in-person prescribing on clinical outcomes^[1][2].

A stronger next step would sample more platforms and patient profiles. Researchers could preserve the questions asked and follow what happens after approval. Follow-up access and product identity would be useful outcomes; so would discontinuation and adverse events. Those data could show whether a thin intake process translates into measurable harm. For now, the conclusion is limited: in this 49-site sample, prescriptions and shipments were common, while direct clinician interaction was often missing.

What readers should take from the study

Convenience is not the problem by itself. The concern is a process that treats a prescription as a checkout step instead of a clinical decision. Someone considering a GLP-1 medicine should expect a licensed prescriber to review relevant history and explain material risks and benefits. There should also be time for questions, plus a clear route for follow-up. A qualified healthcare professional who knows the person's history is the right person to assess whether a medicine is appropriate.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.