On July 6, 2026, JAMA published a study that had been in progress for over a year: a controlled test of what actually happens when a patient tries to obtain a GLP-1 prescription through telehealth, conducted by a research team including a Yale medical student who personally went through the process at 49 different platforms.[1] The study's core question was narrow and testable — does a prescription require meaningful clinical interaction — and its answer was unambiguous enough to generate coverage across STAT, the Washington Post, and PhillyVoice within 48 hours of publication.[2][3]
Study design
The lead researcher, posing as a 237-pound patient with no disqualifying medical history, went through the standard consumer signup flow at 49 telehealth websites that advertise semaglutide or tirzepatide for weight loss.[4] This was not a survey of company policies or a review of marketing claims — it was a direct test of the actual patient-facing process, the same one any real applicant would encounter.
The study tracked several variables at each site:
- Whether a prescription was ultimately issued
- The modality of clinical interaction, if any — synchronous video, phone call, asynchronous messaging, or none
- Whether the platform provided any mechanism to contact a provider with questions, either before or after prescribing
Findings
Of the 49 sites tested, 45 issued a prescription. Roughly two-thirds of those prescriptions were issued without the applicant interacting with a clinician in real time — the process was entirely asynchronous, typically an online questionnaire reviewed (or auto-approved) without a live exchange. Thirteen sites used video consultation as part of the process; three used a phone call.[4] Separately, approximately 8% of sites tested provided no mechanism at all to contact a clinician, either before the prescription was issued or after, for questions about side effects or dosing.[5]
Two-thirds of the sites that prescribed a GLP-1 medication did so with zero interaction between the applicant and a clinician.
What the study does not establish
A methodologically careful reading requires separating what this study demonstrated from what it's often summarized as showing in secondary coverage.
It does not measure patient outcomes. The study assessed process — whether clinical interaction occurred — not whether patients on any of these platforms experienced worse health outcomes than patients on platforms with synchronous review. That's a separate, unanswered research question.
It tested one applicant profile. A single simulated patient with no disqualifying history was tested across all 49 sites. Whether these same platforms would appropriately screen out a genuinely disqualified applicant — history of pancreatitis, thyroid cancer, pregnancy — was not directly tested by this methodology.
The 49 sites are not individually named in the published study. Secondary coverage and platforms themselves have not published which specific companies were tested, which limits the ability to hold any single named platform accountable based on this data alone.
Why this design still matters
Despite these limits, testing the actual patient-facing process — rather than surveying advertised policies — is a meaningfully stronger methodology than most consumer-facing telehealth research to date. Self-reported compliance and marketing claims are exactly what a secret-shopper design is built to bypass.
How this connects to the broader research picture
The study's findings sit alongside a separate May 2026 Journal of Medical Internet Research analysis on the clinical support gap in telehealth GLP-1 care, which found that patients in reviewed trials lost meaningful lean muscle mass alongside fat, with limited registered dietitian involvement across the studies reviewed.[6] Read together, the two publications point toward the same structural issue from different angles: the prescribing step in telehealth GLP-1 care has scaled faster than the ongoing clinical support infrastructure around it.
Policy context
The study's publication came roughly one week after the July 1, 2026 launch of the Medicare GLP-1 Bridge demonstration program, and amid an active DEA rulemaking process on permanent telemedicine controlled-substance prescribing standards — meaning its findings entered a policy environment already actively reconsidering telehealth prescribing rules, rather than a static regulatory backdrop.[7]
Compare current GLP-1 telehealth providers
Source GLP-1 tracks pricing, licensing, and provider structure across the market.