Skip to main content
Back to Our Blog
Playbook
TelehealthDigital Health

10 Common Compliance Mistakes in Telemedicine and Digital Health

Camino Strategy Group, LLC

I spend most of my week inside telemedicine and digital health companies, and the same compliance mistakes come up over and over, whether the company is a two-person startup or a platform operating in 40 states. None of these mistakes come from bad intentions. They come from founders, marketers, and operators who are moving fast in an industry that spent years telling itself it was unregulated, and is now finding out that regulators were simply catching up.

That catch-up is happening right now. In 2026 alone, the California Attorney General settled corporate practice of medicine cases against Carbon Health and Aspen Dental, the FTC sued Hims & Hers, state medical boards started issuing notices about peptides, and a CBS News investigation found companies selling an unapproved drug while displaying LegitScript certification. Digital health is different from a traditional clinic because everything is visible: your website, your ads, your intake flow, your contracts, and your pricing are all public or discoverable, and they are all evidence of how your business actually runs.

Here are the ten mistakes I see most often, why they matter, and what is happening in the news that shows why they matter now.

1. Assuming the rules are the same in every state

The biggest mistake I see is treating the United States as one market. Telemedicine lets you reach patients in all 50 states, but each of those states has its own medical board, nursing board, pharmacy board, and attorney general, and they do not coordinate with each other. A model that is perfectly fine in Florida can be a problem in California, and a service that is legal in one state can be something a board in the next state has told its licensees not to touch.

Peptides are the clearest example this year. In May 2026, the Alabama Board of Medical Examiners issued a notice stating that physicians, PAs, and CRNPs may not prescribe, administer, dispense, or even recommend non-FDA-approved research-grade peptides, and that a patient consent form does not protect the clinician. In August, the South Carolina Board of Medical Examiners followed with its own notice warning that research-grade peptides may not be used in patient care, and reminding practitioners that every drug and compounding ingredient has to come from a facility permitted by the South Carolina Board of Pharmacy, including out-of-state suppliers. Mississippi boards issued similar warnings.

Meanwhile, Louisiana went the other direction. Its 2026 peptide law (SB 253) prohibits licensing boards from stopping prescribers from providing peptides that come from an FDA-registered 503B outsourcing facility or a 503A compounding pharmacy using API from an FDA-registered manufacturer. So in the same year, one state's board says do not recommend them at all, another says only from permitted pharmacies, and a third passes a law protecting access. If you run one peptide protocol nationally with one pharmacy, you are almost certainly out of step somewhere.

The same is true for collaboration ratios, ownership rules, controlled substance prescribing, informed consent, and what counts as a valid patient relationship. Your state map is not a one-time project; it is something you maintain.

2. Confusing corporate practice of medicine with supervision and collaboration

I hear some version of this almost every week: "We have a collaborating physician, so we're covered." Those are two completely different sets of rules answering two completely different questions.

Corporate practice of medicine (CPOM) is about ownership and control. It asks who owns the entity that provides medical care, and whether anyone who is not licensed is influencing clinical decisions. Supervision and collaboration rules are about scope of practice. They ask whether a PA or NP has the physician relationship their state requires to practice, how many clinicians one physician can oversee, and what has to be documented in a practice agreement.

You can have a perfectly executed collaborative practice agreement and still be violating CPOM because a lay company owns the clinical entity. You can also have a properly structured friendly PC and still be out of compliance because your NPs in a reduced-practice state have no collaborating physician on file. More than 30 states restrict the corporate practice of medicine, and in the last two years several of them, including California and Oregon, have tightened their rules. Each layer has to be built and maintained on its own.

3. Letting the MSO run the clinical side in practice

A lot of companies have the right paperwork and the wrong behavior. The friendly PC exists, the physician owner is on file, the management services agreement says all the right things, and then the CEO or the head of operations is the one hiring clinicians, setting their pay, writing protocols, and deciding which patients get which medication. The physician owner signs whatever is sent to them and is otherwise not involved.

The physician owner should be approving clinical hires, signing off on protocols, reviewing clinician performance, and making clinical policy decisions, and there should be a record that they did it: meeting notes, approvals, chart review logs. A medical director whose name is on the website but who has never reviewed a chart is not oversight, it is a liability.

To be clear, the physician owner does not have to do all of this personally. Many PC owners delegate day-to-day clinical oversight to a medical director, and your MSO team can still do the work of drafting protocols, sourcing candidates, and running the schedule. What matters is that the delegation is written down, the medical director is qualified to hold that role, and the PC (through its owner or its delegate) is actually looped in and approving clinical decisions before they take effect.

Here is what I look for when I review a structure:

  • Who approves clinical hires and terminations? It should be the PC owner or a medical director they have delegated that authority to in writing, not the CEO or the head of operations.
  • Who signs off on protocols and formulary changes, and is there a dated record of that approval?
  • Does the PC owner or medical director review clinician performance, chart audits, and adverse events, and is there evidence they did (meeting notes, review logs, sign-offs)?
  • Is the delegation from the PC owner to the medical director documented, and does it say what the PC owner still reserves for themselves?
  • Does the management services agreement give the MSO approval or veto rights over anything clinical, including staffing levels, visit length, or which patients get which treatment?
  • How are clinicians paid? Compensation tied to the number of prescriptions or approvals is a red flag.
  • Do your platform rules (visit time limits, auto-refills, default approvals) constrain clinical judgment, and who signed off on them?

If you cannot point to a document for most of these, the oversight probably is not happening the way your contracts say it is.

This is exactly what regulators are looking at. In June 2026, the California Attorney General sued Carbon Health and its co-founder, alleging the MSO had complete authority over hiring, firing, and compensating licensed clinicians, and that the MSO's board discussed clinician staffing and physician pay. The case settled within a week for $4.4 million against the company and $100,000 against the co-founder personally, with a permanent injunction against those MSA terms. Regulators named an individual founder, which should get every executive's attention.

The federal criminal side is looking at the same thing. When Done Global's founder and clinical president were convicted in November 2025 on controlled substance charges, prosecutors used California CPOM principles as evidence, pointing to platform rules that limited visit length, automated refills, and paid clinicians based on the number of patients who received prescriptions. The physician-owned PC on paper did not protect them.

4. Writing job descriptions that are outside the scope of a license

This one surprises people because a job description feels like an HR document, not a compliance document. But a job description is a written record of what your company expects a person to do, and if it describes work their license does not allow, it becomes evidence of unlicensed practice.

The examples I see most often are health coaches who are expected to answer dosing questions or adjust titration schedules, medical assistants who are expected to triage symptoms and decide whether a patient needs a visit, "care coordinators" or "patient advocates" doing clinical follow-up with no licensed clinician in the loop, and RNs being asked to do things that require a prescriber. In telemedicine, these roles often sit inside the MSO and report to operations, which makes it worse, because now an unlicensed person employed by a lay company is doing clinical work.

Titles are part of this too. I regularly see RNs or NPs given the title "medical director" when the state's rules for that role, or the specific service line (med spa, IV therapy, weight loss, ketamine), require a physician, or require specific training or experience the person does not have. A title does not create authority the license does not give. If someone is acting as your medical director, check that they actually qualify to be one in every state where they are overseeing care, and that the person who is supposed to be supervising them is really doing it. This is where it is so important to go back to each state's scope-of-practice rules and confirm what each role is actually allowed to do, rather than building roles around what the business needs and assuming the license will stretch to fit.

Before you post a role, ask two questions: does this person's license (or lack of one) allow every task listed, in every state they will touch? And is a licensed clinician at the PC supervising the clinical parts of the job? If the answer to either is no, rewrite the role before you hire for it.

5. Putting things in contracts that state law does not allow

Founders and investors want everything documented, which I understand. If you are funding a business, you want to know what happens if the physician owner leaves, dies, or stops cooperating. The problem is that some of what people want in writing is exactly what makes the structure unlawful, and putting it in a signed contract does not make it enforceable. It just creates a clean exhibit for a regulator or a plaintiff.

The most common example is the succession plan. Continuity agreements (also called stock transfer restriction or succession agreements) that let the MSO pick the replacement physician, approve or block share transfers, or force an equity transfer when the management agreement ends are what turn a friendly PC into a captive PC. In Art Center Holdings v. WCE CA Art, a California trial court found that a private equity-backed MSO used a continuity agreement to control the firing of the PC's physicians, which violated CPOM. On appeal, the California Attorney General filed an amicus brief in March 2026 urging an even stricter reading, and the case was still pending at last report. The AG's May 2026 Aspen Dental settlement also included terms addressing owner-replacement rights, and the Carbon Health judgment prohibits MSO ownership interests in the PC through assignable options.

It is not only succession. I regularly see non-competes in states that restrict them for physicians, MSO fees calculated as a percentage of clinical revenue in states with fee-splitting rules, and MSO approval rights over clinical decisions buried in a schedule of services. Oregon's 2025 law (SB 951) now restricts several of these terms directly. A well-built structure gives investors protection through the management agreement and the economics, not through control of the physician's ownership.

6. Running asynchronous care where the state requires a real-time visit

Async is efficient, scalable, and patients like it, which is why so many companies build their entire model around a questionnaire and a chart review. The problem is that states do not agree on whether a questionnaire can establish a valid patient relationship. Some states, like California, allow asynchronous evaluations if they meet the standard of care. Others, like Texas, have historically required a synchronous audio-video visit to establish the relationship before prescribing. Many states say an online questionnaire alone is never enough, and some require periodic in-person follow-up for certain medications.

This is getting attention outside of compliance circles too. An AP analysis published in September 2026 of nearly 50 telehealth companies selling GLP-1s found that fewer than a third required any real-time video or audio visit, and some approved prescriptions within minutes. When a model looks like that from the outside, regulators start asking whether a real clinical evaluation is happening at all.

Controlled substances add another layer. The DEA's pandemic-era flexibilities, which let practitioners prescribe controlled substances by telemedicine without a prior in-person exam, are currently extended only through December 31, 2026. The DEA sent a final special registration rule to the White House for review in August 2026, but the text has not been published as of this writing. If your model includes testosterone, ketamine, stimulants, or any other scheduled drug, you need a plan for January, and you need to check state law, which is often stricter than the federal rule regardless of what happens with the DEA.

7. Running tracking pixels on your website and intake flow

This is probably the most common problem I find, and the one founders are least aware of, because the person who installed the Meta Pixel or Google tag was a marketer doing their job, not someone thinking about health data. If your pixel fires on your intake form, your checkout page, or a page that names a condition or medication, you may be sending health information to an advertising platform every time a patient clicks.

A lot of companies assume this is only a HIPAA issue, and since they are not sure HIPAA applies to them, they assume they are fine. The FTC does not need HIPAA. It has used its general authority over deceptive practices and its Health Breach Notification Rule to bring cases against GoodRx ($1.5 million) and BetterHelp ($7.8 million) in 2023, and Cerebral ($7 million) in 2024, after the FTC said Cerebral's tracking tools shared the information of nearly 3.2 million people. Those orders also banned the companies from using health data for most advertising going forward, which for a DTC brand is a business problem, not just a fine.

In July 2026, the FTC, joined by California and Utah, sued Hims & Hers, alleging the company used Meta Pixel and Meta's Conversions API to send information about what users did on its website to advertising platforms, despite marketing its service as private. On November 10, five more states joined the case. Hims denies the allegations and the case is ongoing, but the theory is the same one the FTC has been using for years: what you promise in your privacy policy and your ads has to match what your code actually does. On top of the FTC, states including Washington, California, Connecticut, and Maryland have passed consumer health data laws, and Washington's My Health My Data Act lets consumers sue directly.

Audit every tag on your site, especially on intake and checkout pages, and make sure your marketing team and your compliance team are looking at the same list.

8. Marketing like a consumer brand instead of a healthcare company

If I had to pick the area where companies get the most wrong, it is marketing. Digital health companies are usually built by people who are very good at it, and the playbook that works for a supplement brand, a fitness app, or a DTC skincare line does not translate to prescription care. In most industries, there are relatively few guardrails on how you acquire customers and what you can say to them. In healthcare, you are advertising services delivered by licensed professionals and, often, prescription drugs, which means the FTC, the FDA, state medical and pharmacy boards, state attorneys general, and the ad platforms themselves all have a say in what you put out. A claim that is just aggressive copy for a consumer brand can be a board complaint, a warning letter, or a lawsuit for a healthcare company.

The marketing mistakes I see most often:

  • Fake, purchased, or edited testimonials. Reviews written by the team, testimonials from people who were never patients, AI-generated reviews, or real reviews edited to sound better. The FTC's rule on consumer reviews and testimonials, in effect since October 2024, bans fake reviews and testimonials, including AI-generated ones, prohibits buying reviews, and lets the FTC seek civil penalties of more than $50,000 per violation against knowing violators.
  • Undisclosed influencer and insider endorsements. Paid creators, affiliates, employees, and clinicians promoting the program without clearly disclosing the relationship.
  • Outcome guarantees and atypical results. "Lose 20% of your body weight" or a dramatic before-and-after with no context about what a typical patient experiences.
  • Implying a compounded drug is the same as a brand. Calling it "generic," saying it has "the same active ingredient," or using the brand name in a way that suggests equivalence.
  • Branded vials and product imagery. Covered in the next section, but it belongs on this list.
  • Clinical claims with nothing behind them. "Clinically proven," "doctor-formulated," or "backed by science" without substantiation.
  • Clinician titles and credentials. An NP or PA calling themselves "Dr." in a patient-facing ad, or a clinician described with credentials or specialties they do not hold, is a state law problem in many places.
  • Pricing and subscription terms that are not clear. When a patient is charged, when refills bill, and how to cancel all need to be obvious. The FTC's Hims & Hers complaint includes billing and cancellation allegations alongside the privacy claims.

The FDA has been very clear about this in the last year. Since its crackdown on direct-to-consumer drug advertising began in September 2025, the agency has sent roughly 80 warning letters and 40 untitled letters in September 2025, 30 more in March 2026, and 25 more in June 2026, most aimed at telehealth companies marketing compounded GLP-1s. The letters target the same patterns over and over: calling a compounded product "generic," saying it has "the same ingredients" as a brand, or implying it was clinically tested.

Every piece of marketing should be reviewed by someone who understands healthcare rules before it goes live, and the clinical team (through the PC) should sign off on anything that makes a clinical claim. Marketing is often where the MSO's influence over clinical care shows up most clearly, so it matters for corporate practice of medicine as much as it does for advertising law.

9. Branded vials and imagery on compounded medications

I understand why companies want their logo on the vial. It looks premium, it builds the brand, and every competitor is doing it. But when a telehealth company puts its own name or trademark on a compounded drug, it creates the impression that the company made the drug, or that it is a manufactured product the company owns, when it was actually prepared by a pharmacy for an individual patient.

The FDA called this out specifically. In the March 2026 round of warning letters, it cited telehealth companies for advertising compounded products branded with the telehealth firm's name or trademark without qualification, which obscures who actually compounded the medication. Depending on the state, putting your trademark on a compounded product can also raise issues under the pharmacy practice act, because it can look like the company repackaged or prepared the drug without a pharmacy license.

Branding also puts you in front of drug manufacturers. Eli Lilly and Novo Nordisk have been suing telehealth companies over compounded GLP-1s for two years, and Lilly's complaints against Mochi Health and Fella Health included false advertising and corporate practice of medicine claims alongside the compounding allegations. Glossy branded vials in your ads are exactly the kind of screenshot that ends up as an exhibit. Brand your experience, your packaging, and your care, and let the pharmacy label be the pharmacy label.

10. Not understanding vendor due diligence, and treating LegitScript as a hall pass

Large companies outsource vendor due diligence because it is a compliance function, and it is one most startups skip entirely. I see companies choose an EMR, a pharmacy, a lab, a white-label clinical group, or a marketing agency because a friend used them, without asking whether the vendor will sign a BAA, what security certifications they hold, which states they are licensed in, how they handle adverse events, or whether they have had their own enforcement history. When a vendor fails, regulators and patients look at the company whose name is on the website, not the vendor.

LegitScript is a good illustration of how this works from the other side. LegitScript is a private company that performs due diligence for payment processors and advertising platforms, which is why you need it to run ads on Meta and Google or process payments for certain products. Some founders treat certification as proof that their entire business is compliant, and as a pass on the marketing rules once they have it. It is neither. Certification is a point-in-time review that you have to maintain, and it does not replace FTC rules, state advertising rules, or board requirements.

Certification can also be lost. Zealthy lost its LegitScript certification in 2025 after failing to disclose DOJ litigation. And in September 2026, a CBS News investigation found that several companies selling retatrutide, an experimental drug that is not FDA-approved, were displaying LegitScript certification. That story raised questions about the certification process itself, but for operators the lesson is simpler: the seal does not protect you if what you are selling or saying is not lawful, and the companies in that story are now the example everyone points to.

Why this matters now

Every one of these mistakes has the same root: the business moved faster than its compliance infrastructure. That was survivable a few years ago, when regulators were still figuring out how to approach telemedicine. It is much less survivable now that state attorneys general, the FTC, the FDA, the DOJ, drug manufacturers, and medical boards are all actively looking, and when the evidence they use is your own website, ads, contracts, and job postings.

The good news is that none of this is hard to fix if you catch it early. A structure built correctly from the start, with a physician owner who is actually involved, contracts that fit state law, a state-by-state map you keep current, and marketing that matches what your code and your care actually do, costs far less than unwinding a problem after a regulator or a competitor finds it.

How I help

I help telemedicine and digital health companies build, implement, and operate compliant structures, from PC/MSO formation and contracts to marketing review, vendor due diligence, and multi-state expansion. If any of these ten sounded familiar, reach out at caminostrategygroup.com and we can walk through where your business stands.

Related reading

This post is general information and is not legal advice. Laws and enforcement priorities change quickly, so talk with qualified counsel about your specific situation.

Sources

Have a compliance question?

Reading about it is step one. Send it our way.

If this article hits close to home — your structure, your states, your launch — tell us what you’re building. We read every submission and reply by email within one business day.

No sales calls. Educational content only — not legal advice.