Key Takeaways
- Nolla Health has launched a one‑year pilot in Utah that uses AI to assess mild‑to‑moderate acne and recommend prescription topical treatments without an in‑person doctor visit.
- Patients pay a $4.99 subscription, submit basic health data and facial photos; the AI maps skin lesions and suggests a regimen, with prescriptions sent to the user or a local pharmacy.
- Weekly tele‑check‑ins monitor progress; cases that do not improve are escalated to licensed clinicians for further evaluation.
- The pilot is overseen by Utah’s Office of AI Policy, which granted Nolla a special sandbox status to test the model while requiring monthly reports and involvement of the state medical licensing board.
- State officials see the program as a potential remedy for healthcare worker shortages, long wait times, and high costs, while clinicians hope AI can free dermatologists to focus on more serious conditions like melanoma.
- Initial safety controls require physician review of AI recommendations for the first 100 patients; oversight will relax after subsequent phases, contingent on outcome data.
- Nolla is exploring expansion to other indications such as urinary tract infections, influenza, radiology, and eventually birth‑control prescribing, pending proof of concept.
Patient Onboarding and Data Collection
Nolla Health’s pilot begins with a simple digital enrollment process. Interested individuals pay a $4.99 monthly subscription, after which they provide basic health information—including family history—and upload several facial photographs taken under varying lighting conditions. As Dr. Zaid Fadul, Nolla’s chief medical advisor, explained, “Patients who pay the $4.99 subscription and enroll provide basic health information, including family history, and upload photos of their face taken in different lighting.” These images feed the proprietary AI algorithm, which constructs a three‑dimensional map of the patient’s skin to identify lesion type, severity, and distribution.
AI‑Driven Treatment Recommendation
Once the skin map is generated, the AI system formulates a personalized treatment plan focused on topical acne medications. The recommendation is then transmitted either directly to the patient’s home or to a local pharmacy of their choice. Fadul noted that the approach is intentional: “We thought that was a good place to start because you can deal with it quite well with topical prescriptions.” By limiting the initial scope to topical agents, the startup aims to minimize risk while demonstrating the AI’s ability to assess dermatologic conditions accurately.
Monitoring and Escalation Pathways
Patients enrolled in the program participate in weekly follow‑up check‑ins via telehealth to track progress and side‑effects. If a patient responds well, they continue on the prescribed regimen; if the therapy fails to produce desired results, the case is escalated. As Fadul described, “Either patients are started and doing well and they keep moving along, or they’re started and not getting the results they want, and they get bumped up to one of our clinicians.” Licensed medical providers are available around the clock through the platform to review non‑topical therapies or complex cases that fall outside the AI’s current capabilities.
Regulatory Sandbox and State Oversight
Utah’s Office of AI Policy granted Nolla a special sandbox designation, allowing the company to operate under relaxed medical‑practice rules for a limited period. Office director Zach Boyd explained the rationale: “We can let them run a business model that is kind of different from the normal regulations for a little while and try to assess as a state whether we think it could be beneficial.” The oversight framework includes mandatory monthly reports detailing patient volume, clinical outcomes, and any adverse events. Importantly, members of Utah’s medical licensing board helped shape the pilot’s design, ensuring that state clinical standards remain a reference point.
Physician Review Phases and Safety Controls
To safeguard patient safety, the pilot incorporates a staged physician‑review process. During the first 100 enrolled patients, a licensed clinician examines each AI‑generated prescription before it is dispensed. For the subsequent approximately 500 patients, reviewers will assess recommendations after the AI has already issued the prescription, shifting oversight to a post‑hoc audit. Fadul emphasized the rigor behind this approach: “That first hundred patients is really critically important, in addition to the 10,000 plus cases that we tested ahead of time. This has already been tested out. This has already been thoroughly mapped out and rigorously processed, and this is just a chance to do it live.” After these phases, oversight will taper further contingent on demonstrated safety and efficacy.
Motivations Behind Choosing Utah
Fadul highlighted Utah’s healthcare landscape as a motivating factor for selecting the state as the pilot site. He pointed to systemic pressures that many patients experience: “People are dissatisfied with the healthcare system, and we have a shortage of workers. We have long wait times, expensive prices.” By delegating routine acne management to an AI‑driven platform, Nolla hopes to alleviate dermatologist workload, allowing specialists to concentrate on higher‑acuity conditions such as melanoma. Fadul posed a rhetorical question to underscore the trade‑off: “If you’ve got a melanoma problem, do you want your dermatologists who are busy seeing a lot of basic acne cases, or focusing more on potentially lifesaving intervention for melanoma?”
Future Expansion and Broader Applications
Looking ahead, Nolla envisions scaling the AI model beyond acne. Fadul indicated interest in expanding the service to adolescents aged 13 and older, pending data validation: “I think one day we may, if the data proves out like I think it will, it’s hopefully something that’ll be from 13 [years old] and up.” Additionally, the startup is exploring applications in radiology, infectious‑disease diagnostics (e.g., urinary tract infections and influenza), and eventually hormonal therapies such as birth control—though the latter remains under investigation. The company’s website and Fadul’s comments suggest a strategic roadmap that leverages early success in dermatology to build a broader AI‑enabled telehealth ecosystem.
Conclusion: Balancing Innovation with Patient Safety
Nolla Health’s Utah pilot represents a concrete experiment at the intersection of artificial intelligence, dermatology, and regulatory flexibility. By offering a low‑cost, subscription‑based pathway for acne treatment, the program aims to mitigate access barriers while generating real‑world evidence to inform broader adoption. State officials will monitor outcomes closely, and the staged physician‑review protocol seeks to ensure that innovation does not compromise safety. Whether the model succeeds in reducing wait times, lowering costs, and freeing clinicians for more complex care will hinge on the data collected over the coming year—a data set that, if favorable, could pave the way for AI‑first approaches across multiple medical specialties.
https://www.ksl.com/article/51633260/utah-pilot-lets-artificial-intelligence-prescribe-acne-treatments