DermGPT and the Expanding Role of AI in Clinical Practice

Artificial intelligence (AI) is no longer a distant concept in dermatology—it is already embedded in daily workflows. Yet how clinicians are using it varies widely. In an interview with Dermatology Times, Faranak Kamangar, MD, a dermatologist with Sutter Health in Palo Alto, California, outlined how platforms such as DermGPT are pushing the field beyond basic information retrieval toward real clinical utility.

“I would say the large majority of clinicians are still using AI as a search bot,” Kamangar said. “They’re kind of using it like Google.”

That gap—between what AI can do and how it is currently used—defines this moment in the field. While many clinicians remain in the exploration phase, others are beginning to integrate AI more deeply into the mechanics of care delivery, from documentation to patient communication and administrative workflows.¹

MORE ON AI IN DERMATOLOGY

One of the most immediate impacts is on efficiency. Dermatology clinics are fast-paced, with complex cases often squeezed between routine visits.² AI offers a way to manage that load in real time rather than deferring it. “You want to make that complex task less complex in your head, so you do it right then and there,” Kamangar explained. “Maybe write the note right then and there, create the patient handout so you can give it to the patient.”

This shift toward closing the loop during the visit has ripple effects. Incomplete notes can lead to prior authorization denials, therapy delays, and time-consuming follow-up. By integrating AI at the point of care, clinicians can avoid what Kamangar described as “the forever-calling insurance world”—a familiar frustration in modern practice.

AI’s reach is also extending beyond physicians. In some cases, adoption has been driven organically by clinical staff. Kamangar described an unexpected moment when clinicians revealed they had already been using AI for triage. “We had a group of nurses from a local academic center reach out to us, saying, ‘The system’s down. We use this for our triage,’” she said. “We didn’t realize that use case was being done, so we actually created an agent specifically for that.” These kinds of bottom-up use cases highlight how AI is quietly reshaping workflows across the care team.

Patient communication is another area undergoing rapid change. Traditional handouts, often generic and easily discarded, have long been a weak point in patient education. AI enables clinicians to generate highly specific, patient-centered materials in seconds. For acne management, for example, instructions can be tailored to include multiple topical agents, sequencing, adverse event management, and supportive skin care recommendations.

“You’re able to create handouts that are really unique…you can do that in about a minute,” Kamangar said.

The benefits extend beyond education. By proactively answering common questions, these tools can reduce message volume and improve adherence. When patients do reach out, AI-assisted responses can dramatically cut down physician time. “A 2- to 3-minute task can go down to 30 seconds,” she noted—an incremental gain that becomes significant across hours of patient messages.

AI is also emerging as a cognitive support tool in complex cases. Dermatologists can use it to sanity-check differential diagnoses or consider additional workup in real time. “I have a patient presenting with this, this, and this. I’m thinking these 2 things. Am I correct, or what else am I missing?” Kamangar said, describing a common prompt. Used this way, AI acts less as an authority and more as a sounding board, helping clinicians refine their thinking without replacing it.

Still, Kamangar emphasized that enthusiasm must be tempered with caution. Patient privacy remains a key concern, particularly given the way large language models process information. “Don’t include patient information,” she advised. “Even on sites that say ‘HIPAA compliant,’ I think AI is a different level.”

Accuracy is another consideration. AI tools can generate confident, well-structured responses even in the face of uncertainty, making clinician oversight essential. Rather than autonomous decision-making, Kamangar envisions AI functioning within defined boundaries, similar to other members of the care team operating under protocol.

Ultimately, the promise of AI in dermatology is not about replacing clinical expertise, but amplifying it. For most clinicians, the goal is not to practice differently but to practice more efficiently, with fewer bottlenecks and better patient engagement. As tools such as DermGPT continue to evolve, that mindset shift—from curiosity to integration—may define the next phase of AI in dermatology.

References

1. Maleki Varnosfaderani S, Forouzanfar M. The role of AI in hospitals and clinics: transforming healthcare in the 21st century. Bioengineering (Basel). 2024;11(4):337. doi:10.3390/bioengineering11040337
2. Katta R, Strouphauer E, Ibraheim MK, Li-Wang J, Dao H. Practice efficiency in dermatology: enhancing quality of care and physician well-being. Cureus. 2023;15(5):e39195. doi:10.7759/cureus.39195