PracticalDermatology

DermGPT, a new AI (artificial intelligence)-powered tool for improving clinic productivity, was recently unveiled at a limited press briefing. The first-ever dermatology chatbot is powered by proven, curated AI. Developed by Faranak “Fara” Kamangar, MD, a board-certified dermatologist based in Palo Alto, CA, DermGPT aims to improve clinical efficiency. Dr. Kamangar leads PSO Telehealth, focusing on psoriasis; she also practices at Sutter Health in the Bay Area and is the department chair with Palo Alto Medical Foundation.

“DermGPT allows dermatologists to ask questions in natural language and receive relevant, helpful answers, just like they would with real-life colleague,” Dr. Kamangar said. “With the growing volume of medical information, it can be difficult to obtain the necessary information in a timely manner.”

An engineer by training prior to becoming a physician, Dr. Kamangar built DermGPT to be unique in the world of large language models because of its proprietary database. “Instead of someone using ChatGPT where the information may not be scientifically sound or sourced from peer-reviewed medical journals, DermGPT only pulls from trusted, physician-approved sources,” she explained.

Dr. Kamangar hopes DermGPT serves dermatologists as a timesaving enterprise chatbot, an AI-powered dermatological colleague trained with chat, reasoning, writing abilities, and customized with a highly expert-curated dermatology-specific database.

“Every industry requires knowledge assistants to apply across business functions and get jobs done faster,” Dr. Kamangar said. “Healthcare is at the top of this list, with increasing clinical demands. Dermatologists are especially impacted, with most of us seeing 30-plus patients a day, many of whom require specialty medications, prior insurance authorizations. Minutes saved in a day because of DermGPT will yield more efficient patient care, decreased wait times to see a dermatologist, and less physician burnout.” Practical Dermatology^© spoke with Dr. Kamangar about her app.

What propelled you to create your AI platform?

Dr. Kamangar: As our clinical demands increase and physicians become exceedingly overburdened, AI assistance will be a game changer for dermatologists. My main driver was to improve patient care and streamline dermatology practices by assisting with our ever-increasing clinical tasks and making dermatology information ready at our fingertips. DermGPT can leverage the power of machine learning to enhance accuracy, efficiency, and accessibility in dermatology. I am motivated to create technology that is physician-first, because I have seen how bad technology has led to further burnout, instead of helping physicians.

Can you describe the steps a doc would take to use the platform with a patient?

Dr. Kamangar: There are multiple use cases for a dermatology chatbot that is powered with vast dermatology peer-reviewed knowledge. A physician might use DermGPT to look up dermatology information including diagnoses, treatments, and prior authorization criteria. They can also use DermGPT for patient management and shared decision-making tools such as patient educational handouts and prescription programs. A dermatologist may also find it helpful to answer patient inquiries such as health online messaging and direct messaging. Dermatology providers and staff--RNs, MAs, and ancillary staff--can all utilize this trusted source of information to help serve a patient.

How do you envision it being used by dermatologists in everyday practice?

Dr. Kamangar: For my practice, my RNs, MAs, and I have DermGPT up during our entire clinical day. We use it to look up information quickly to answer clinical queries, look up data, answer patient questions, and access handouts.

What is the learning curve for using the platform?

Dr. Kamangar: The platform’s user interface is designed to be intuitive and user-friendly, with clear navigation and easy-to-understand instructions. This can help users quickly learn how to navigate the platform.

What are the pros and cons of the app?

Dr. Kamangar: It is also continuously updated and improved (unlike ChatGPT which hasn’t been connected to the Internet since September 2021) and will remain the most updated source of material. The con of any AI tool is that it is not a human, nor is it trying to be. AI tools excel in efficient and data-driven tasks. They lack human creativity, intuition, and the ability to think outside the box. So, like all AI tools, DermGPT is most powerful when paired with a dermatology provider or staff who can complement its skills with their higher-level creative problem-solving. One of the best pros of DermGPT is that it will save dermatologists time so that they have more energy to engage in more complex thinking and human relationships which then strengthens our ability to care for patients.

Is there an ideal scenario for its use?

Dr. Kamangar: In what situations should the platform not be used, if any? The ideal scenario for DermGPT is as an assistant to help save time and improve access to dermatologic data for a dermatologist and their staff. DermGPT does not replace medical care or decision-making, it should not be used as a stand-alone decision maker, but as an assistant.

How does your platform fit into the growing landscape of AI in dermatology?

Dr. Kamangar: Much of dermatology AI development is around visual AI tools to help diagnose from images. DermGPT is different and is the first language-based model. Language-based models are the solution for practice management. They aid in decision-making by providing easily accessible high-quality data, patient education, research, and data analysis. By integrating with existing EHR systems, DermGPT can streamline documentation and facilitate seamless information exchange.

DermGPT aims to assist with the following common tasks in a dermatological practice setting:

• Provide healthcare providers a quick reference for medication options, dosing, and interactions.
• Summarize diagnosis and treatment guidelines for complex cases.
• Suggest appropriate labs, imaging, and procedures for diagnostic workups.
• Automate documentation such as visit notes, orders, and prescriptions.
• Ask questions about diagnoses.
• Write insurance authorization letters.
• Create patient handouts regarding drugs, after-care, and more.
• Summarize a provider’s medical notes.
• Assist with study guides for Board Exams.
• Provide informed consent language regarding procedures and telehealth consults.