Tips for AI and Health Information Technology in Dermatology Practice, With Jane M. Grant-Kels, MD

A presentation titled ‘Making the Most of Health Information Technology: Tools for the Dermatology Practice’ was given at the 2026 American Academy of Dermatology (AAD) Annual Meeting in Denver, Colorado.1

Jane M. Grant-Kels, MD, was the session presenter, and she was interviewed with HCPLive about key takeaways on the topics highlighted in ths talk. Grant-Kels is known for her role as the vice chair of the Department of Dermatology, the founding director of the Cutaneous Oncology Center and Melanoma Programs, and as a professor of dermatology, pathology, and pediatrics at the University of Connecticut Health Center.

This HCPLive interview Q&A features Grant-Kels discussing notable highlights from her talk at AAD 2026:

HCPLive: We want to ask about your other AAD session on making the most of health information technology tools for dermatology practice. Would you tell us about some of the tools or practices that you highlight within that session?

Grant-Kels: A lot of the new tools involve artificial intelligence. Artificial Intelligence is now being used to help streamline paperwork, fill out forms, communicate with patients on the website, and is also now being used for scribe work. So very often, you'll go into a room with a new doctor or a doctor, and they'll say, ‘Do you mind if I record this session?’ Because the recording is what's required for the AI to create a note out of the conversation that you're having…Those are great things. I'm a big advocate for doctors not being afraid of AI, but embracing it and making their practices more efficient. From the patient's perspective, there are concerns, though. First, a lot of patients are afraid of artificial intelligence and that you need to get their informed consent for you to use it, particularly for an AI scribe.

Then the other big issue is privacy, because how that data is saved is complex, and as we all know, there is no such thing as 100% secure data anymore, and AI only makes that more of a difficulty, because that data, especially a visit with the doctor, where all the information is transcribed, is no longer as secure as it was when we had written records. The electronic medical record, per se, makes data available to others, but it can also be broken into. So those are the main, main issues. The other issue is access to something called distributive justice. AI is very expensive and to include and who is going to have the ability to have that part of their practice, but are going to be physicians who are practicing in more affluent areas, rather than physicians who are taking care of Medicaid patients or rural doctors who may not have access to a higher population that can pay more, so that you know the availability of all this new technology is also not going to be distributed fairly because of the cost involved.

Those are the major issues. Informed consent, autonomy, the patient's right to allow or not allow AI to be part of the visit, and distributive justice. Who's going to have access to all of these new technologies that are very expensive? Those are the major issues. Then, there's also the concern that AI is going to replace healthcare providers in some form or fashion, whether it's the physicians themselves, whether you will need fewer physicians, or it's the secretary at the front desk or the administrator who does the paperwork.

The question is, how do we embrace AI incorporated into our practices and ensure that people have job training so that they know how to use the AI and can be employed in some fashion, to assist in the healthcare system? So those are the major concerns that I'm going to bring up about the employment of the new technologies…For anybody who's called a number like you call a department store or business, and you're never talking to a human being anymore, that's beginning to happen in medicine now. You call on someone's office, and you're not talking to an answering service. You're not talking to their front desk service.

Very often, you're talking to an AI that sounds very human-like and may be answering your questions. And you know, there's a concern about accuracy, if they're deliberating with the patient, and even in some cases, giving advice to the patient. A lot of these AI services, where the patient has access to their electronic medical record and can ask questions, in many cases. Now, the initial questions may be handled by AI, and there's some concern about how accurate that is as well.

HCPLive: Have you noticed in particular that there is one form or another of this technology that is more widely implemented in your own practice or any other notable products you highlight in your session?

Grant-Kels: There's been an explosion of products that are available. I mean, all you have to do is type in AI and medical, you know, office care, and it explodes with different companies that you know, there are many, many companies, and I applaud them trying to streamline prior authorization. So when you come to me as a patient, and I write a prescription, the insurance company and their wisdom turns it down, and then you have to fill out a prior authorization to rationalize why you want that drug.

This is very time-consuming for offices, and so many systems now offer AI to prepare the paperwork. Of course, at the same time, the insurer is also employing AI to reject the new prior authorization that you've submitted. But there's been an explosion. The main electronic health record that a lot of people are using in hospitals is, obviously, Epic. Epic is now employing AI in some of its technology. So that's what I'm exposed to, mostly at my institution. But in the private sector, people can use whatever they want.

HCPLive: Do you feel there is a concern, maybe among medical residents, about AI scribes and AI note takers removing the ability to learn certain things as thoroughly?

Grant-Kels: A lot of people, historically, have worked as scribes before medical school. That's when they finished college. They're either taking some pre-med, and they're working part-time, ascribed, learning the lingo, learning the medical, electronic medical record, and those jobs are definitely at risk, and I think a lot of people who have enhanced people's education and prepared them better for medical school. There is that concern. I think, from my experience, I find residents and medical students worried about what their role will be and whether they're going to be replaced with AI, for example, in radiology. Most X-rays are read by AI and then reviewed by a radiologist in pathology.

Now, a lot of slides are reviewed by AI and then checked by a pathologist…Dermoscopic images can now be evaluated by AI. I teach courses on demoscopy, and I'm also a dermatopathologist, and residents ask, ‘Do I have to learn this if AI is going to do it for me?’ The answer is yes, because AI needs the human as well as the AI together to be a partner, because it sometimes makes mistakes, just like we as humans make mistakes. The other problem with a lot of the AI is [one that is] in dermatology. A lot of the training sets have not included all skin types, and so there's a big concern whether the AI is more prone to make mistakes with darker skin patients, because it hasn't been trained properly.

So there are a lot of concerns about the accuracy of AI, the privacy, the informed consent, the autonomy of the patient, and the role that we as physicians will have with AI as it gets better, because AI tends to get better and better. So you know, there's this always and it's learning from its mistakes. So the question is, will it someday? Will there not be a need for dermatopathologists? Will there not be a need for demoscapists? Will there not be a need for radiologists? And my answer is yes, there will be a need. It will enhance our job by circling an area with its notes on pathology. But that doesn't mean we don't look at the slide, or that we don't look at the X-ray, but there is that concern among younger physicians today.

HCPLive: Can you envision a world in which these new technologies enhance the jobs of some of these clinicians, so they will actually be better able to tailor things specifically toward patients a little bit more, rather than the current system?

Grant-Kels: Physicians are in a very bad situation. Our overhead is going up, just like everybody else's, and our reimbursements have gone down. And so to keep a practice viable, physicians have been forced to see more patients. It's, believe me, doctors don't enjoy going at that speed. I do think AI might help, because it might certainly, in theory, if AI writes the note, that's more time I can look at the patient rather than staring at the computer and typing. I mean, that's the goal. If AI can synthesize and give me a synopsis of what previously happened at previous visits and give me a summary paragraph, I can go into a room more alert to what the patient's concerns are.

I mean, that is possible. But there's a limit to how fast doctors can go, because there's still that human need to come into contact with a patient, and put hands on the patient in regards to reassurance and explaining things to the patient. I do think that for AI, we need to partner with it. I also think that the system needs to change. Reimbursements need to rise with overhead, so that needs to be addressed as well, especially since AI is very expensive. These technologies are not cheap. They come at a great cost to the physician's offices. So there needs to be a compromise, because you can't keep going faster and faster and still get any.

The quotes contained in this discussion were edited for the purposes of clarity.

References

1. Grant-Kels J, et al. F062 Making the Most of Health Information Technology: Tools for the Dermatology Practice. Sesson presented at: 2026 American Academy of Dermatology Annual Meeting; March 27–31, 2026; Denver, CO.