Every day, new articles are published highlighting applications emerging for Artificial Intelligence (AI) use in almost every aspect of our lives. While many of these writers present doomsday scenarios, addressing the many positive advances AI can achieve, particularly in healthcare, is essential.
In healthcare, AI has the potential to streamline administrative tasks, improve medical diagnosis, contribute to personalized medicine, and advance drug and medical technology. AI is a tool that should not be underestimated. Yes, there are concerns that AI may diminish human connection in medicine, and some argue that using ChatGPT to respond to patients will be robotic and impersonal. However, a recent study from JAMA (Journal of the American Medical Association) suggests the contrary.
In fact, in the study, AI was perceived to not only provide higher quality medical advice, but also to be more empathetic than human physicians. Harvard researchers selected medical questions from a Reddit social media platform, typically answered by verified health professionals. Both ChatGPT and real doctors responded to these questions. Another team of doctors, unaware of whether AI or a physician generated the answer, ranked responses on a 1-5 scale of “Good” to “Not good” advice and “Empathetic” to “Not Empathetic.”
Interestingly, physicians ranked fellow physicians as far less empathetic than ChatGPT. Empathy is “the ability to understand and share the feelings of another.” The study concluded that AI was perceived to understand patients’ feelings better than human doctors.
Why these results? Shouldn’t a human doctor have a greater capacity to understand feelings better than technology? ChatGPT can likely mimic empathy. When measuring empathy, it’s a response people have to certain cues which a computer can replicate. For the most part, the responses by Chat GPT were longer. The length makes it seem as though more time was taken to respond carefully and answer concerns. In addition, the responses by ChatGPT were likely written in layman’s terms, making them more relatable to those without medical expertise.
Looking at this study, one might think that doctors should be nervous about AI replacing them. After all, AI can spit out medical facts instantaneously and can also be perceived to relate to patients’ feelings better. However, AI should not be viewed as a replacement but rather as a companion to the doctor. AI can make doctors more efficient, allowing healthcare workers to spend less time on administrative tasks and more time actually connecting with patients.
Similar to how a Tesla doesn’t eliminate the need for a driver, AI doesn’t eliminate the need for a physician. Think of AI as the equivalent of a medical student who generates responses at the speed of light, but those responses need review. After all, AI may “hallucinate” and give incorrect facts and made-up sources. Physicians are needed for quality control and review. Even with this required review, AI can be an integral new tool to free up more time for physicians. Responsibly implementing AI can bring about medical advances and reduce fears about this new technology.
There are many ways in which AI may reduce the time for tasks as well as the work burden on physicians, which can help reduce burnout and improve patient safety. It can assist physicians in combing through patient medical records and documenting patient encounters, keeping healthcare workers off computer screens during patient visits. AI can also streamline the ability of physicians to respond to patients’ questions remotely. It can create an environment for healthcare workers to connect better with patients, lower costs, improve outcomes, and decrease burnout.
AI will not replace physicians, but it can be used to empower them to improve the healthcare experience for patients and themselves. Medicine has become more complex, and AI presents an opportunity to obtain a better handle on that complexity and benefit patients, physicians, and the whole healthcare system. Ultimately, we think embracing AI in medicine will reduce tedious tasks for physicians, improve patient safety, and result in more efficient and compassionate care.
________________________________________________________________________________________________
(Left) Dr. Christine Collins is an Emergency Medicine resident at Cooper University Hospital in Camden, NJ. Following in the footsteps of Dr. Mazzarelli (Dr. Mazz), she completed a one month internship on The Michael Smerconish Program.
(Right) Anthony Mazzarelli, MD, JD, MBE is Co-President ans CEO of Cooper University Health Care in Camden, NJ. and the Dean of Clinical affairs for Cooper Medical School of Rowan University. He is a physician, lawyer, bioethicist, and healthcare executive. He is still a practicing emergency medicine physician.
________________________________________________________________________________________________
Ayers JW, Poliak A, Dredze M, et al. Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum. JAMA Intern Med. 2023;183(6):589–596. doi:10.1001/jamainternmed.2023.1838
