As a physician, I’m captivated by artificial intelligence’s (AI) potential to revolutionize care, while simultaneously acknowledging the apprehensions that accompany such advancements — in particular the more topical large language models (LLMs) and generative AI.
Ready or not, there’s no question that AI will impact our industry; in fact, the market is expected to grow from $14.6 billion in 2023 to a staggering $102.7 billion in just five years.
An expanding array of use cases
In my mind, there are three general use cases for how AI can be best used in health care. Two seem to be discussed with the most frequency: administrative and clinical. In administrative use cases, AI can help reduce the burden on staff by managing patient data, scheduling patients, billing, and similar tasks. With staffing challenges, many hospitals and health systems are already applying AI in this way.
The second area, clinical AI, is likely the most commercially and academically developed. This involves using AI to support the diagnostic process and assist with clinical decision-making. There are a myriad of mature solutions in this field with reach into every field of medicine — from radiology to the emergency department. And this makes sense, doctors and clinical researchers have always looked for new tools to make them better at helping patients. Why should AI be any different?
The third, and least explored, use of AI in healthcare is to improve the quality and fidelity of the interaction between a doctor and his or her patient. This is something that was largely impossible prior to the advent of scalable generative AI and LLM solutions. Modern health care has a communication deficit that AI can most certainly help fill. We doctors want more time with our patients, and our patients want more time with us.
But time is in short supply. The trick therefore is to make the most of the time we have together.
As an emergency department (ED) physician, I’ve seen firsthand how AI can help patients understand their healthcare journey and prepare for the limited time they have with me. In my ED we use AI in a variety of ways to accomplish this goal. For example, we use it to predict treatment times so that our patients’ families who aren’t able to come to the ED know what to expect. This has dramatically affected the experience of receiving care as measured on standard surveys and reduced walk-outs, which has broad clinical and financial implications for hospitals.
The role of AI in patient experience and engagement
Health care is a service industry. And in almost every other service industry, companies have adopted robust AI software platforms to interact with customers and their staff. Imagine taking a multi-stop international flight to Europe without your trusty Delta app. Now imagine what that trip would be like with several delays and rescheduled legs. Without your app, a nightmare. People expect a certain level of digital interaction from the services they choose to use, and the digital consumer landscape in health care is woefully behind.
Poor patient experience does more than lose business for health systems; it impacts health outcomes, too. If patients leave your hospital feeling they didn’t get the attention they deserved, they may be less likely to follow your post-discharge instructions, use their medications properly, or set up a follow-up appointment.
Good communication is critical to improving patient experience. And the ED is a real proving ground for this, especially if yours is (cough, cough) overcrowded or understaffed. AI can relieve some of that burden while contributing to a positive patient experience.
Just as they use their cell phones to track Amazon deliveries, patients in AI-supported EDs can use a mobile app to know exactly how long they’ll have to wait, view test results explained in plain language, communicate with staff, and even view educational content curated to their specific condition in much the same way YouTube suggests content you may want to watch next. The overall purpose of these tools is to keep patients informed in a timely way so that their interactions with clinicians are as deep and fruitful as possible.
Modern health care suffers from an ailment of time. Doctors, weighed down by administrative tasks, are forced to mete out time with our patients and their families. Further, we spend so much of the visit catching patients and their families up that we can miss an opportunity to foster the human connection that is the capstone of a successful doctor and patient interaction. While it sounds counterintuitive, AI provides an opportunity to bring back the human element by giving us back our most valuable asset: time.
Meeting patient — and regulatory — expectations
The 21st Century Cures Act introduced a new and interesting dynamic to the doctor-patient relationship. It requires doctors to give patients access to all key information in their health records as soon as it’s available. Recent research shows patients overwhelmingly prefer access to information immediately, even if they must wait to talk to their doctor about what it means.
While well-intentioned, a major challenge with this provision is that there is nothing in the law encouraging doctors to help their patients understand this clinical information. So, even though patients can often access their lab results or clinical notes immediately, these notes may not make sense or even lead to misinterpretation. There is no reason we should expect patients to be able to understand our highly technical notes without our help, but the law is placing these notes and results in their hands often hours or days before we can speak to them about the content and implications.
I believe this is where AI, specifically generative AI based on LLMs, can bridge the communication gap. AI can distill complex results and define technical jargon from medical notes and summarize this information safely. In fact, research has demonstrated that AI-driven health literacy interventions can empower patients with relevant knowledge, improve adherence to treatment plans, and foster shared decision-making.
What I am talking about here is not a replacement for bedside interaction. It is a temporary measure meant to elevate a patient’s understanding of their medical condition so that when the doctor-patient visit does occur, both parties can pick up where they left off and have a collaborative and constructive visit.
By bridging the information gap, and with the immediacy many patients are coming to expect, generative AI can democratize healthcare knowledge and make it more accessible to patients, regardless of their level of health literacy.
Instead of a doctor taking time to explain every lab result, every note, and every diagnosis in a way the patient can fully grasp, AI can do that automatically. This allows the conversation between clinician and patient to start at a higher level. The patient is ready to ask important questions, making more use of the provider’s valuable but limited time, strengthening human moments, and improving hands-on care.
As AI marches on, so too should the healthcare industry
AI can be a helpful tool to enhance the doctor-patient relationship. Human empathy, compassion, and expertise will always be irreplaceable, and AI should complement these qualities to ensure the delivery of patient-centered care.
What our industry needs right now is ongoing research and collaboration between AI developers, health systems and their doctors, and regulatory bodies. These stakeholders and others must work together to ensure the responsible and ethical implementation of AI technologies.
Like many of my colleagues, I am optimistic about the role of AI in how we care for our patients. With thoughtfulness and care, I believe we can mold AI into one of the most powerful tools we have to care for our patients. To do so, we must consider the different roles that AI can play in the healing process and approach them individually with diligence and bravery.
About Dr. Justin Schrager
Dr. Justin Schrager is an Emergency Medicine physician, the co-founder and Chief Medical Officer of Vital Software, Inc. He actively practices medicine while leading the development of innovative solutions to improve patient care through technology, including the use of artificial intelligence (AI). He is committed to advancing the responsible use of AI in healthcare and is dedicated to improving healthcare delivery and outcomes for patients