- Executive summary
- Headaches in the healthcare industry
- The changing role of doctors
- The AI-assisted doctor will see you now
- Making medicine better for everyone
- How do doctors and patients feel about generative AI?
- Should doctors be worried about their jobs?
- Learnings
Executive summary:
Generative AI has generated a lot of excitement in recent years. However, there are few sectors where its impact could be so transformative as in healthcare, where it has the potential to save countless lives. A growing number of hospitals are now implementing generative AI to help doctors streamline administrative tasks, improve diagnostic accuracy, and provide a more personalised patient experience.
- Companies like Tortus and Nextech have developed AI-powered assistants for doctors that can record their conversations with patients and automatically create visit summaries.
- According to a February 2024 survey by Athinahealth, 83% of doctors believe that AI can help alleviate their administrative burden.
- PYMNTS Intelligence research indicates that the market for generative AI in healthcare could reach nearly US$22 billion by 2032.
- “What this technology does is it allows me to focus on the person in front of me—the most important person, the patient,” says Dr Shiv Rao, founder and chief executive of healthcare startup Abridge.
- “I don’t believe that AI will replace human physicians, but those who use AI will be so much more capable than those who don’t,” says Dr Bernard Chang, dean for medical education at Harvard Medical School.
It appears likely that generative AI will play an important role in the future of healthcare. However, this transition raises several challenging questions. How will the role of doctors change? Will we still be facing shortages of qualified professionals? And how might generative AI impact the patient experience? No definitive answers exist as of yet, so we should be careful not to get too carried away on our journey into healthcare’s world of tomorrow.
Headaches in the healthcare industry
Doctors have one of the most difficult and stressful jobs in the modern world. Between time constraints, information overload and repetitive administrative tasks, there is no shortage of obstacles standing between them and their ability to provide optimal care. Unable to give each patient the attention they deserve, doctors frequently misdiagnose them and subject them to unnecessary and often costly tests. This, in turn, can lead to longer wait times and inflict severe stress on the patients—often for nothing.
Generative AI could prove a game-changer for healthcare in this regard. Doctors could stand to benefit the most, helping them automate administrative work such as updating patient records and sending written correspondence. It could also help uncover relevant information about the patient’s health more quickly, enabling doctors to make more accurate diagnoses and devise more personalised treatments for their patients. On a wider scale, it could dramatically alter how we provide and receive medical care. So, let’s take a closer look at how the healthcare industry is being shaken up by this emerging technology, and consider its implications for the role of the doctor going forward.
The changing role of doctors
Generative AI can help doctors automate various administrative tasks, make more accurate diagnoses, and provide a more personalised patient experience.
Over the course of a typical workday, a doctor is expected to carry out a wide range of routine and repetitive administrative tasks, such as scheduling appointments, managing and updating patient records, writing prescriptions, and interpreting test results. While these tasks are important, they take valuable time away from providing actual care to patients and can force doctors into autopilot. Generative AI can help automate many of these tasks, reducing the administrative burden on doctors and freeing them up to focus more on their patients.
Generative AI can also help doctors arrive at the right diagnosis sooner. It can analyse patient records to identify key pieces of information the doctor might have missed and suggest alternative diagnoses or treatments that may be more suitable for the patient’s condition and symptoms. It can also be used to analyse medical images, including X-rays, MRI, and CT scans, to uncover hidden patterns that may not be immediately apparent to humans and diagnose diseases with greater accuracy and at a much earlier stage, significantly increasing the patient’s chances of recovery.
However, the use of generative AI in healthcare settings raises some serious ethical concerns. To be even remotely effective, generative AI models need to be trained on massive datasets. While healthcare providers do possess large amounts of personal data about their patients, this data is highly sensitive and has strict rules that govern how it can be used and for what purpose. Another problem is that this data is often biased and fails to accurately reflect the full diversity of the patient pool, which can cause certain segments of the population to be diagnosed more frequently.
A far more serious issue is that generative AI has a tendency to produce factually incorrect information, which can, for obvious reasons, be particularly troublesome in a healthcare context. The lack of transparency and explainability related to how AI algorithms arrive at a certain decision is yet another point of contention. How can a doctor be confident in the output produced by generative AI if they don’t know the reasoning behind it? Or better yet, will patients be able to trust that the decision will be in their best interest if doctors themselves can’t explain why a specific procedure needs to be done?
“Generative AI is a new and exciting tool to help us in our mission to simplify how our practices deliver excellent patient care”.
Bill Lucchini, chief operating officer of Nextech
The AI-assisted doctor will see you now
There are a growing number of tools that can help doctors automate some of the more mundane tasks and give their full attention to their patients.
A number of companies have developed AI-powered tools aimed specifically at doctors. Generative AI startup Tortus, for example, has developed a virtual assistant named Operating System Leverage in Electronic Records (O.S.L.E.R.), with the goal of automating the generation of clinical documentation. It uses speech recognition technology to record and transcribe conversations between doctors and patients, automatically generating visit summaries, lab requests, and medical code suggestions at the end of each visit. This enables doctors to focus fully on the patient sitting in front of them rather than their computer screen, improving the patient experience and minimising their administrative workload.
“I know that over-worked, stressed clinicians can’t deliver the best quality care to patients, and that this negatively impacts their patients’ experiences and outcomes,” says Dr Dom Pimenta, co-founder and chief executive officer of Tortus. “We created O.S.L.E.R to free clinicians from their keyboards and the burden of administrative tasks. The technology delivers an immediate improvement to the engagement between a clinician and patient, which is the foundation of every healthcare system.” To assess the tool’s effectiveness in clinical settings, the company established a partnership with Great Ormond Street Hospital, one of the world’s leading children’s hospitals, where O.S.L.E.R. will be used to document notes and letters in outpatient clinics.
There are, of course, numerous companies that develop products with similar functionalities. One of them is the healthcare technology company Nextech, which offers three AI tools that can help doctors improve their efficiency. The first is a virtual assistant which greets patients online, responds to their inquiries, and schedules doctor appointments. The AI Scribe, on the other hand, listens in as the doctor interacts with the patient and automatically fills out the patient’s chart with relevant medical information. Finally, the AI Support tool streamlines access to support resources and enables doctors to instantly get answers to their questions. “Generative AI is a new and exciting tool to help us in our mission to simplify how our practices deliver excellent patient care”, says Bill Lucchini, chief operating officer of Nextech.
Not to be outdone by its Western counterparts, the Hong Kong-based Centre for Artificial Intelligence and Robotics (CAIR) has launched its own AI assistant for doctors named Cares Copilot. Based on Meta’s Llama 2 large language model, the tool can process many different types of data, including text, images, videos, and voice. It can also analyse MRI and CT scans, as well as ultrasound images, helping doctors make more accurate diagnoses and find the most appropriate treatments. The chatbot was trained on millions of records, including medical literature, teaching materials, and expert guidelines, which allows it to access relevant information within seconds. Perhaps most notably, it does so with up to 95% accuracy, according to CAIR.
“It’s no longer a question of whether LLMs will replace doctors in the clinic—they won’t—but how humans and machines will work together to make medicine better for everyone.”
Ethan Goh, a healthcare AI researcher at Stanford’s Clinical Excellence Research Center
Making medicine better for everyone
Experts largely agree that generative AI can help doctors do their jobs, but it is unlikely to fully replace them.
While the adoption of generative AI in healthcare is still in its infancy, most experts agree it could have a significant impact on the sector’s future. “We can use this technology and all of the data we’re collecting about the patient, to get insights from a multimodality perspective—insights from things like imaging, echocardiograms, and electrocardiograms, that maybe I as a human can’t see, but AI can and allow me to act on it…before events happen,” explains Dr Ashley Beecy, medical director of AI operations at NewYork-Presbyterian (NYP), one of New York’s leading hospital systems.
Probably the biggest immediate benefit of generative AI is reducing the amount of time doctors lose to administrative work. “After I see a patient, I have to write notes, I have to place orders, I have to think about the patient summary,” says Dr Shiv Rao, founder and chief executive officer of healthcare AI startup Abridge. “So what this technology does is it allows me to focus on the person in front of me—the most important person, the patient—because when I hit start, have a conversation, then hit stop, I can swivel my chair and within seconds, the note’s there.”
Others are quick to emphasise that the technology still has certain limitations. “Quite often, these AI models are not designed in an inclusive way, or they have some hidden biases in the models,” explains Dr Dhireesha Kudithipudi, founding director of the MATRIX AI Consortium at the University of Texas at San Antonio. “In those contexts, human intervention becomes important to mitigate the blind spots that AI has and avoid these biases being translated into actual impact.” David Atashroo, chief medical officer of AI-powered surgical management platform Qventus, also urges caution. “This means not only developing models with rigorous quality checks, but also incorporating regular assessments by human experts to validate their performance,” he says.
Most experts seem to agree that generative AI is very unlikely to replace human doctors. “We know AI is going to transform how healthcare is delivered. It already is doing that,” argues Dr Bernard Chang, dean for medical education at Harvard Medical School. “I don’t believe that AI will replace human physicians, but those who use AI will be so much more capable than those who don’t.” Ethan Goh, a healthcare AI researcher at Stanford’s Clinical Excellence Research Center, shares a similar view. “It’s no longer a question of whether LLMs will replace doctors in the clinic—they won’t—but how humans and machines will work together to make medicine better for everyone.”
How do doctors and patients feel about generative AI?
While the majority of doctors are ready to embrace generative AI, patients appear to be slightly more hesitant about the idea.
Ask almost any doctor and they will tell you they are overworked. A February 2024 survey from healthcare technology company Athenahealth reveals that more than 90% of doctors feel burned out on a regular basis. Among those, 60% blame their struggles on the amount of paperwork they are required to fill out every day. To keep up with the administrative workload, doctors work 15 hours per week on average outside their normal hours. When asked whether AI could help them address this problem, 83% of respondents answered positively.
It appears that a growing number of doctors are jumping on the generative AI bandwagon. According to an April 2024 survey from information services company Wolters Kluwer, 68% of doctors say that they are enthusiastic about the benefits generative AI can bring to their industry. When asked what those benefits might be, 81% said that they believed generative AI could help improve interactions with patients, with a further 68% stating it could help them search medical literature more quickly, and 54% expressing a belief that it would save them 20% or more time by summarising patient data. Perhaps most importantly, 40% of respondents said they are ready to use generative AI when interacting with patients, while nearly 90% would be comfortable with it if they knew where the data came from.
Patients, on the other hand, appear slightly more hesitant. A 2023 PYMNTS Intelligence study found that 60% of respondents are uncomfortable with the idea of their healthcare provider using AI. 57% of respondents said that they believed that the use of AI for diagnoses or treatment recommendations could have a negative impact on the patient-provider relationship. Despite these concerns, 42% have expressed interest in the implementation of AI. The potential benefits of AI are not lost on patients, with 40% saying they believed it would help reduce mistakes, while 38% believe it would lead to better patient outcomes.
Taken as a whole, these figures indicate that the adoption of generative AI within healthcare will continue to accelerate in the coming years. This is further supported by the findings presented in a 2023 Capgemini Research Institute survey, which revealed that 56% of pharma and healthcare organisations have started exploring generative AI’s potential, while 37% have already started working on some pilots. The generative AI market for healthcare is predicted to grow from US$1 billion in 2022 to nearly US$22 billion by 2032, according to PYMNTS Intelligence.
Should doctors be worried about their jobs?
Generative AI is set to take on a more prominent role in healthcare in the future, but doctors don’t need to fear for their jobs just yet.
As generative AI continues to advance, it’s likely to take over a growing number of tasks that were previously performed by doctors. Does this mean that the technology may one day replace human doctors entirely? Probably not. While generative AI has an unprecedented ability to process massive amounts of data quickly and uncover hidden insights, it still lacks certain qualities that are essential for effective medical care, such as empathy and intuition. A doctor typically employs a holistic approach that combines vast medical knowledge with a nuanced understanding of the patient’s individual needs and life circumstances, something that generative AI still cannot—and may never be able to—replicate.
While it may not take away their jobs, generative AI will almost certainly change the way doctors work. No longer required to handle mundane tasks like record keeping, doctors will be able to expend more energy engaging meaningfully with with their patients, exploring innovative treatment approaches, or even furthering their education. The latter is particularly important as the growing implementation of generative AI in healthcare will require doctors to obtain a new set of skills that will enable them to realise the technology’s full potential. They will need to learn not only how to use AI tools but also how to critically evaluate their outputs and understand their limitations to ensure optimal outcomes for their patients.
Learnings:
It is readily apparent that generative AI has the potential to address some of the challenges that currently plague our healthcare systems, such as the increasing demand for medical services and the growing administrative burden placed on doctors. Combining human expertise with advanced algorithms can help hospitals significantly reduce the number of diagnostic errors and ensure that each patient receives appropriate treatment.
- Generative AI can help doctors automate administrative tasks and make more accurate diagnoses.
- It can also help detect diseases at an earlier stage, increasing the chances of patient recovery.
- While it could increase productivity, it’s unlikely that generative AI will ever fully replace human doctors.
- An April 2024 survey by Wolters Kluwer reveals 68% of doctors are enthusiastic about the benefits generative AI can bring to healthcare.
- A 2023 PYMNTS Intelligence study revealed that 60% of consumers are uncomfortable with the idea of their healthcare provider using AI.
As impressive as generative AI might seem, we must keep in mind the profound ethical and societal implications of integrating such a powerful tool into a field as intimate and consequential as medicine. How do we approach the issues of privacy, accountability, and trust? How do we mitigate the risks of perpetuating biases and ensure that there are no negative consequences for the patients?
