Every industry is reckoning with the role artificial intelligence will play in its future, but the possibilities for healthcare are especially promising. From the first contact with patients to the operating room, AI is already transforming and personalizing patient care. Research suggests that in 2021 the AI healthcare market was worth around 11 billion U.S. dollars worldwide. By 2030, that number could reach $188 billion.
Done right, AI can personalize the patient experience, speed up care, and improve outcomes. “The hype around AI in healthcare is now translating into practical applications and tangible benefits for patients,” according to Dr. Clément GOEHRS, MD, MSC, and CEO & Co-Founder of Synapse Medicine. With that in mind, let’s explore some current trends and future possibilities for AI in patient care.
How AI is transforming clinical decision support
AI is already hard at work assisting clinicians and transforming patient experiences. As Dr. Stephen Lane, MD, MPH, and Chief Medical Officer of Health Gorilla, pointed out in a webinar hosted by Synapse Health and Health Gorilla, “Today, mature [electronic health records] EHRs provide extensive automated decision support, like drug-drug and drug-allergy interaction checks. They're integral to everyday clinical practice, reminding clinicians of critical considerations like a patient's creatinine clearance before prescribing specific medications.”
A well-trained model can help accurately diagnose patients, predict how they will react to drugs, and expedite diagnosis. As Medical Economics points out, “This not only expedites the diagnostic journey but also ensures a higher level of precision, minimizing errors and maximizing treatment efficacy.”
Oftentimes, however, AI comes into play before a patient even sees a doctor. Chatbots can handle intake, ask patients about their symptoms, perform triage, or just schedule an appointment. This frees up doctors and their staff to focus on higher-level care where AI is increasingly useful. For instance, AI can detect abnormalities in scans the human eye may easily gloss over.
It can also provide prescription support with personalized medical recommendations:
The examples are myriad, but AI still faces challenges, especially in the life-or-death market of healthcare.
Challenges for AI in healthcare
For all its usefulness, AI is only as good as the data it is trained on. As Dr. Lane said of the increased reliance on AI, “It raises questions about the reliability of these systems and the currency of their data. For instance, how does a physician know if their EHR database is up-to-date with the latest drug interactions?” The questions about data, however, go far beyond currency.
Health Data Management reports, “According to the American Civil Liberties Union, inherent biases in the data used to train AI language models have been found in numerous examples, resulting in gender and racial unfairness.” It will “be incumbent upon healthcare enterprises and vendors to monitor healthcare data for bias errors.” The teams behind the tools — and research in general — are increasingly aware of these blind spots and are working to eliminate them.
“Our goal is to ensure that clinical trials consider population biases, a factor critical for AI and machine learning applications in clinical decision support systems,” said Shelly Spiro, Executive Director of the Pharmacy HIT Collaborative (PHIT), in the same webinar. “We aim to redefine the terminology used…” she added. “This change is crucial for accurately identifying clinical manifestations and side effects noted in clinical trials. Before applying AI to these tools, such foundational changes are essential.”
Getting the data right is important, but so is getting buy-in from the clinicians using the technology. Luckily, the scales are beginning to tip when it comes to adoption.
Elsevier Health’s Clinician of the Future 2023 report found just 11% of clinical decisions are currently assisted by generative AI tools. However, the future is brighter for AI as 48% of respondents say using generative AI tools in clinical decision-making will be desirable in 2-3 years. Additionally, “Over half (51%) of clinicians welcome the prospect of medical students using generative AI-powered tools as part of their medical education in the next 2-3 years.” This may indicate the next generation of doctors will pave the way for AI to become a full partner in patient care.
The bigger hurdle when it comes to adoption may be patients themselves. According to Pew research, 60% of U.S. adults “say they would feel uncomfortable if their own health care provider relied on artificial intelligence to do things like diagnose disease and recommend treatments; a significantly smaller share (39%) say they would feel comfortable with this.” Clinicians must tread lightly in pushing AI on patients; moving slowly may be the key to ultimate success.
Clinical decision support and the future of AI in precision medicine
Researchers are already finding new and innovative ways to use AI and machine learning to transform healthcare. Take, for instance, an algorithm developed by Yale researchers in 2023 that can diagnose aortic stenosis from an echocardiogram performed by a primary care doctor, eliminating the need for a doppler exam with a radiologist. This is just one example of how machine learning can speed up diagnosis — and cut costs for the patient — through pattern recognition.
AI is also helping caregivers cope with the demands of our current moment. Another Yale research project deployed AI to predict COVID-19 outcomes for emergency room patients. The Yale Daily News reports, “hospitals often run out of beds during COVID-19 outbreaks. AI-powered predictions could help determine which patients need inpatient care and which patients can safely recover at home.”
Given patients’ skepticism, much of the daily use of AI in healthcare settings is still focused on automating repetitive tasks despite its potential. However, in underserved areas with a shortage of healthcare providers, AI could prove life-changing.
As the National Institutions of Health (NIH) reports, when combined with wireless technology and remote tracking, AI can revolutionize healthcare in locations where providers — especially specialists — are few and far between. This can also bring down healthcare costs: “It is estimated that AI applications can cut annual US healthcare costs by USD 150 billion in 2026. A large part of these cost reductions stem from changing the healthcare model from a reactive to a proactive approach, focusing on health management rather than disease treatment.”
AI can also help target treatment to the individual. The NIH says, “Precision medicine provides the possibility of tailoring healthcare interventions to individuals or groups of patients based on their disease profile, diagnostic or prognostic information, or their treatment response.”
Ultimately AI may usher in a new world of bespoke treatment that is personalized to each individual — the impacts of which may be staggering.
