AI in Healthcare: Why Doctors Need to Learn Data Science
Artificial intelligence (AI) is revolutionizing healthcare, but there’s a glaring gap in the system: most clinicians and medical students still view data science training as optional. This needs to change, and fast. According to a forthcoming Medscape report, AI Adoption in Healthcare, a staggering 85% of practicing physicians agree that the rise of AI in medicine demands significant changes in medical education and training.
Experts echo this sentiment. To effectively use AI tools while maintaining clinical judgment and safeguarding patients, doctors must understand how AI works and the data it relies on. The future of healthcare depends on a new breed of physicians who can skillfully navigate this technology.
Why Doctors Must Stay in Control
AI is not a magic wand. While it holds the potential to transform healthcare, it’s not infallible. Mistakes will happen, and that’s why doctors must remain in control. Shauna Overgaard, PhD, senior director for AI Strategy & Frameworks at the Mayo Clinic Center for Digital Health, emphasizes the importance of autonomy. “It’s never about just trusting AI,” Overgaard told Medscape Medical News. “We want physicians to maintain their judgment, their empathy, their perception of control—all things that contribute to [their] skepticism.”
Doctors need to know when to trust AI and when to question it. Just as they’re trained to evaluate a study’s methodology or assess a new drug, they’ll need to scrutinize AI algorithms. Overgaard explains that physicians must ask critical questions: Does the AI’s output make sense? What data informed its predictions? Which patients benefit from the algorithm, and where does it fall short?
“Clinicians don’t need to outsmart the system,” Overgaard adds, “but they do need to know when to call BS.” Ultimately, the responsibility for care decisions remains with the physician. As Oren Mechanic, MD, MPH, a private practice physician in Miami, puts it, “We can’t put in the chart, AI told us to make this decision.”
To strike the right balance between collaboration and skepticism, doctors need more training—starting now.
Building a Data Science Foundation
The ideal starting point for AI education is medical school. Peter Steel, MD, an emergency medicine physician and AI researcher at Weill Cornell Medical School, envisions data science becoming a core part of the curriculum. “Doctors-in-training will get a fundamental understanding of AI,” Steel says. This includes learning about data analytics, machine learning, limitations, and bias.
With this foundation, future doctors will be better equipped to critically interpret AI outputs and cross-check them against their clinical judgment. They’ll be able to identify bias, recognize hallucinations, and know how to respond when AI recommendations conflict with their own decisions.
According to Mechanic, laying this foundation could be as simple as offering a single course. As students encounter AI in various training and clinical settings, their skills will naturally evolve. Beyond technical knowledge, doctors will also need to learn how to communicate AI-based insights to patients. Steel highlights the importance of explaining where AI outputs come from and how reliable they are, even when the solutions are complex or not entirely explainable. Building trust with patients will be key.
Currently, data science training is mostly offered as elective courses driven by student interest, says Kim Lomis, MD, vice president of Medical Education Innovations at the American Medical Association. But Lomis warns that it’s now “urgent” for medical schools to integrate AI training into their required curriculum.
However, this is easier said than done. Medical schools are already overwhelmed with teaching an extensive curriculum in just four years. Additionally, many institutions lack the expertise to teach emerging AI content. Steel suggests that medical schools may need to partner with advanced data science faculty to meet this challenge.
Training Across All Levels of Medicine
AI education isn’t just for medical students. The entire clinical workforce needs to understand how to deliver AI-augmented, clinician-led care. Yet, much like undergraduate medical education, AI training for graduate and postgraduate levels remains largely elective.
Some institutions, like the Mayo Clinic and the University of Louisville, offer master’s programs in AI specifically for medical professionals. However, Overgaard notes that these programs are more suited for doctors aiming to lead AI initiatives in their specialties. For most physicians, shorter courses or certificates in healthcare AI—offered by institutions like Stanford, Dartmouth, Harvard, and MIT—may be more practical.
But external education won’t be the only solution. Healthcare systems themselves will need to take responsibility for training their workforce. Steel points out that if clinicians feel unprepared or apprehensive, they won’t adopt AI tools, no matter how advanced they are. To see a return on their AI investments, healthcare systems must equip their frontline workers with the necessary skills.
Overgaard adds that many clinicians are already seeking out AI education. The Mayo Clinic, for example, has heavily invested in AI training, incorporating it into their programs and partnering with developers like Google. Smaller healthcare systems are likely to benefit from a trickle-down effect, as larger institutions test and refine AI tools and share their findings.
“We have a responsibility as this organization with an incredible amount of funding and intellectual capacity not just to serve ourselves,” Overgaard says, “but set up systems, make recommendations, and create frameworks other [health] systems can adopt and improve upon.”
The Road Ahead
As AI continues to reshape healthcare, the need for comprehensive education becomes increasingly urgent. From medical school to postgraduate training and beyond, clinicians must be equipped to navigate this new frontier. The stakes are high, but the potential rewards—improved patient outcomes, greater efficiency, and groundbreaking innovations—are worth the effort.
Doctors who can master AI will not only enhance their own practice but also help shape the future of medicine. The question is no longer whether AI will play a role in healthcare, but how well-prepared clinicians will be to harness its power.
Originally Written by: Donavyn Coffey
