AI in the clinic: a scoping review of AI and clinical expertise
This scoping review of 58 empirical studies from 2022 onwards indicates that while AI tools significantly enhance physician clinical performance—particularly for novices—current evidence does not support fears of AI-induced deskilling, though further longitudinal research is needed to fully assess long-term impacts on clinical expertise.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a medical clinic as a busy kitchen. For years, the head chefs (experienced doctors) have cooked meals using their own intuition and years of training. Recently, a new type of "smart assistant" (Artificial Intelligence) has entered the kitchen. This isn't just a simple timer; it's a sophisticated tool that can taste the soup, suggest ingredients, and even spot a burnt crust before the chef does.
However, some people are worried. They fear that if chefs rely too much on this smart assistant, they might forget how to cook on their own. They worry the chefs will lose their "taste buds" (clinical expertise) and become dependent on the machine.
This paper is like a team of food critics who went into 58 different kitchens to see exactly what happens when these smart assistants are used. Here is what they found, explained simply:
1. The "Smart Assistant" is Mostly a Helper, Not a Replacement
The researchers looked at studies from 2022 to 2025. They found that in almost every kitchen they visited, the smart assistant helped the chefs cook better.
- The Result: In 84% of the cases, the doctors' performance actually got better when they used the AI.
- The Analogy: Think of the AI as a super-powered pair of glasses. When a chef puts them on, they can see the ingredients more clearly. The paper found that these glasses helped doctors spot problems (diagnose illnesses) much more accurately.
2. The "Rookies" Benefit the Most
Here is the most interesting part: The new chefs (novice doctors) gained the most from these glasses.
- The Result: When inexperienced doctors used the AI, they often cooked as well as the master chefs who didn't use any help at all.
- The Analogy: Imagine a cooking student who usually burns the toast. With the smart assistant, they suddenly make toast as perfect as a 20-year veteran. The AI seems to be a great equalizer, helping the less experienced catch up to the experts quickly.
3. The "Skill Loss" Fear Wasn't Found (Yet)
This is the big question: If the chefs use the glasses all the time, will they forget how to cook without them?
- The Result: The researchers looked for evidence of this "skill loss" (deskilling). They found almost none. In fact, only one study out of the 58 even checked what happened after the chefs took the glasses off. In that single case, the chef's performance did drop slightly. But in the other studies that checked, the chefs stayed sharp or even got better.
- The Analogy: It's like a musician using a digital tuner. The fear is that if they use the tuner every day, they'll forget how to tune their own instrument. This paper says, "We haven't found proof that they are forgetting yet." Most of the time, the musicians are still playing great, even if they aren't being tested on what happens when the tuner is taken away.
4. The Kitchen is Still a Bit Small
While the results look promising, the researchers noted a few limitations in how the studies were done.
- The Sample Size: Many of the "kitchens" they visited were very small. Some studies only had a handful of doctors (sometimes as few as 2 or 6).
- The Time: Most of these studies were like snapshots taken in a single day. They didn't watch the chefs cook for months or years to see if the smart assistant caused long-term problems.
- The Focus: The studies mostly looked at how well the doctors could diagnose (spot the burnt toast). They didn't really look at other important skills, like how the doctors talked to patients, worked in teams, or handled ethical dilemmas.
The Bottom Line
The paper concludes that right now, the fear that AI will make doctors "dumber" isn't supported by the evidence we have. Instead, the evidence suggests AI is a powerful tool that helps doctors, especially the newer ones, perform better.
However, the authors warn us not to get too excited just yet. We need to watch these kitchens for a longer time (long-term studies) and with more chefs (larger groups) to make sure the "smart assistant" doesn't eventually cause the chefs to lose their touch. For now, the AI looks like a helpful sous-chef, not a threat to the head chef's skills.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.