Comparison Reveals Human Physicians Outperform Virtual Diagnosticians

Increasingly powerful computers using ever-more sophisticated programs are challenging human supremacy in diverse areas. Can current digital diagnostics programs outperform, or even match, human physicians? The answer, according to a new study, is still “far from it.”

The findings of the study, led by researchers at Harvard University Medical School (Boston, MA, USA), show that physicians’ performance is vastly superior as they made a correct diagnosis more than 2x as often as 23 commonly used symptom-checkers. The analysis is believed to be the first direct comparison between human-made and computer-based diagnoses.

Diagnostic errors stem from failure to recognize a disease or to do so in a timely manner. Physicians make such errors roughly 10-15% of the time, researchers say. Over the last two decades, computer-based checklists and other digital apps have been increasingly used to reduce medication errors or streamline infection-prevention protocols. Lately, experts have wondered whether computers might also help improve clinical diagnoses and reduce diagnostic errors. Many people use apps or Internet programs to check their symptoms or to self-diagnose, yet how these computerized symptom-checkers fare against physicians has not been well studied.

In the study, 234 internal medicine physicians were asked to evaluate 45 clinical cases, involving both common and uncommon conditions with varying degrees of severity. For each scenario, physicians had to identify the most likely diagnosis along with two additional possible diagnoses. Each clinical vignette was solved by at least 20 physicians.

The physicians outperformed the symptom-checker apps, listing the correct diagnosis first 72% of the time, compared with 34% of the time for the digital platforms. 84% of clinicians listed the correct diagnosis in the top 3 possibilities, compared with 51% for the digital symptom-checkers. The difference between physician and computer performance was most dramatic in more severe and less common conditions, and smaller for less acute and more common illnesses.

“While the computer programs were clearly inferior to physicians in terms of diagnostic accuracy, it will be critical to study future generations of computer programs that may be more accurate,” said senior investigator Ateev Mehrotra, associate professor of healthcare policy HMS.

Despite outperforming the machines, physicians still made errors in about 15% of cases. Developing computer-based algorithms to be used in conjunction with human decision-making may help further reduce diagnostic errors. “Clinical diagnosis is currently as much art as it is science, but there is great promise for technology to help augment clinical diagnoses,” said Prof. Mehrotra, “That is the true value proposition of these tools.”

The study, by Semigran HL et al, was published online October 10, 2016, in the journal JAMA Internal Medicine.

Related Links:
Harvard University Medical School