Patients can be persuaded to place almost as much confidence in diagnostic AI as they have in primary care physicians.

The conclusion comes from a study in which PCPs defeated machines at earning patient partiality by just six percentage points, 53% to 47%. Researchers at the University of Arizona conducted the experiment. PLOS Digital Health published the resulting study report May 19.

Senior author Marvin Slepian, MD, and colleagues ran the project in two phases. The first was a pretest of the materials in which the team interviewed 24 demographically diverse patients using a structured format. The second was a randomized, blinded survey of more than 2,600 healthcare consumers whose number was oversampled for minority representation.  

For the survey, the team presented participants with clinical scenarios randomly scrambling several variables. These included disease severity, whether AI is proven more accurate than human specialists, whether the AI diagnostician is personalized to the patient through listening and/or tailoring, whether the AI diagnostician avoids racial and/or financial biases, whether the PCP promises to explain and incorporate the advice, and whether the PCP nudges the patient toward AI as the established, recommended and easy choice.

Asking participants to choose one or the other—the doctor or the AI—the researchers found the aforementioned 53% to 47% score. Importantly, responses were weighted to represent the ethnic, racial and demographic makeup of the U.S. adult population, Slepian and co-authors report.

In unweighted experimental contrasts of qualified respondents, acceptance of AI—or AI “uptake”—significantly rose under three conditions:

1. a PCP’s explanation that AI has proven superior accuracy,
2. a PCP’s nudge toward AI as the established choice, and
3. reassurance that the AI clinic had trained counselors to listen to the patient’s unique perspectives.

Interestingly, patients weren’t swayed to choose a human doctor over AI when the hypothesized diagnosis was a dreaded disease such as leukemia (versus a comparatively benign condition like sleep apnea). Also noteworthy: Compared to White participants, Black participants selected AI less often, and Native Americans selected it more often. Also, older respondents were less likely to choose AI, as were those who identified as politically conservative or viewed religion as important.  

In general, higher educational levels corresponded with greater acceptance of AI for medical diagnostics. “To ensure that the benefits of AI are secured in clinical practice, future research on best methods of physician incorporation and patient decision making is required,” the authors comment.

The study is available in full for free, and the university has posted its own news coverage.

Buzzworthy developments of the past few days.

• Men and women experience Alzheimer’s disease with symptoms specific to their respective sexes. And researchers in Texas have used machine learning to find the genes behind the binary. The work is ongoing at Baylor College of Medicine and Texas Children’s Hospital. Nature Communications has published a study describing it. Baylor biochemist and molecular biologist Olivier Lichtarge, MD, PhD, says the team’s innovative AI approach “lets us exploit a massive amount of evolutionary data efficiently.” Free journal study here, Texas Children’s Hospital summary here.
   
• Large language AI deployed in healthcare could cost lives as well as save them. So warns the World Health Organization. WHO believes the chance of being harmed by the technology at some point along the patient journey is a bracing 1 in 300, with error originating in biased or fumbled data representing a gaping pitfall. The World Economic Forum unpacks the risk, and Axios has posted a brisk analysis.
   
• Variables affecting health status at the level of the individual are seven times more costly than medical errors. So reminds New York-based startup Laguna Health. The company has closed on a $15 million funding round it will use to advance AI-powered “contextual care” with an eye on reining in individual patients’ costs while helping address the problem at the population level. Announcement.
   
• Physicians would need to work 27 hours each 24-hour day to handle every clinical and administrative task on their daily to-do lists. Citing the Journal of General Internal Medicine study behind that finding, IKS Health in Dallas and Abridge in Pittsburgh have announced they will partner “expansively” on AI-based ways to head off or relieve physician burnout. Announcement here.
   
• Doctors’ bedside visits to hospital inpatients need to be abstracted into medical codes or the key caregivers won’t get paid. CodaMetrix (Boston) has launched an autonomous AI product to help with that. The company is piloting the system with Henry Ford Health in Michigan. Details here.
   
• AI can identify rare as well as common abnormalities in preborn babies. Researchers in France validated the informed hunch when they tested an ultrasound decision-support product marketed by Sonio (Paris). The academics believe the technology stands to improve perinatal care while curtailing parental anxiety caused by serial testing. Full study published in Ultrasound in Obstetrics & Gynecology (select PDF).
   
• Virtual reality can both prepare future workers and stir young people to choose healthcare over other industries. Count immersive learning startup Transfr among those who believe so. The company is marketing its simulations to schools, staffing services and employers to “create pathways to high-growth careers” in healthcare. Announcement.  
   
• A deep learning tool has helped radiologists cut interpretation times by 40%. Developed at the University of Zurich, the software works by quickly bringing up relevant prior exams in reading workflows. This saves radiologists time they would have spent mousing, clicking and scanning. Study published in Academic Radiology, findings summarized in Health Imaging.