Ran Shaul

While the final treatment decision is with the physician, AI may have a role in assisting decision making in virtual urgent care.

AI Bests Clinicians in Urgent Care Treatment Recommendations, with Zehavi Kugler, MD, and Ran Shaul

Artificial intelligence (AI) is increasingly transforming medical care, evolving from automating administrative tasks to influencing clinical decision-making. AI-driven tools can now analyze vast datasets to identify disease patterns, assist in imaging interpretation, and suggest evidence-based treatment options. While these technologies have the potential to enhance efficiency, accuracy, and patient outcomes, they also raise concerns regarding bias, data security, and the need for clinician oversight. AI’s continued integration into healthcare must be carefully managed to complement human expertise and improve the quality and accessibility of patient care.

A study evaluating one such application of an AI model was presented at the 

American College of Physicians (ACP) Internal Medicine (IM) Meeting 2025

, held April 3-5, in New Orleans, Louisiana, by Zehavi Horowitz-Kugler, MD, a primary care physician at a private practice in Israel and vice president of Medical Sciences at K Health.

spoke with Kugler and Ran Shaul, cofounder and chief product officer of K Health, to learn more about the study’s findings and the Cedars-Sinai Connect AI-assisted virtual urgent care clinic that they developed in collaboration with Cedars-Sinai. The study compared initial AI recommendations with final recommendations of physicians who had access to the AI recommendations and may or may not have viewed them.

Kugler and colleagues found that among 461 weighted visits, AI recommendations were more frequently rated as optimal (77.1% [95% CI, 72.7-80.9]) compared with treating physician decisions (67.1% [95% CI, 62.9-71.1]) by third party physician adjudicators. Quality scores were equal in 67.9% (95% CI, 64.8-70.9) of cases, better for AI in 20.8% (95% CI, 17.8-24.0), and better for treating physicians in 11.3% (95% CI, 9.0-14.2) of cases.

Physician adjudicators were not blinded to the source of the recommendations, representing one limitation of the new study.

“[Some] physicians are more attached to the way that they're practicing and would even tell themselves that [some cases are] trivial, like treating UTI or an upper respiratory infection, and [are] not something that you need help with. I think the study demonstrated that that's wrong, and it could be valuable, even for the more simple bread and butter cases,” Kugler told 

Kugler and Shaul's disclosures include K Health.

Zeltzer D, Kugler Z, Hayat L, et al. Comparison of Initial Artificial Intelligence (AI) and Final Physician Recommendations in AI-Assisted Virtual Urgent Care Visits. Presented at: ACP IM Meeting 2025; April 3-5; New Orleans, Louisiana. doi: 10.7326/ANNALS-24-03283

Ran Shaul

Articles