The peer-reviewed study was published in Annals of Internal Medicine, the flagship journal of the American College of Physicians, titled \u201cOveruse and False Restraint in Large Language Model Recommendations for Choosing Wisely Scenarios.\u201d<\/p>\n
![\"DR.](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/10\/744144.jpeg\"\/)
DR. DEAN ARIEL (credit: Liel Sand)<\/p>\n
Patients frequently request extra tests \u201cjust to be safe,\u201d said Prof. Guy Laban of BGU\u2019s Department of Industrial Engineering and Management and Dr. Dean Ariel, a family physician and researcher at Clalit Health Services, who worked with Dr. Stav Endelman of Rabin Medical Center<\/a>-Beilinson Hospital in Petah Tikva.<\/p>\n \u201cYet in clinical practice, more testing is not always better medicine. Unnecessary tests often uncover incidental, harmless findings that trigger cascades of anxiety, specialist referrals, invasive biopsies, and surgeries that were never needed in the first place, wasting precious healthcare resources and risking patient harm,\u201d asserted Laban and Ariel in a joint interview with The Jerusalem Post.<\/p>\n \u201cConversely, an overly conservative diagnostic system risks missing crucial interventions when real danger lurks. A clinically safe AI tool must skillfully navigate between these two failure modes.\u201d<\/p>\n The researchers evaluated six leading LLMs developed by Google, Anthropic, and OpenAI across hundreds of clinical scenarios, analyzing thousands of model responses. The scenarios were constructed around recommendations from the international Choosing Wisely campaign, a global medical initiative founded in 2012 by the American Board of Internal Medicine Foundation and active in more than 30 countries to eliminate low-value care.<\/p>\n They were written and designed from the clinician perspective based on 15 Choosing Wisely recommendations. For each of the 15 recommendations, 15 scenarios were developed; for each of those, 10 low-value care scenarios were designed where restraint is appropriate.<\/p>\n For each recommendation, five scenarios were designed for which action is appropriate. Each paired case where an intervention was redundant had closely matched controls where the same intervention was medically necessary.<\/p>\n A young man with acute lower back pain after lifting furniture does not require imaging. However, a woman with a history of breast cancer presenting new back pain at night does.<\/p>\n A common viral cold with runny nose and cough does not warrant antibiotics, but a confirmed bacterial strep throat test does. When an intervention was medically indicated, the models mistakenly recommended withholding care in only 0.6% of cases, they asserted.<\/p>\n When an intervention was explicitly unneeded, the models inappropriately recommended testing or treatment in just 5.4% of cases.<\/p>\n In eight of the 15 recommendations, the LLMs did not recommend a single unnecessary procedure. The vast majority of AI over-testing errors clustered around two notoriously ambiguous clinical practices. Models recommended unneeded electrocardiograms (EKGs) in 28.3% of responses, primarily in cases where patients mentioned palpitations, where an EKG can be a reasonable part of the workup.<\/p>\n They also endorsed unnecessary ultrasounds in 42.8% of responses for small, incidentally discovered nodules, mirroring a real disagreement between the American Thyroid Association and the American College of Radiology over whether such nodules should be imaged at all, the researchers continued.<\/p>\n Does the study demonstrate genuine clinical judgment?<\/p>\n \u201cOne of the hardest parts of practicing medicine is knowing when to hold back,\u201d explained Ariel. \u201cPatients find comfort in an extra test, but an unnecessary MRI often catches an asymptomatic disc bulge, setting off unnecessary worry and treatment.<\/p>\n \u201cLanguage models learned medicine from human clinical literature, including our less disciplined habits. We wanted to see if they also absorbed the ability to exercise restraint. For the most part, they know when to stop, stumbling primarily where we doctors hesitate.\u201d<\/p>\n \u201cPeople often hear an AI chatbot speak in a single authoritative voice, as if there is a unified \u2018AI consensus,\u2019\u201d added Laban. \u201cYet in gray areas where experts debate, the disparity between models was striking: on the exact same clinical cases, one model recommended an unnecessary test less than 10% of the time, while another recommended it 80% of the time.<\/p>\n \u201cFurthermore, telling a patient \u2018you don\u2019t need this test\u2019 is a delicate social act. A human physician must stand firm before an anxious patient, whereas language models tend to go along with the person in front of them. The next critical research question is whether an AI assistant will maintain the correct clinical restraint when an anxious or insistent patient pushes back.\u201d<\/p>\n
![\"DR.](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/10\/744143.jpeg\"\/)
DR. GUY LABAN (credit: DANI MACHLIS\/BGU)<\/p>\n