Wrong automated advice cuts clinicians' accuracy, meta-analysis finds
Physician in the Loop, Oct 24, 2026
Listen to this post
read by AI voices ยท all episodesTHE SHORT READ
Clinicians had about 10 times the odds of a correct decision when automated advice was right than when it was wrong, a meta-analysis of 46 reports published Oct. 23 in npj Digital Medicine found.
In four real-world studies of computerized electrocardiogram readings, 26.6% of erroneous machine interpretations went uncorrected and unrecognized, the authors reported. Studies of current AI and large language models documented harmful reliance on wrong output but were too varied to pool, and only one study reported harm at the patient level. Human oversight is protective "only if clinicians recognize and resist erroneous output," wrote the authors, from West China Hospital at Sichuan University.
Elsewhere, Quest Diagnostics told regulators an unauthorized party reached patient records through the Quanum system clinicians use to order lab tests, and the Department of Health and Human Services moved to keep its workforce on Anthropic's Claude without a competitive bid.
THE DETAILS
- FYIEvidence Oct 23, 2026The review in npj Digital Medicine pooled 46 reports covering 41 independent groups of clinicians. In three adjusted cohorts the odds of a correct decision were 10.19 times higher with correct advice than with incorrect advice, and a sensitivity analysis limited to studies from 2020 onward found a similar but less precise result. In four real-world studies of computerized ECG interpretation, 26.6% of erroneous readings were neither corrected nor recognized, with wide variation among studies. The authors wrote that human oversight "should be evaluated as an error-interception capability rather than assumed to ensure safety."Source: npj Digital Medicine
- WatchIncident Oct 22, 2026
Quest said the full scope, including which data and how many patients were affected, is still under investigation.
Quest learned of the access on Oct. 21 and disclosed it in a filing with the Securities and Exchange Commission dated Oct. 22. The affected system, Quanum Lab Services Manager, is used by certain clients to place test orders and to receive and look up results, the company said. Quest paused access to the system, notified federal law enforcement and hired outside cybersecurity experts, and it has not said how many records were involved, according to MedTech Dive. The company said the incident has not materially disrupted its operations or its ability to serve patients.Sources: Quest Diagnostics SEC filing, MedTech Dive - FYIRegulation Oct 20, 2026The Department of Health and Human Services posted a notice of intent on Oct. 19 for a "limited continuity of services award" to Anthropic, saying the capability "is specific to one provider," FedScoop reported. The notice lists no value and covers chat, drafting, summarization, data analysis and administrative controls; it says HHS data is excluded from model training by default and stays in the continental United States, according to Becker's. HHS rolled out Claude across the department in December 2025, and Anthropic's discounted government-wide deal ends Oct. 31. Jessica Tillipman, a procurement law scholar at George Washington University, questioned whether the notice justifies skipping competition, FedScoop reported.Sources: FedScoop, Becker's Hospital Review
- FYIEvidence Oct 22, 2026The Center of Surgical and Medical Intelligence sits within the system's Smith Institute for Evidence, Innovation and Policy and will build unified data systems, benchmarks and tools spanning the patient journey, Beth Israel Lahey Health said Oct. 22. Its directors are Dr. Adam Rodman, a hospitalist at Beth Israel Deaconess Medical Center, and Dr. Gabriel Brat, a trauma surgeon and intensive care physician there. "AI systems are going to succeed or fail based on the quality of our data," Brat said. The system did not disclose funding; it has more than 6,300 physicians.
- ActRegulation Oct 20, 2026
FDA asks providers to check model numbers against its Oct. 20 table; devices already in use may continue if they show no problems.
The Food and Drug Administration updated its letter to health care providers on Oct. 20, adding the Endocapsule small intestinal capsule set, the Evis Exera II small intestinal videoscope, a deflectable videoscope and two bronchovideoscopes made in Japan to an import alert first issued in June 2025. The agency said it does not recommend canceling or delaying procedures without discussing benefits and risks with patients. A warning letter dated Oct. 7 cited quality system and adverse event reporting violations, including risk files that omitted hazards such as air embolism and death, MedTech Dive reported. Olympus said it does not expect a material impact on customers.Sources: FDA, MedTech Dive - FYIDeployment Oct 22, 2026The tool lets case managers query charts and see changes from the past 24 hours, and Sentara plans to extend it across acute care and into ambulatory care, Fierce Healthcare reported Oct. 22. Executives projected savings of nearly 90,000 hours, the equivalent of more than 40 full-time employees and nearly $5 million in labor costs; the figures are projections, not measured results. Sentara runs 12 hospitals and more than 400 care sites in Virginia and northeastern North Carolina.Source: Fierce Healthcare
- FYIRegulation Oct 21, 2026Australia's Therapeutic Goods Administration surveyed about 100 digital scribe suppliers, received 40 responses and reviewed 12 products, including those with the largest market shares, Healthcare IT News reported. The agency did not name the eight and did not test safety or performance; reviewed products included Heidi, Lyrebird Health and Microsoft's, according to The Medical Republic. The agency said disclaimers are "not sufficient" to settle whether a product is a device and plans targeted compliance activity.Sources: Healthcare IT News, The Medical Republic
- FYIEvidence Oct 21, 2026The Phantom X sensor sits under the skin over the residual limb's muscles and sends wireless commands to a prosthesis, MobiHealthNews reported Oct. 21. The early feasibility study at Cabrini Health in Melbourne is designed to enroll 10 people with below-the-elbow amputations, with about 20 weeks of follow-up; each outpatient implant took about an hour without general anesthesia. The device is investigational and not approved for commercial use, the company said, and no results have been published.Source: MobiHealthNews