JMIR Publications has released two feature articles in its News and Outlook section. Shalini Katuria Narang’s “Can human-like reasoning be replicated in large-scale language models for clinical decision-making?” and Sara Novak’s “How healthcare professionals can cope with digital fatigue” provide complementary insights into the capabilities of artificial intelligence in diagnosis and the real-world fatigue faced by healthcare professionals managing digital systems.
Are LLMs equivalent to doctors in clinical reasoning?
In “Can Humanlike Reasoning Be Replicated in Large Language Models for Clinical Decision-Making?,” Narang discusses recent research comparing diagnostic reasoning in OpenAI’s o1 model to that of physicians. Dr. Narang reports that the model matched or exceeded human performance across three stages of care triage: arrival, first contact with a physician, and admission, with the largest performance gap occurring during initial ER triage, when the most limited information was available. Adam Rodman, a hospitalist and one of the study’s researchers, said the results validate the model’s diagnostic performance but do not mean the system is ready for independent deployment.
Although LLM was good at integrating text-based information, real-world clinical settings rely heavily on non-textual input such as visual and auditory cues collected during physical examinations. Rodman points out that while LLM is great at synthesizing selective data and gathering verbal information, it can’t replace a doctor’s ability to physically examine a patient, listen to stray voices, and integrate messy information from multiple uncurated sources.
Rather than replacing doctors, Narang writes, the future of AI in healthcare requires collaborative integration and careful evaluation. The researchers emphasize the need to conduct prospective trials in real-world settings to safely evaluate new multimodal models. A promising application for the technology, Rodman said, is to catch diagnostic errors before they occur and serve as a second opinion that alerts doctors that they may be on the wrong track.
Face the digital workload
Novak explores a paradoxical phenomenon that has emerged as clinical processes become increasingly integrated with digital tools: digital fatigue. “How Healthcare Workers Deal with Digital Fatigue” explores how, despite the benefits of automation and increased access in clinical care, many healthcare workers struggle to manage digital interfaces and meet the continued demands of responding to redundant alerts. Hearings from experts including physician Hassan Benchekrohn and digital fatigue researchers Rachel Houpsik and Audrey Hajnovak outline how the prevailing piecework health care system already limits the amount of time providers can spend with each patient, and how the additional administrative burden of navigating complex platforms creates a compounding cycle that leads to clinician burnout.
Mr. Novak outlines strategies for organizations and individuals to manage these risks. Experts are urging health systems to streamline digital workflows by reducing low-value prompts, such as non-life-threatening allergy alerts, and eliminating redundant alerts that are often ignored. Restructuring tasks into a team-based system proactively shares responsibilities, such as managing your inbox or refilling your medication, and prevents the accumulation of after-hours administrative work.
Ultimately, it’s up to healthcare organizations to recognize digital tasks as a formal part of their daily routine and ensure they are provided with sufficient time and training during patient assignments, but on an individual level, healthcare workers can schedule digital detox breaks and delay email delivery until after work hours to protect recovery time. “Like any other occupational risk, digital fatigue needs to be taken seriously,” Novak wrote.
sauce:
References:
- Naran, South Carolina (2026). Can human-like reasoning be replicated in large-scale language models for clinical decision-making? Medical Internet Research Journal. DOI: 10.2196/103526. https://www.jmir.org/2026/1/e103526
- Novak, S. (2026). How healthcare workers can manage digital fatigue. Medical Internet Research Journal. DOI: 10.2196/104196. https://www.jmir.org/2026/1/e104196
