AI summary · may contain errors
In 19-hospital, 10-country emergency departments, the 0/1 h high-sensitivity troponin pathway is non-inferior for 30-day death or new type 1 myocardial infarction versus 0/3 h.
- 6 cites
Moving in medRxiv, Brain, Clinical Infectious Diseases, The Lancet Digital Health, JAMA Internal Medicine, Journal of the American College of Cardiology, Nature Medicine and European Heart Journal.
Rebuilt
Nothing in this topic has moved yet this week.
Newest first · last 60 days
AI summary · may contain errors
In 19-hospital, 10-country emergency departments, the 0/1 h high-sensitivity troponin pathway is non-inferior for 30-day death or new type 1 myocardial infarction versus 0/3 h.
AI summary · may contain errors
A moderated multi-agent LLM diagnostic team improved differential recall on curated CPC and MedCaseReasoning vignettes (top-3 +5.5 and +3.0 points) but reversed on 364 real emergency notes (top-3 −6.4 points).
AI summary · may contain errors
The 2024 McDonald diagnostic criteria for multiple sclerosis broaden acceptable evidence, but their safe use outside expert centres is uncertain given variable imaging, labs, and interpretation.
AI summary · may contain errors
HINTS is most diagnostically effective for acute vestibular syndrome patients with continuous dizziness and spontaneous nystagmus at rest, with accuracy dropping when used in inappropriate populations.
AI summary · may contain errors
VIBE (Violation of Interpersonal Bayesian Expectation) formalizes how psychiatrists’ clinical impressions arise from high-level social prediction errors when interpersonal cues disconfirm expected reciprocity.
AI summary · may contain errors
An AI system using frontier language models achieved 83.2% diagnostic accuracy on New England Journal of Medicine clinicopathological conference cases while selecting cost-effective tests.
AI summary · may contain errors
In a blinded evaluation on 30 psychiatric vignettes, clinician-rated reasoning separated five LLM pairs despite clinician-determined diagnostic accuracy separating none.
AI summary · may contain errors
An LLM-powered, self-administered conversational vignette assessment in Vietnam achieved human-level agreement on transcript coding (90.1% and 89.2%) and AUROCs above 0.91.
AI summary · may contain errors
Public debate about “missed” diagnoses highlights hindsight bias, urging earlier patient engagement in diagnostic reasoning to improve how diagnostic uncertainty is communicated.
AI summary · may contain errors
RoB2-Evaluator, a rule-constrained GPT-5.2 system for Cochrane RoB 2 assessment, achieved 90.0% temporal agreement (weighted kappa = 0.93) across repeated RCT ratings after one week.
Journals this month
Topics moving, week to 17 Nov 2026
By field