National Healthcare Group

Abstract Background Large language models such as ChatGPT have rapidly become popular tools for seeking health information, with surveys indicating that nearly half of consumers use generative AI for health-related inquiries. Despite documented accuracy limitations and hallucination risks, patients increasingly consult these tools when making healthcare decisions, effectively using them as supplementary or alternative sources of health information alongside or in lieu of traditional medical consultation. Objectives To systematically identify, appraise, and synthesise qualitative research exploring the lived experiences of patients who use large language models to inform healthcare decision-making, examining motivations, information-seeking behaviours, trust and scepticism, perceived impacts on healthcare decisions, and the meaning patients ascribe to AI-generated health information. Methods This protocol follows PRISMA-P and ENTREQ guidelines. Systematic searches will be conducted across MEDLINE, PsycINFO, CINAHL, Web of Science, ACM Digital Library, and Scopus from November 2022 through December 2025. Studies employing qualitative methods to explore patient experiences with LLM-based health information seeking will be included. Study selection and quality assessment using the CASP Qualitative Checklist will be conducted independently by two reviewers using Covidence. Data will be synthesised using thematic synthesis following Thomas and Harden’s approach. Discussion This protocol establishes a rigorous framework for synthesising qualitative evidence on an emergent healthcare phenomenon with significant implications for patient safety, health literacy, shared decision-making, and the evolving patient-provider relationship in the age of generative AI.

Clinical neurology is predicated on deciphering the intricate functions and dysfunctions of the nervous system. This educational review explores the fundamental principles of localization and diagnostic reasoning that are essential for neurological practice. We frame these principles within a philosophical context, referencing Plato's Allegory of the Cave to demonstrate the fundamentally indirect nature of neurological assessment, wherein clinical signs and symptoms reflect underlying neuronal processes. The application of the scientific method, balancing reductionist and contextualist approaches, is discussed. Bayes' Theorem is presented as a framework for probabilistic clinical reasoning. We examine the interplay between intuitive pattern recognition, analytical reasoning, and metacognition in developing diagnostic acumen. A systematic model (Symptoms → Syndrome → Anatomy → Pathology → Etiology) is described as a cornerstone for avoiding errors. Practical applications of localization are illustrated through common clinical scenarios. Finally, we highlight cognitive biases that contribute to diagnostic errors and propose strategies for mitigation.

Major procedural complications are an unavoidable feature of interventional cardiology and often represent pivotal inflection points in an operator's professional development. Beyond their immediate clinical consequences, these events can profoundly influence confidence, judgment, risk tolerance, and subsequent technical growth. Whether complications stimulate learning or precipitate persistent distress depends largely on how they are interpreted, discussed, and supported. This review examines the emotional and cognitive responses of interventional cardiologists following major procedural complications. We explore how cognitive bias, cognitive heuristics, perfectionism, moral injury, and threats to professional identity shape recovery, decision-making, and future procedural behavior. These forces may foster adaptive learning or drive maladaptive responses, including rumination, self-doubt, withdrawal, and risk aversion. Drawing on evidence from cognitive psychology, aviation safety, and established peer-support frameworks, we outline practical strategies to transform complications into opportunities for emotional and technical growth. Institutional approaches include psychologically safe debriefing, process-focused morbidity and mortality conferences, peer-support programs, leadership development, and accessible mental health resources. Individual strategies such as internal validation, narrative reframing, patient reconnection, mentorship, and structured reflective practice help restore confidence, preserve moral integrity, and reinforce professional identity by explicitly separating decision quality from outcomes. Supporting operator recovery after procedural complications is not ancillary to patient care; it is foundational to sustained technical excellence, sound clinical judgment, and compassionate practice. Integrating structured, stigma-free recovery pathways into interventional training and institutional culture should be recognized as a core component of patient safety and professional sustainability.

Background:Generative artificial intelligence (GenAI) is increasingly used to draft multiple-choice questions (MCQs) for health professions education, but much evidence concerns raw model outputs, expert ratings, or item difficulty alone. Educators edit GenAI drafts before use, and whether such items are psychometrically ready for postgraduate assessment remains unclear. Objective:This study aimed to compare human-edited GenAI-assisted and educator-crafted MCQs for postgraduate Family Medicine Applied Knowledge Test-level assessment, examining difficulty, discrimination, reliability, distractor functioning, and participant perceptions. Methods:We conducted a blinded cross-sectional, within-participant comparative psychometric evaluation in Singapore. Sixty best-of-five single-best-answer MCQs were evaluated, 30 human-edited GenAI-assisted items and 30 educator-crafted items, topic-matched across postgraduate FM domains and randomized across 2 assessment sets. Eligible participants were postgraduate doctors enrolled in FM residency or postgraduate family medicine programs, preparing for the Applied Knowledge Test, and blinded to item origin; incomplete paired responses were excluded. Outcomes included paired total scores, score correlation and agreement, Kuder-Richardson Formula 20 reliability, item difficulty index, corrected point-biserial discrimination, distractor functioning, and perceived difficulty, clarity, and relevance. Analyses used paired-sample tests, Pearson correlation, Fisher exact tests, and item-level psychometric statistics, with α=.05 and Bonferroni correction within comparison families. Results:Of 74 participants, 73 completed both item sets and were included in the analysis. The final sample comprised 36 graduate diploma in FM trainees, 5 MMed FM trainees, and 32 FM residents. Paired-sample testing showed lower scores on GenAI-assisted than educator-crafted items (mean 19.12, SD 2.83 vs mean 21.10, SD 3.42 out of 30; mean difference -1.97, 95% CI -2.72 to -1.23; P<.001; Cohen d=0.62), indicating that GenAI-assisted items were not easier. Scores were positively correlated (r=0.49, 95% CI 0.30-0.64; P<.001), but Bland-Altman analysis indicated limited agreement. Kuder-Richardson Formula 20 reliability was lower for GenAI-assisted items (0.38 vs 0.60). Mean difficulty index did not differ significantly (0.64 vs 0.70; mean difference -0.07, 95% CI -0.19 to 0.06; P=.29), and more GenAI-assisted items fell within the acceptable difficulty range (18/30, 60.0% vs 13/30, 43.3%). However, mean corrected point-biserial discrimination was lower for GenAI-assisted items (0.09 vs 0.18; mean difference -0.08, 95% CI -0.16 to -0.01; P=.04), and negative discrimination was more common (6/30, 20% vs 3/30, 10%). GenAI-assisted items also had more nonfunctioning and negatively discriminating distractors, although these differences were not statistically significant. Participant ratings of perceived difficulty, clarity, and practice relevance did not differ by origin. Conclusions:Human-edited GenAI-assisted MCQs can achieve plausible difficulty, but difficulty and surface acceptability did not ensure assessment readiness. Using trainee response data, this study extends work on raw outputs or expert opinion. GenAI should be used as a drafting adjunct within educator-led workflows prioritizing key verification, distractor engineering, pilot testing, empirical item analysis, and repair before item-bank or summative use.