Volume 13, Issue 1

The phase of postgraduate medical training and practice is notoriously difficult because junior physicians or medical residents find themselves stuck in a tenuous situation of having to handle newfound heavy clinical work and responsibilities while scaling a steep learning curve on the job. In recent years, increased focus on diagnostic error has led to increasing calls to re-evaluate how clinical reasoning is cultivated in medical training, with emphasis on pedagogical interventions that aim to sharpen clinical judgments while minimising cognitive errors. Against this backdrop, we herein review the concept of “cognitive load” in post-graduate training and clinical practice, and discuss its relevance to effective metacognitive learning amidst clinical duties and to optimisation of medical decision-making in real-world settings by reducing cognitive errors in the form of bias and noise. We then outline pedagogical and workplace-based interventions that may target the twin problem of intrinsic and extrinsic cognitive load in clinical learning and work, and specifically advocate metacognitive-based practices that promote iterative cycles of cognitive schema re-calibration and professional development.

Introduction Neurodegenerative diseases such as multiple sclerosis (MS), Alzheimer’s disease (AD), and Parkinson’s disease (PD) share overlapping clinical and pathological features, complicating early diagnosis and management. Demyelination, a key pathological hallmark, underscores the importance of accurately assessing white matter (WM) integrity. Content Myelin water imaging (MWI), an advanced non-invasive MRI technique, quantifies the myelin water fraction (MWF) and offers high specificity for detecting myelin abnormalities. This systematic review explores the feasibility and diagnostic utility of MWI across MS, AD, and PD by analyzing 21 high-quality studies from major databases, following PRISMA guidelines. Summary MWI consistently revealed reduced MWF in MS patients across various WM regions, lesion types, and disease stages, including responsiveness to early treatment. In AD, MWF decline correlated with disease progression and apolipoprotein E4 (APOE4) genotype, supporting its potential in early diagnosis. Findings in PD were inconsistent, reflecting secondary or minimal myelin involvement in its pathology. Outlook MWI shows strong promise as a non-invasive imaging biomarker, particularly in MS and AD. Standardization of acquisition protocols, integration with multimodal imaging, and further longitudinal studies are essential to establish its clinical utility and support broader implementation in neurodegenerative disease diagnostics.

Diagnostic uncertainty remains a persistent challenge in improving diagnostic practice, particularly for individual physicians. This paper applies the BANI (Brittle, Anxious, Nonlinear, Incomprehensible) framework to examine the nature of uncertainty encountered by physicians. Using clinical case examples, I demonstrate how BANI manifests in the diagnostic process and propose GRACE 2 , a six-element behavioral framework designed to support physicians in managing diagnostic uncertainty. GRACE 2 integrates cognitive flexibility, adaptive reasoning, and empathic communication. Together with BANI, it offers a structured and actionable approach to enhancing physicians’ diagnostic decision-making, thereby contributing to the broader goal of improving diagnostic safety for patients.

Understanding the pathologic basis of diseases and their clinical correlates has been growing in parallel to the relevant advances in science and medicine. However, most reformed medical school curricula have mainly addressed the overall content integration with a less emphasis on explicitly structuring the integration of pathophysiology or other relevant basic sciences in clinical skills (CS) courses. Clinicopathologic Correlations (CPCOR), when effectively designed in CS courses, link the clinical findings to their related basic science fundamental changes. Regular highlighting of relevant CPCORs in CS courses enhances student acquisition of clinical reasoning skills and at the same time triggers their translational scientific curiosity. The six-step CPCOR process, detailed in the manuscript, starts with developing session learning objectives that guide CPCOR content development relevant to the weekly CS case. A typical CPCOR session includes pre-encounter and post-encounter small group activities in which students formulate broad and narrow differential diagnosis respectively. Throughout the session, students discuss risk factors, etiopathogenesis, and history and physical examination findings for the identified differential diagnoses. These small group activities are enhanced by a large group session delivered by a Pathologist-Clinician team that leads student centered CPCOR discussions relevant to the weekly CS case. In addition to the enhanced clinical reasoning skills, the implemented CPCOR process augmented the curricular emphasis of lifelong learning skills while reinforcing the importance of the pathologic basis of the clinical findings. Our streamlined CPCOR process can easily be adapted into other medical school curricula to meet relevant needs of integration and clinical reasoning enhancements.

Illness scripts are mental networks of information about medical diseases and are used by clinicians to efficiently make diagnostic and treatment decisions. The traditional components of an illness script include the epidemiologic risk factors, pathophysiology, and clinical findings of a disease [Feltovich PJ, Barrows HS. Issues of generality in medical problem solving. In: Schmidt HG, De Volder ML, editors. Tutorials in problem-based learning. Assen/Maastricht: Van Gorcum; 1984:128–42 pp]. Here, we coin the term illness script builder templates (ISBTs), which are educational tools to help learners develop their illness scripts. While the actual concept of ISBTs is not new, we aim to explicitly distinguish these discrete learning tools from illness scripts themselves, which are mental frameworks. ISBTs have significant pedagogic value. They are being more widely adopted in clinical curricula in medical school, as they have been well-received by students and educationally effective. ISBTs allow the integration of new knowledge to be intuitive, as they take advantage of our predilection toward story-based learning. Limitations to ISBTs also exist, especially when considering their optimal length and structure as well as complexity for different levels of learners. We additionally explore the specific strengths and limitations of ISBTs and propose strategies to maximize their use in education and clinical practice.

The coronavirus disease 2019 (COVID-19) pandemic has placed laboratory medicine at the forefront of public health and clinical care. Larger use of social media and official communication platforms raised public awareness of laboratory science, driving demand for rapid, accurate diagnostic information and shifting expectations around access and interpretation of testing. Laboratory medicine, rooted in accuracy, precision, reproducibility and clinical relevance, has advanced from basic diagnostics to sophisticated molecular and data-driven platforms. Yet, literature and policy on coordinated international laboratory networks, especially for surveillance and emergency response, remain limited. This opinion paper introduces the concept of “global-of-care testing”, encompassing globally connected diagnostic infrastructures with regional adaptability, robust governance, and sustained investment in technology and workforce. Laboratory network design must account for geography and population density in allocating facilities. Integrated systems require automation capable of interfacing across multiple platforms (preanalytical processing, clinical chemistry, immunochemistry, hematology, coagulation, urinalysis and even molecular diagnostics and mass spectrometry) to optimize workflows, support real-time decision-making, facilitate remote collaboration and maintain rigorous quality assurance. A decentralized yet interconnected model allows peripheral laboratories to actively participate in clinical decision-making through shared protocols, telemedicine and integrated data, ultimately reducing turnaround times, improving responsiveness and enhancing patient-centred care. Embedding Value-Based Laboratory Medicine (VBLM) within this framework ensures that diagnostics are aligned with health outcomes in a multidisciplinary ecosystem organized around patient needs. The future of laboratory medicine will hence depend on evidence-based reforms that integrate technology, reorganize systems and reinforce governance for promoting quality, equitable access and sustainable precision healthcare.

Objectives Referral documentation may either contribute to diagnostic excellence or play a role in diagnostic errors (DEs), but its exact impact remains unclear. This study investigates the association between referral documentation and DEs among patients initially evaluated by another hospital or department and subsequently referred to the general internal medicine (GIM) outpatient clinic of an acute care tertiary hospital in Japan. Methods This cross-sectional study analyzed outpatients who visited the GIM outpatient clinic between April 1, 2017 and March 31, 2023. Patients initially evaluated at another medical facility or department, who then visited the GIM outpatient clinic, and were subsequently readmitted unexpectedly within 14 days after GIM outpatient clinic visit were included. DEs were identified using the Revised Safer Dx Instrument. Errors were analyzed using the Diagnostic Error Evaluation and Research (DEER) taxonomy. Logistic regression analysis was performed to assess the relationship between referral letters and DEs. Results Of 80 patients, 29 (36.3 %) experienced DEs. Referral letters were present for 52 (65.0 %) patients. The proportion of DEs was lower in the referred patients compared to non-referred patients (25.0 vs. 57.1 %; p-value=0.004). After adjusting for age, sex, race, multimorbidity, type of previous physicians, and post-graduate year of the GIM physician, the presence of a referral letter was associated with a substantially likelihood of DEs (OR=0.20, 95 % CI: 0.06–0.62, p-value=0.005). Conclusions The presence of a referral letter facilitates accurate diagnoses while markedly reducing DEs. Healthcare systems should consider promoting the proper use of referral systems.

Objectives Internists do not feel competent in diagnosing and treating common dermatologic conditions. Teaching clinical reasoning principles in graduate medical education can improve trainees’ diagnostic accuracy, but previously published dermatology curricula did not emphasize these skills. We developed a novel curriculum applying clinical reasoning concepts to teach internal medicine (IM) residents how to describe dermatologic lesions, develop differential diagnoses, and use deliberate reflection to improve diagnostic accuracy for four common dermatologic complaints. Methods Five asynchronous, interactive 10-min online modules were developed and administered to all 152 IM residents at a large academic residency program in 2023. Residents were evaluated for their ability to describe dermatologic lesions, their diagnostic accuracy, and their deliberate reflection skills. Residents completed this novel assessment before, immediately after, and four months after the curriculum. Linear mixed effects regression models were used to assess changes in assessment scores over time. Results One hundred eleven of 152 residents (73 %) participated in the study. Total assessment scores improved between pre-test and post-test (mean difference 0.98, 95 % CI [0.32, 1.64], p=0.004), but not between pre-test and delayed post-test. Residents who completed 4 or 5 modules improved from pre-test to post-test in the description component (mean difference 0.46, 95 % CI [0.01, 0.91], p=0.043) and the final diagnosis/treatment component (mean difference 0.69, 95 % CI [0.22, 1.17] p=0.004), but not the deliberate reflection component. Conclusions An interactive, asynchronous clinical reasoning-based dermatology curriculum can improve IM resident knowledge of common dermatologic complaints, particularly immediately after participation and if most modules are completed.

Objectives During the SARS-CoV-2 pandemic, new patient evaluations in pediatric rheumatology were performed using telehealth. Given the pediatric rheumatology workforce shortage, telehealth may be a way to efficiently triage referrals. The objective was to assess the utility of telehealth visits as a diagnostic tool to accurately assess the need for in-person evaluation. Methods This was a retrospective cohort study of patients evaluated by telehealth for a new patient visit from March 1 to June 30, 2020 at a tertiary center. Electronic health record documentation from subsequent rheumatology, specialty, and primary care encounters over the subsequent 4 years were reviewed. The primary outcome was diagnostic concordance, defined as consistency in the documented diagnostic reasoning, between the initial telehealth video visit and in-person follow-up visits. Results During the study period, there were 111 telehealth visits, 80 (72 %) of which had follow-up data. 55/80 had in-person rheumatology evaluations. Only 9 % patients had discordant diagnoses, all of whom had initial concern for inflammatory arthritis during the telehealth visit but a diagnosis of a non-inflammatory condition after in-person evaluation. Nine patients with a significant rheumatic disease were identified via telehealth. There were no unplanned ED visits or hospital admissions following telehealth visits. 33 % of patients were found to not warrant rheumatologic follow-up after the telehealth visit. Conclusions For pediatric rheumatology new patient evaluations, diagnostic accuracy via telehealth evaluation was high. Providers triaged patients with chronic rheumatologic conditions for in-person evaluations and were able to accurately identify benign conditions that did not require in-person follow-up.

Objectives In the prognostic evaluation of syncopal patients, previous episodes have a somewhat ambivalent role, as they are usually associated to reflex syncope, with a benign prognosis, but according to some authors when frequent require further investigations even in case of a negative ED workup. Methods Retrospective observational study in a cohort of 378 ED syncopal patients to investigated the presence of previous episodes and their prognostic significance, in predicting cardiac syncope, syncope associated to underlying acute diseases and syncope severe due to consequential risk. Results We found previous syncopal episodes in over half of patients (57.4 %); mostly few (up to two episodes in total in 63.8 %), with a trends over time very rare or occasional, in a small minority complicated by major injuries. As regards the prognostic role of syncopal recurrences, we found: 1. no association between their total number in life with any outcome considered; 2. a close association between presence of acute principal diseases and absence of previous episodes; 3. a close association between severe syncope due to consequential risk and a frequent or recurrent time course of previous spells and previous episodes worsened by major injuries. Conclusions Previous episodes, although frequent, don’t correlate with adverse outcomes so don’t help to identify high risk patients. Conversely, a first syncopal episode, especially in elderly patients, requires careful evaluation to rule out any underlying acute condition. Additionally, attention should be also paid to patients with previous episodes clustered in the last period or complicated by severe injuries, as they negatively impact quality of life.

Objectives Interpreting skin findings can be challenging for both laypersons and clinicians. Large language models (LLMs) offer accessible decision support, yet their diagnostic capabilities for dermatological images remain underexplored. This study evaluated the diagnostic performance of LLMs based on image interpretation of common dermatological diseases. Methods A total of 500 dermatological images, encompassing four prevalent skin conditions (psoriasis, vitiligo, erysipelas and rosacea), were used to compare seven multimodal LLMs (GPT-4o, GPT-4o mini, Gemini 1.5 Pro, Gemini 1.5 Flash, Claude 3.5 Sonnet, Llama3.2 90B and 11B). A standardized prompt was used to generate one top diagnosis. Results The highest overall accuracy was achieved by GPT-4o (67.8 %), followed by GPT-4o mini (63.8 %) and Llama3.2 11B (61.4 %). Accuracy varied considerably across conditions, with psoriasis with the highest mean LLM accuracy of 59.2 % and erysipelas demonstrating the lowest accuracy (33.4 %). 11.0 % of all images were misdiagnosed by all LLMs, whereas 11.6 % were correctly diagnosed by all models. Correct diagnoses by all LLMs were linked to clear, disease-specific features, such as sharply demarcated erythematous plaques in psoriasis. Llama3.2 90B was the only LLM to decline diagnosing images, particularly those involving intimate areas of the body. Conclusions LLM performance varied significantly, emphasizing the need for cautious usage. Notably, a free, locally hostable model correctly identified the top diagnosis for approximately two-thirds of all images, demonstrating the potential for safer, locally deployed LLMs. Advancements in model accuracy and the integration of clinical metadata could further enhance accessible and reliable clinical decision support systems.

Objectives The challenges posed by questionable journals to academia are very real, and being able to detect hijacked journals would be valuable to the research community. Using an artificial intelligence (AI) chatbot may be a promising approach to early detection. The purpose of this research is to analyze and benchmark the performance of different AI chatbots in identifying hijacked medical journals. Methods This study utilized a dataset comprising 21 previously identified hijacked journals and 10 newly detected hijacked journals, alongside their respective legitimate versions. ChatGPT, Gemini, Copilot, DeepSeek, Qwen, Perplexity, and Claude were selected for benchmarking. Three question types were developed to assess AI chatbots’ performance in providing information about hijacked journals, identifying hijacked websites, and verifying legitimate ones. Results The results show that current AI chatbots can provide general information about hijacked journals, but cannot reliably identify either real or hijacked journal titles. While Copilot performed better than others, it was not error-free. Conclusions Current AI chatbots are not yet reliable for detecting hijacked journals and may inadvertently promote them.

Objectives The aim was to reduce blood loss related to laboratory blood sampling by decreasing the discard blood volume from the central venous catheter (CVC) from 10 to 3 mL, and replacing the 3 mL tube for complete blood count (CBC) with a 2 mL tube. A comparison study was performed, and after introduction of this modified protocol the decrease in withdrawn blood was calculated. Methods In 52 adult inpatients at the ICU 3 mL of blood was discarded from the CVC, and blood was drawn in a 5 mL serum tube and a 2 mL EDTA tube (test samples), followed by subsequent sampling for routine laboratory testing using the same blood tubes (routine samples). In paired test and routine samples, clinical chemistry testing and CBC were performed. Results High correlations (ρ>0.90) were obtained for the majority of parameters. Small constant difference was obtained for hemoglobin. Small significant biases were observed for hemoglobin, MCHC, albumin, ALT, calcium, chlorides, MCV and glucose, however, being below the biases of the analytical methods. No statistically significant differences between test and routine samples were found for any of the assessed tests (p>0.05). Implementation of the 3 mL discard blood volume protocol and transition to a smaller volume tube for CBC reduced the amount of withdrawn blood by almost half (46 %). Conclusions The 3 mL discard blood volume from the CVC can be safely used for routine hematology and clinical chemistry testing. The dual approach assessed herein can contribute to reduction of iatrogenic blood loss.

Objectives The aim of this study is to assess the discrepancies between potassium values obtained from point-of-care testing and laboratory analyzers, and to propose a straightforward method to indirectly assess the likelihood of hemolysis in whole blood samples. Methods The data were collected from 409 hospitalized patients (199 males, 210 females; median age 78 years, interquartile range 55–87) in whom were simultaneously requested the blood gas profile, including electrolytes (Na, K, Cl and Ca 2+ ) on the GEM Premier 5000 blood gas analyzer and K in plasma samples on the Cobas analytical system. Results The Wilcoxon test showed a significant difference (p<0.0001) for K measurements, with a number of positive differences between Cobas and GEM Premier 5,000 of 110 and negative differences of 293. The median value of the hemolysis index was 7 (95 % CI for the median 6 to 8). The differences between the two K measurements obtained from plasma and whole blood samples showed a lowest and highest differences of −1.7 and 4.6 mmol/L. The median difference was 0.63 (95 % CI for the median: 0.54–0.73) and the interquartile range (IQR) was 0.48–0.9. Overall, 89 samples (21 %) displayed a difference larger than the relative change value. Conclusions The presence of hemolysis in plasma samples should alert laboratory professionals to carefully evaluate the corresponding whole blood results. Concordant values indicate the presence of hemolysis in the whole blood sample, while discordant values suggest that hemolysis only affects one of the two samples.

Objectives Hereditary transthyretin amyloidosis with polyneuropathy (ATTRv-PN) is a rare, progressive neurodegenerative disorder caused by mutations in the transthyretin ( TTR ) gene. The disease leads to systemic amyloid deposition, primarily affecting the nervous system and, in some cases, the heart. Early diagnosis and monitoring are critical for effective management, yet reliable biomarkers remain limited. This study aimed to investigate the role of serum glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) as biomarkers in ATTRv-PN. Methods A retrospective observational study was conducted at the University Hospital Paolo Giaccone, enrolling ATTRv-PN patients, asymptomatic TTR mutation carriers, and healthy blood donors. Serum GFAP and NfL levels were measured using a fully automated immunoassay (Lumipulse G1200). Results A total of 119 participants were included: ATTRv-PN (n=23), carriers (n=27), and healthy controls (n=69). GFAP levels were significantly elevated in ATTRv-PN patients compared to carriers and healthy controls (p<0.001), with the highest levels observed in individuals with the V122I mutation. The median NfL levels were also significantly elevated in ATTRv-PN patients (30.74 pg/mL) compared to carriers (11.59 pg/mL) and healthy controls (12.86 pg/mL) (p<0.001). Additionally, a significant negative correlation was observed between NfL levels and clinical severity scores, indicating its association with disease severity. Conclusions These findings support the usefulness of serum NfL as a prognostic tool in ATTRv-PN and highlight the potential involvement of astrocyte activation in disease pathology. Further longitudinal studies are needed to validate these biomarkers for clinical application.

The pursuit of medical diagnosis has long been shaped by an epistemic framework that assumes diseases have inherent, discoverable essences. This essentialist approach, deeply rooted in Aristotelian thought, has historically guided diagnostic reasoning and classification for over a century. However, the rise of artificial intelligence (AI) is catalyzing a philosophical and practical shift toward nominalism – a framework in which diagnoses are derived from dynamic, data-driven pattern recognition rather than fixed disease categories. This transition, if it occurs, would be revolutionary, exposing core limitations of essentialist thinking and reframing diagnosis as a process rather than a static conclusion. In doing so, it challenges the conventional concept of an ‘endpoint diagnosis’ – the idea that diseases can be definitively and completely categorized. Instead, diagnosis emerges as a contingent narrative point within broader clinical trajectories, calling for a reimagining of diagnostic reasoning in the AI era.

Objectives Diagnostic errors are a significant source of patient harm, often arising from cognitive biases and communication breakdowns. Diagnostic timeouts (DTOs) offer a structured opportunity to reassess diagnoses, but their implementation is inconsistent. This study aimed to evaluate a nurse-driven DTO framework designed to enhance interprofessional communication and diagnostic accuracy in a pediatric hospital setting. Methods A multidisciplinary committee developed and piloted a DTO framework on a 48-bed pediatric acute care unit. Bedside nurses received structured education on DTOs, and standardized prompts were integrated into nursing workflows to identify patients who may benefit from a DTO. Feasibility and acceptability were assessed through voluntary surveys, post-DTO documentation, and qualitative feedback. Results Among 90 surveyed nurses, 28 % reported that a DTO would have been beneficial for at least one patient in the eight weeks prior to the intervention period. Over six months, 13 nurse-initiated DTOs occurred, with nearly half (46 %) prompting additional workup. Two DTOs led to escalation of care. Qualitative feedback highlighted improved communication, increased psychological safety, and enhanced teamwork. Conclusions Pilot implementation of a formalized nurse-driven DTO framework was well-received, reinforcing its role in structured diagnostic reassessment. Future efforts will expand DTO implementation and evaluate its impact on psychological safety.

Objectives The misdiagnosis of appendicitis remains frequent. Better understanding of its clinical evolution over time would decrease the incidence of misdiagnosis. Case presentation At the Society to Improve Diagnosis in medicine conference in Cleveland Ohio in October 2023, Alice Tapper and her father, CNN journalist Jake Tapper, presented her case of misdiagnosed appendicitis. Thanks to meticulous notes and a detailed timeline, the case vividly demonstrates the protean nature of appendicitis as well as the cognitive pitfalls of clinicians who treat it. Conclusions McBurney's point tenderness is over-emphasized as the key to the diagnosis of appendicitis.

Objectives We will explore the diagnostic similarities of spindle cell neoplasms and the attributed heuristics that lead to misdiagnosis biases. The biases explored will include availability bias and anchoring bias, with a discussion on the events leading to their formation. Case presentation A 58-year-old African American male with a past medical history of well-controlled HIV presented to the dermatology clinic for a two-year history of several persistent skin nodules on his lower legs. One lesion on his left lateral calf, a 1.5 cm dome-shaped nodule with a centralized keratinous plug, was suspicious for squamous cell carcinoma (SCC), prompting a shave biopsy. The dermatopathology report identified the lesion as dermatofibrosarcoma protuberans (DFSP) with CD34 positivity and the patient was referred for Mohs Micrographic Surgery. Frozen sections during Mohs surgery revealed concern for an alternative diagnosis, which was then confirmed as Kaposi Sarcoma. Conclusions This case highlights the susceptibility of dermatology to misdiagnosis. Availability bias in the clinical setting led to an inadequate biopsy method. Further anchoring bias then potentially influenced histologic interpretation and management decisions. Insufficient appreciation of Kaposi Sarcoma development in the setting of well-controlled HIV also further influenced the diagnosis rendered. Mohs Surgery evaluation allowed for de-biased clinical and histologic assessment, correcting diagnosis. Several overlying factors, such as time pressures, knowledge gaps, and technique limitations, create a reliance on cognitive heuristics. Recognizing these external pressures can help clinicians enhance diagnostic accuracy by systematically considering alternative diagnoses.