Cognitive mechanisms underlying anchoring bias in diagnosis: a randomized controlled experiment.

Diagnostic errors have been attributed to failure to sufficiently reflect on initial diagnoses. However, evidence of the benefits of reflection is conflicting. This study examined whether reflection upon initial diagnoses on difficult cases improved diagnostic accuracy and whether reflection triggered by confrontation with case evidence was more beneficial than simply revising initial diagnoses. Participants were physicians in Bern, Switzerland, registered for the 2018 Swiss internal medicine certification exam. They diagnosed written clinical cases, providing an initial diagnosis by following the same instructions and returning to the case to provide a final diagnosis. The latter required different types of reflection depending on the physician's experimental condition: return without instructions, identify confirmatory evidence, identify contradictory evidence, or identify both confirmatory and contradictory evidence. The authors examined diagnostic accuracy scores (range 0-1) as a function of diagnostic phase and reflection type. One hundred and sixty-seven physicians participated. Diagnostic accuracy scores did not significantly differ between the 4 groups of physicians in the initial (I) or the final (F) diagnostic phase (mean [95% CI]: return without instructions, I: 0.21 [0.17, 0.26], F: 0.23 [0.18, 0.28]; confirmatory evidence, I: 0.24 [0.19, 0.29], F: 0.31 [0.25, 0.37]; contradictory evidence, I: 0.22 [0.17, 0.26], F: 0.26 [0.22, 0.30]; confirmatory and contradictory evidence, I: 0.19 [0.15, 0.23], F: 0.25 [0.20, 0.31]). Regardless of type of reflection employed while revising the case, accuracy increased significantly between initial and final diagnosis, I: 0.22 (0.19, 0.24) vs F: 0.26 (0.24, 0.29); P < .001. Physicians' diagnostic accuracy improved after reflecting upon initial diagnoses provided for difficult cases, independently of the evidence searched for while reflecting. The findings support the importance attributed to reflection in clinical teaching. Future research should investigate whether revising the case can become more beneficial by triggering additional reflection.

While diagnosing problems, physicians frequently switch from nonanalytical to reflective reasoning. The conditions inducing doctors to reflect are largely unknown. The authors investigated whether a shift to reflection occurs when physicians perceive a case as problematic, and its effects on diagnostic accuracy. The authors conducted two within-subjects experiments in Brazilian teaching hospitals in 2007. In Experiment 1, 20 medical residents diagnosed the same 10 clinical cases under two experimental conditions: a nonproblematic versus a problematic context. (The latter was created by informing participants that other physicians failed to diagnose the case previously.) In addition, participants judged whether a set of medical concepts were related to the case, and response time was measured. In Experiment 2, 18 residents diagnosed two cases while thinking aloud. The authors hypothesized that a case perceived as problematic would trigger reflection, leading to higher diagnostic accuracy, lower response times for recognizing concepts (Experiment 1), more time for diagnosing, and more elaborate think-aloud protocols (Experiment 2). Experiment 1: Accuracy of diagnosis was significantly higher within the problematic context, and participants were faster in deciding whether concepts were related to the case. The same cases were evaluated as more complex and less frequently seen. Experiment 2: Time spent on diagnosis, memory for case findings, and inferences derived from the cases were significantly higher within the problematic context. A context perceived as problematic induced reflection in the participating clinicians, as indicated by lower response times, more time spent on diagnosis, and more elaborate protocols. Reflective reasoning comprised more careful analysis of findings and alternative diagnoses, and increased diagnostic accuracy.

Experienced clinicians derive many diagnoses intuitively, because most new problems they see closely resemble problems they've seen before. The majority of these diagnoses, but not all, will be correct. This study determined whether further reflection regarding initial diagnoses improves diagnostic accuracy during a high-stakes board exam, a model for studying clinical decision making. Keystroke response data were used from 500 residents who took the 2010 American Board of Internal Medicine (ABIM) Internal Medicine Certification Examination. Data included time to initial response on each question, whether the answer was correct, and whether or not the resident changed her or his initial response. The focus was on 80 diagnosis questions that comprised realistic clinical vignettes with multiple-choice single-best answers. Cognitive skill (ability) was measured using overall exam scores. Case complexity was determined using item difficulty (proportion of examinees that correctly answered the question). A hierarchical generalized linear model was used to assess the relationship between time spent on initial responses and the probability of correctly answering the questions. On average, residents changed their responses on 12% of all diagnosis questions (or 9.6 questions out of 80). Changing an answer from incorrect to correct was almost twice as likely as changing an answer from correct to incorrect. The relationship between response time and accuracy was complex. Further reflection appears to be beneficial to diagnostic accuracy, especially for more complex cases.

Introduction: Occupational Accidents with Exposure to Biological Material (OAEBM) represent a significant threat to the health of healthcare professionals and are a global public health concern. Studies have shown that medical residents, due to their training phase and involvement in invasive procedures, are particularly vulnerable to needlestick and sharps injuries, exacerbated by hospital-related stress.The objective of the study was to analyze the prevalence of percutaneous injuries with the risk of exposure to biological material among medical residents in Brazil from 2020 to 2024. Methodology: This is a cross-sectional study with epidemiological data regarding the prevalence of needlestick and sharps injuries among medical residents in Brazil, using data from the Notifiable Diseases Information System (SINAN) for the period from 2020 to 2024. Results: Surgical procedures consistently represented the primary cause of occupational accidents related to exposure to biological material. In 2020, these procedures accounted for 61% of cases. In 2021, there was a slight reduction to 56%, but surgical procedures remained the main cause. In 2022, the percentage rose to 58%, remained constant in 2023, and showed a slight reduction of 1% in 2024. Despite the minor annual variation, surgical procedures continued to be the leading cause of such accidents. Conclusion: The data highlight the importance of reinforcing preventive measures, especially for medical residents due to their limited practical experience. Regular training, adherence to safety protocols, and proper use of personal protective equipment (PPE) are crucial. Continuous data analysis helps identify and mitigate emerging risks.

52 Objective: The diagnosis of ischemic stroke subtype is difficult in the emergency setting when based only upon a non-contrast CT scan and clinical findings. Accurate diagnosis may be important because prognosis depends upon the size and location of the infarct and emergency therapeutic decisions rest upon attempts to improve prognosis. Diffusion/perfusion weighted MR identifies acute ischemia and ischemic injury but is expensive, and not consistently accessible nationwide. We investigated the ability of a readily available CT contrast study (CT angiography, CTA; and whole brain CT perfusion, CTP) to enhance diagnostic accuracy of stroke subtype. Methods: All patients (1/97–12/98) who received a CTA within 24 hours of stroke onset (mean=4.6 hrs) and a follow-up CT or MRI within 2 weeks were analyzed (N=40). Stroke neurologists made stroke subtype diagnoses based upon: 1) non contrast CT and clinical vignette; 2) #1 and CTA; 2) #1, #2 and CTP. Diagnostic accuracy at each sequential step was measured against the gold standard based upon all available clinical, lab and follow-up imaging information. Results: The addition of the contrast CT study (CT+CTP)led to a statistically significant, relative improvement in accuracy of 1)infarct localization, 100%; 2) involved vascular territory, 69%; 3) occluded vessel, 94%; 4) TOAST stroke subtype, 44%; and 5) Oxfordshire stroke subtype, 81%. CTA led to significant improvement in diagnostic accuracy of vessel occlusion and TOAST subtype. CTP led to significant improvement in diagnostic accuracy of infarct localization, vascular territory and Oxfordshire classification. Conclusion: The addition of a contrast CT study to evaluate the intracranial vessels (CTA) and whole brain perfusion (CTP) enables highly accurate diagnosis of stroke subtype in the emergency setting. The ability of this widely accessible, emergency neuroimaging technique to predict functional outcome and guide therapeutic decisions can now be investigated.

Does GPT-4 have neurophobia? Localization and diagnostic accuracy of an artificial intelligence-powered chatbot in clinical vignettes

BackgroundPrimary health care (PHC) doctors’ numbers are dwindling in high- as well as low-income countries, which is feared to hamper the achievement of Universal Health Coverage goals. As a large proportion of doctors are privately educated and private medical schools are becoming increasingly common in middle-income settings, there is a debate on whether private education represents a suitable mean to increase the supply of PHC physicians. We analyse the intentions to practice of medical residents in Brazil to understand whether these differ for public and private schools.MethodsDrawing from the literature on the selection of medical specialties, we constructed a model for the determinants of medical students’ intentions to practice in PHC, and used secondary data from a nationally representative sample of 4601 medical residents in Brazil to populate it. Multivariate analysis and multilevel cluster models were employed to explore the association between perspective physicians’ choice of practice and types of schools attended, socio-economic characteristics, and their values and opinions on the profession.ResultsOnly 3.7% of residents in our sample declared an intention to practice in PHC, with no significant association with the public or private nature of the medical schools attended. Instead, having attended a state secondary school (p = 0.028), having trained outside Brazil’s wealthy South East (p < 0.001), not coming from an affluent family (p = 0.037), and not having a high valuation of career development opportunities (p < 0.001) were predictors of willingness to practice in PHC. A low consideration for quality of life, for opportunities for treating patients, and for the liberal aspects of the profession were also associated with future physicians’ intentions to work in primary care (all p < 0.001).ConclusionsIn Brazil, training in public or private medical schools does not influence the intention to practice in PHC. But students from affluent backgrounds, with private secondary education, and graduating in the rich South East were found to be overrepresented in both types of training institutions, and this is what appears to negatively impact the selection of PHC careers. With a view to increasing the supply of PHC practitioners in middle-income countries, policies should focus on opening medical schools in rural areas and improving access for students from disadvantaged backgrounds.

BackgroundRecent changes in medical education calls for a shift toward student-centered learning. Therefore, it is imperative that clinical educators transparently assess the work-readiness of their medical residents through entrustment-based supervision decisions toward independent practice. Similarly, it is critical that medical residents are vocal about the quality of supervision and feedback they receive. This study aimed to explore the factors that influence entrustment-based supervision decisions and feedback receptivity by establishing a general consensus among Taiwanese clinical educators and medical residents regarding entrustment decisions and feedback uptake, respectively.MethodsIn Q-methodology studies, a set of opinion statement (i.e., the Q-sample) is generated to represent the phenomenon of interest. To explore the factors that influence entrustment-based supervision decisions and feedback receptivity, a Q-sample was developed using a four-step approach: (1) literature search using electronic databases, such as PubMed and Google Scholar, and interviews with emergency clinical educators and medical residents to generate opinion statements, (2) thematic analysis and grouping using The Model of Trust, the Ready, Wiling, and Able model, and the theory of self-regulated learning, (3) translation, and (4) application of a Delphi technique, including two expert panels comprised of clinical educators and medical residents, to establish a consensus of the statements and validation for a subsequent Q-study.ResultsA total of 585 and 1,039 statements from the literature search and interviews were extracted to populate the sample of statements (i.e., the concourse) regarding entrustment-based supervision decisions for clinical educators and feedback receptivity emergency medicine residents, respectively. Two expert panels were invited to participate in a Delphi Technique, comprised of 11 clinical educators and 13 medical residents. After two-rounds of a Delphi technique, the panel of clinical educators agreed on 54 statements on factors that influence entrustment-based supervision decisions and were categorized into five themes defined by the Model of Trust. Similarly, a total of 60 statements on the factors that influence feedback receptivity were retained by the panel of medical residents and were categorized into five themes defined by the Ready, Willing, and Able model and the theory of self-regulated learning.ConclusionThough not exhaustive, the key factors agreed upon by clinical educators and medical residents reflect the characteristics of entrustment-based supervision decisions and feedback receptivity across specialties. This study provides insight on an often overlooked issue of the paths to teaching and learning in competency-based residency training programs. Additionally, incorporation of the Delphi technique further adds to the existing literature and puts emphasis as an important tool that can be used in medical education to rigorously validate Q-statements and develop Q-samples in various specialties.

The diagnostic accuracy of differential diagnoses generated by artificial intelligence chatbots, including ChatGPT models, for complex clinical vignettes derived from general internal medicine (GIM) department case reports is unknown. This study aims to evaluate the accuracy of the differential diagnosis lists generated by both third-generation ChatGPT (ChatGPT-3.5) and fourth-generation ChatGPT (ChatGPT-4) by using case vignettes from case reports published by the Department of GIM of Dokkyo Medical University Hospital, Japan. We searched PubMed for case reports. Upon identification, physicians selected diagnostic cases, determined the final diagnosis, and displayed them into clinical vignettes. Physicians typed the determined text with the clinical vignettes in the ChatGPT-3.5 and ChatGPT-4 prompts to generate the top 10 differential diagnoses. The ChatGPT models were not specially trained or further reinforced for this task. Three GIM physicians from other medical institutions created differential diagnosis lists by reading the same clinical vignettes. We measured the rate of correct diagnosis within the top 10 differential diagnosis lists, top 5 differential diagnosis lists, and the top diagnosis. In total, 52 case reports were analyzed. The rates of correct diagnosis by ChatGPT-4 within the top 10 differential diagnosis lists, top 5 differential diagnosis lists, and top diagnosis were 83% (43/52), 81% (42/52), and 60% (31/52), respectively. The rates of correct diagnosis by ChatGPT-3.5 within the top 10 differential diagnosis lists, top 5 differential diagnosis lists, and top diagnosis were 73% (38/52), 65% (34/52), and 42% (22/52), respectively. The rates of correct diagnosis by ChatGPT-4 were comparable to those by physicians within the top 10 (43/52, 83% vs 39/52, 75%, respectively; P=.47) and within the top 5 (42/52, 81% vs 35/52, 67%, respectively; P=.18) differential diagnosis lists and top diagnosis (31/52, 60% vs 26/52, 50%, respectively; P=.43) although the difference was not significant. The ChatGPT models' diagnostic accuracy did not significantly vary based on open access status or the publication date (before 2011 vs 2022). This study demonstrates the potential diagnostic accuracy of differential diagnosis lists generated using ChatGPT-3.5 and ChatGPT-4 for complex clinical vignettes from case reports published by the GIM department. The rate of correct diagnoses within the top 10 and top 5 differential diagnosis lists generated by ChatGPT-4 exceeds 80%. Although derived from a limited data set of case reports from a single department, our findings highlight the potential utility of ChatGPT-4 as a supplementary tool for physicians, particularly for those affiliated with the GIM department. Further investigations should explore the diagnostic accuracy of ChatGPT by using distinct case materials beyond its training data. Such efforts will provide a comprehensive insight into the role of artificial intelligence in enhancing clinical decision-making.

52 Objective: The diagnosis of ischemic stroke subtype is difficult in the emergency setting when based only upon a non-contrast CT scan and clinical findings. Accurate diagnosis may be important because prognosis depends upon the size and location of the infarct and emergency therapeutic decisions rest upon attempts to improve prognosis. Diffusion/perfusion weighted MR identifies acute ischemia and ischemic injury but is expensive, and not consistently accessible nationwide. We investigated the ability of a readily available CT contrast study (CT angiography, CTA; and whole brain CT perfusion, CTP) to enhance diagnostic accuracy of stroke subtype. Methods: All patients (1/97–12/98) who received a CTA within 24 hours of stroke onset (mean=4.6 hrs) and a follow-up CT or MRI within 2 weeks were analyzed (N=40). Stroke neurologists made stroke subtype diagnoses based upon: 1) non contrast CT and clinical vignette; 2) #1 and CTA; 2) #1, #2 and CTP. Diagnostic accuracy at each sequential step was measured against the gold standard based upon all available clinical, lab and follow-up imaging information. Results: The addition of the contrast CT study (CT+CTP)led to a statistically significant, relative improvement in accuracy of 1)infarct localization, 100%; 2) involved vascular territory, 69%; 3) occluded vessel, 94%; 4) TOAST stroke subtype, 44%; and 5) Oxfordshire stroke subtype, 81%. CTA led to significant improvement in diagnostic accuracy of vessel occlusion and TOAST subtype. CTP led to significant improvement in diagnostic accuracy of infarct localization, vascular territory and Oxfordshire classification. Conclusion: The addition of a contrast CT study to evaluate the intracranial vessels (CTA) and whole brain perfusion (CTP) enables highly accurate diagnosis of stroke subtype in the emergency setting. The ability of this widely accessible, emergency neuroimaging technique to predict functional outcome and guide therapeutic decisions can now be investigated.

Online CKD Education for Medical Students, Residents, and Fellows: Training in a New Era

Rare diseases, which affect millions of people worldwide, pose a major challenge, as it often takes years before an accurate diagnosis can be made. This delay results in substantial burdens for patients and health care systems, as misdiagnoses lead to inadequate treatment and increased costs. Artificial intelligence (AI)-powered symptom checkers (SCs) present an opportunity to flag rare diseases earlier in the diagnostic work-up. However, these tools are primarily based on published literature, which often contains incomplete data on rare diseases, resulting in compromised diagnostic accuracy. Integrating expert interview insights into SC models may enhance their performance, ensuring that rare diseases are considered sooner and diagnosed more accurately. The objectives of our study were to incorporate expert interview vignettes into AI-powered SCs, in addition to a traditional literature review, and to evaluate whether this novel approach improves diagnostic accuracy and user satisfaction for rare diseases, focusing on Fabry disease. This mixed methods prospective pilot study was conducted at Hannover Medical School, Germany. In the first phase, guided interviews were conducted with medical experts specialized in Fabry disease to create clinical vignettes that enriched the AI SC's Fabry disease model. In the second phase, adult patients with a confirmed diagnosis of Fabry disease used both the original and optimized SC versions in a randomized order. The versions, containing either the original or the optimized Fabry disease model, were evaluated based on diagnostic accuracy and user satisfaction, which were assessed through questionnaires. Three medical experts with extensive experience in lysosomal storage disorder Fabry disease contributed to the creation of 5 clinical vignettes, which were integrated into the AI-powered SC. The study compared the original and optimized SC versions in 6 patients with Fabry disease. The optimized version improved diagnostic accuracy, with Fabry disease identified as the top suggestion in 33% (2/6) of cases, compared to 17% (1/6) with the original model. Additionally, overall user satisfaction was higher for the optimized version, with participants rating it more favorably in terms of symptom coverage and completeness. This study demonstrates that integrating expert-derived clinical vignettes into AI-powered SCs can improve diagnostic accuracy and user satisfaction, particularly for rare diseases. The optimized SC version, which incorporated these vignettes, showed improved performance in identifying Fabry disease as a top diagnostic suggestion and received higher user satisfaction ratings compared to the original version. To fully realize the potential of this approach, it is crucial to include vignettes representing atypical presentations and to conduct larger-scale studies to validate these findings.

During the care of incapacitated patients, physicians, and medical residents discuss treatment options and gain consent to treat through healthcare surrogates. The purpose of this study is to ascertain medical residents' knowledge of healthcare consent laws, application during clinical practice, and appraise the education residents received regarding surrogate decision making laws. Beginning in February of 2018, 35 of 113 medical residents working with patients within Indiana completed a survey. The survey explored medical residents' knowledge of health care surrogate consent laws utilized in Indiana hospitals and Veterans Affairs (VA) hospitals via clinical vignettes. Only 22.9% of medical residents knew the default state law in Indiana did not have a hierarchy for settling disputes among surrogates. Medical residents correctly identified which family members could participate in medical decisions 86% of the time. Under the Veterans Affairs surrogate law, medical residents correctly identified appropriate family members or friends 50% of the time and incorrectly acknowledged the chief decision makers during a dispute 30% of the time. All medical residents report only having little or some knowledge of surrogate decision making laws with only 43% having remembered receiving surrogate decision making training during their residency. These findings demonstrate that medical residents lack understanding of surrogate decision making laws. In order to ensure medical decisions are made by the appropriate surrogates and patient autonomy is upheld, an educational intervention is required to train medical residents about surrogate decision making laws and how they are used in clinical practice.

Neural networks applied to the oculomotor system (D.A. Robinson). Eye movements in neurological diagnosis: Huntington's disease and congenital nystagmus (D.S. Zee). Neurophysiology of Eye Movements. Localization of targets in humans: the role of ocular muscle proprioception (D. Nommay et al). Effects of severance of the vestibular commissural pathway on the neural integrator of the oculomotor system in cat (E. Godaux, G. Cheron). Representation of three-dimensional eye movements in the cerebellar flocculus of the rabbit (J. van der Steen et al). Effects of floccular injections of noradrenergic agonists and antagonists on adaptive changes in the VOR gain (J. van Neerven et al). The Control of Eye Movements. Human vestibuloocular reflex (VOR) and vestibular velocity storage as influenced by optokinetic stimuli (E. Koenig et al). Vertical gaze stability in cat: otolithic contribution (V.E. Pettorossi et al). Post-rotational nystagmus suppression by the presentation of a single visual target in cat and man (G. Magenes et al). Adaptive mechanisms in the monkey saccadic system (P. Inchingolo et al). Human express saccades: catch trails influence the probability of their occurrence (M. Juttner, W. Wolf). Eye-Head Coordination. Adaptation of eye and head movements to reduced peripheral vision (G.M. Gauthier et al). Coupled and dissociated modes of eye-head coordination in humans to flashed visual target (S. Ron, A. Berthoz). Superior colliculus and feedback control of gaze shift in the head-free cat (D. Pelisson et al). Strategies of eye-head coordination (R. Schmid, D. Zambarbieri). Eye Movements in Pathology. Sensory and motor aspects of congenital nystagmus (R.V. Abadi et al). Head-shaking nystagmus - A clinical tool to detect peripheral and central vestibular asymmetries (M. Fetter et al). Saccadic and smooth pursuit eye movements in olivo-ponto-cerebellar atrophies (M. Spanio et al). Slow eye movement abnormalities in Wallenberg's lateral medullary syndrome (W. Waespe). Eye tracking dysfunctions in schizophrenic patients - Why is there no neuroleptic side-effect? (N. Hock et al). Eye Movements and Cognitive Processes. A mathematical analysis of the convenient viewing position hypothesis and its components (M. Brysbaert, G. d'Ydewalle). Is there an optimal landing position in words during reading of texts? (F. Vitu, J.K. O'Regan). Processing of prepositions as reflected by gaze durations (A. Wilbertz et al). The latency of saccadic eye movements to texture-defined stimuli (H. Deubel, H. Frank). Does thinking aloud influence the structure of cognitive processes? (U. Lass et al). Eye movements in visual search: a test of the limited cognitive effort hypothesis and an analysis of the search operating characteristic (A.M. Jacobs). Applied Research. Colour, effective contrast and search performance (J.L. Barbur et al). Video-oculography - An alternative method for measurement of three-dimensional eye movements (A.H. Clarke et al).

Abstract: Introduction: Emergency medicine is a relatively new medical specialty in Brazil, approved just in 2016. Residency training programs have been implemented ever since. The emergency environment is known to represent a death-and-life tension on the professional team, culminating with high rates of mental illness in this population. The Covid-19 pandemic seems to be affecting these rates of depression, anxiety, and burnout in health professionals. Objective: To assess the symptoms of burnout, depression, and anxiety in Brazilian medical residents of Emergency Medicine during the Covid-19 pandemic and compare the residents’ beliefs regarding clinical practice related to Covid-19 patients. Methods: A quantitative study was conducted with a convenience sample of volunteer medical residents from an anonymous online survey, available during April 2020. This investigation collected sociodemographic information and used the Oldenburg Burnout Inventory (OLBI) to measure burnout; the Patient Health Questionnaire (PHQ-9) to measure depression; and the General Anxiety Disorders (GAD-7) to measure generalized anxiety disorder. This study also developed a Covid-19 Impact Questionnaire (CIQ-19) to assess the residents’ beliefs and clinical practices related to Covid-19 patients. Results: The survey consisted of 63 respondents, about 26,35% of emergency medicine residents in Brazil. Only 39.6% residents felt safe while working with Covid-19 patients. Mild depressive symptoms were found in 68.2% of the residents, followed by anxiety symptoms in 50.7% and burnout in 54.0% overall. About 12% of the residents do nothing about their mental health status, while some prefer to talk with family or friends (36.1%) and discuss with their team support (24.3%) when they need mental health care. Conclusion: Emergency medicine residents have high rates of mental illness and it could get worse when submitted to stressful and unknown situations, such as the Covid-19 pandemic. Initiatives should be made to improve these physicians’ mental health status. It is proposed that health institutions pay medical supervisors a closer and more unique look at physicians in training. A mentoring program proposal is an opportunity to reflect on technical and personal improvements for medical residents.