Diagnosing aortic dissection: A review of this elusive, lethal diagnosis.

Aortic dissection (AD) remains a difficult diagnosis in the emergency setting. Despite its rare occurrence, it is a life-threatening pathology that, if missed, is typically fatal. Previous studies have documented minimal improvement in timely and accurate diagnoses despite the advancement of computed tomography. Previous literature has highlighted aortic dissections as a major cause of serious misdiagnosis-related harm. The aim of this article is to review the available literature on AD, discussing the diversity in presentations and the prevalence of historical and exam features to better aid in the diagnosis of AD. AD remains a difficult diagnosis, even with the widespread prevalence of computed tomography angiography usage. No single feature of the history or physical examination is enough to raise suspicion. The diagnosis should be strongly considered in any patient with chest pain that is severe and unexplained by other findings or testing. Those who do not present with acute pain are often complicated by neurologic deficits, hypotension, or syncope. These patients suffer from a change in mental status limiting their ability to participate in the history and physical examination and have a higher rate of complications and mortality. An educated understanding of the atypical presentations of aortic dissection helps the clinician to realistically rank it on the differential diagnosis, culminating in judicious use of definitive imaging.

Diagnostic errors and characteristics of patients seen at a general internal medicine outpatient clinic with a referral for diagnosis.

OBJECTIVES: Patients referred to general internal medicine (GIM) outpatient clinics may face a higher risk of diagnostic errors than non-referred patients. This difference in risk is assumed to be due to the differences in diseases and clinical presentations between referred and non-referred patients; however, clinical data regarding this issue are scarce. This study aimed to determine the frequency of diagnostic errors and compare the characteristics of referred and non-referred patients visit GIM outpatient clinics. METHODS: This study included consecutive outpatients who visited the GIM outpatient clinic at a university hospital, with or without referral. Data on age, sex, chief complaints, referral origin, and final diagnosis were collected from medical records. The Revised Safer Dx Instrument was used to detect diagnostic errors. RESULTS: Data from 534 referred and 599 non-referred patients were analyzed. The diagnostic error rate was higher in the referral group than that in the non-referral group (2.2 % vs. 0.5 %, p=0.01). The prevalence of abnormal test results and sensory disturbances was higher in the chief complaints, and the prevalence of musculoskeletal system disorders, connective tissue diseases, and neoplasms was higher in the final diagnoses of referred patients compared with non-referred patients. Among referred patients with diagnostic errors, abnormal test results and sensory disturbances were the two most common chief complaints, whereas neoplasia was the most common final diagnosis. Problems with data integration and interpretation were found to be the most common factors contributing to diagnostic errors. CONCLUSIONS: Paying more attention to patients with abnormal test results and sensory disturbances and considering a higher pre-test probability for neoplasms may prevent diagnostic errors in patients referred to GIM outpatient clinics.

Racial and Ethnic Disparities in Occult Hypoxemia Prevalence and Clinical Outcomes Among Hospitalized Patients: A Systematic Review and Meta-analysis.

BACKGROUND: There is growing concern that pulse oximeters are routinely less accurate in hospitalized patients with darker skin pigmentation, in turn increasing risk of undetected (occult) hypoxemia and adverse clinical outcomes. The aim of this systematic review and meta-analysis was to synthesize evidence on racial and ethnic disparities in occult hypoxemia prevalence and clinical impacts of undetected hypoxemia. METHODS: Ovid MEDLINE, Embase, and CINAHL databases were searched for relevant articles published through January 2024. Eligible studies must have been conducted among adults in inpatient or outpatient settings and report occult hypoxemia prevalence stratified by patient race or ethnicity, or clinical outcomes stratified by patient race or ethnicity and occult hypoxemia status. Screening for inclusion was conducted independently by two investigators. Data extraction and risk of bias assessment were conducted by one investigator then checked by a second. Outcome data were synthesized using random-effects meta-analyses. RESULTS: Fifteen primary studies met eligibility criteria and reported occult hypoxemia prevalence in 732,505 paired oximetry measurements from 207,464 hospitalized patients. Compared with White patients, occult hypoxemia is likely more common among Black patients (pooled prevalence ratio = 1.67, 95% CI 1.47 to 1.90) and among patients identifying as Asian, Latinx, Indigenous, multiracial, or other race or ethnicity (pooled prevalence ratio = 1.39, 95% CI 1.19 to 1.64). Findings from studies reporting clinical outcomes suggest that Black patients with undetected hypoxemia may experience poorer treatment delivery outcomes than White patients with undetected hypoxemia. No evidence was found from outpatient settings. DISCUSSION: This review and included primary studies rely on self-identified race or ethnicity, which may obscure variability in occult hypoxemia risk. Findings underscore that clinicians should be aware of the risk of occult hypoxemia in hospitalized patients with darker skin pigmentation. Moreover, oximetry data from included studies suggests that the accuracy of pulse oximeters could vary substantially from patient to patient and even within individual patients. TRIAL REGISTRATION: PROSPERO ( CRD42023402152 ).

Case report: Ultrasound misdiagnoses atypical parathyroid adenoma as malignant thyroid tumor.

Atypical Parathyroid Adenoma (APA) is a type of tumor that lies somewhere between parathyroid adenoma and parathyroid carcinoma. It often affects adults over the age of 60, and the clinical symptoms are consistent with those of hyperparathyroidism. This condition has a low occurrence, and its ultrasonographic signs are strikingly similar to thyroid malignant tumors, making it easily misdiagnosed. As a result, a case of APA ultrasonography misdiagnosis admitted to our hospital was recorded in order to serve as a reference point for APA diagnosis.

Frequency and characteristics of errors by artificial intelligence (AI) in reading screening mammography: a systematic review.

PURPOSE: Artificial intelligence (AI) for reading breast screening mammograms could potentially replace (some) human-reading and improve screening effectiveness. This systematic review aims to identify and quantify the types of AI errors to better understand the consequences of implementing this technology. METHODS: Electronic databases were searched for external validation studies of the accuracy of AI algorithms in real-world screening mammograms. Descriptive synthesis was performed on error types and frequency. False negative proportions (FNP) and false positive proportions (FPP) were pooled within AI positivity thresholds using random-effects meta-analysis. RESULTS: Seven retrospective studies (447,676 examinations; published 2019-2022) met inclusion criteria. Five studies reported AI error as false negatives or false positives. Pooled FPP decreased incrementally with increasing positivity threshold (71.83% [95% CI 69.67, 73.90] at Transpara 3 to 10.77% [95% CI 8.34, 13.79] at Transpara 9). Pooled FNP increased incrementally from 0.02% [95% CI 0.01, 0.03] (Transpara 3) to 0.12% [95% CI 0.06, 0.26] (Transpara 9), consistent with a trade-off with FPP. Heterogeneity within thresholds reflected algorithm version and completeness of the reference standard. Other forms of AI error were reported rarely (location error and technical error in one study each). CONCLUSION: AI errors are largely interpreted in the framework of test accuracy. FP and FN errors show expected variability not only by positivity threshold, but also by algorithm version and study quality. Reporting of other forms of AI errors is sparse, despite their potential implications for adoption of the technology. Considering broader types of AI error would add nuance to reporting that can inform inferences about AI's utility.

Giant Cell Arteritis With Sepsis-Like Symptoms in an Elderly Man: A Case Report.

This case report details the diagnostic challenge and management of an 88-year-old man who presented to a rural Japanese community hospital with sepsis-like symptoms, initially suspected of acute bacterial cholangitis based on his physical and laboratory findings. Despite the antibiotic treatment of tazobactam and piperacillin, the patient's symptoms persisted, leading to further investigations that revealed no signs of infection but notable aortic arch wall thickening on contrast-enhanced computed tomography scans. These findings, combined with the patient's clinical presentation and lack of antibiotic response, redirected the diagnosis toward giant cell arteritis (GCA). The administration of prednisolone of 60 mg daily significantly alleviated symptoms and prevented potential severe complications such as blindness and irreversible neurological damage. This case underscores the importance of considering GCA in elderly patients presenting with systemic inflammatory symptoms and the necessity of timely intervention. It also highlights the challenges in managing high-dose steroid therapy in elderly patients and suggests the potential benefits of integrating immunosuppressants to reduce steroid dependency. This report emphasizes the need for heightened awareness and a comprehensive diagnostic approach in atypical presentations of GCA, particularly in geriatric populations within resource-limited healthcare settings.

A systematic evaluation of the performance of GPT-4 and PaLM2 to diagnose comorbidities in MIMIC-IV patients.

Background: Given the strikingly high diagnostic error rate in hospitals, and the recent development of Large Language Models (LLMs), we set out to measure the diagnostic sensitivity of two popular LLMs: GPT-4 and PaLM2. Small-scale studies to evaluate the diagnostic ability of LLMs have shown promising results, with GPT-4 demonstrating high accuracy in diagnosing test cases. However, larger evaluations on real electronic patient data are needed to provide more reliable estimates. Methods: To fill this gap in the literature, we used a deidentified Electronic Health Record (EHR) data set of about 300,000 patients admitted to the Beth Israel Deaconess Medical Center in Boston. This data set contained blood, imaging, microbiology and vital sign information as well as the patients' medical diagnostic codes. Based on the available EHR data, doctors curated a set of diagnoses for each patient, which we will refer to as ground truth diagnoses. We then designed carefully-written prompts to get patient diagnostic predictions from the LLMs and compared this to the ground truth diagnoses in a random sample of 1000 patients. Results: Based on the proportion of correctly predicted ground truth diagnoses, we estimated the diagnostic hit rate of GPT-4 to be 93.9%. PaLM2 achieved 84.7% on the same data set. On these 1000 randomly selected EHRs, GPT-4 correctly identified 1116 unique diagnoses. Conclusion: The results suggest that artificial intelligence (AI) has the potential when working alongside clinicians to reduce cognitive errors which lead to hundreds of thousands of misdiagnoses every year. However, human oversight of AI remains essential: LLMs cannot replace clinicians, especially when it comes to human understanding and empathy. Furthermore, a significant number of challenges in incorporating AI into health care exist, including ethical, liability and regulatory barriers.

Delayed diagnosis in a child with strangulated mesenteric hernia.

Background: Strangulated intestinal obstruction is a life-threatening condition that should be considered as a differential diagnosis in children with shock. However, it has pitfalls in diagnosis and can lead to diagnostic errors. Case Presentation: A 3-month-old male patient presented with a pale complexion lasting 2 h and abnormal crying. He was in shock with lactic acidosis, altered mental status, and slight abdominal distension. He required volume resuscitation, vasoactive agents, and transfusion. On Day 2, he had marked abdominal distension and acute kidney injury, which required continuous kidney replacement therapy. Contrast-enhanced computed tomography revealed extensive intestinal ischemia. It took 33.5 h from his arrival to the computed tomography, leading to operative management. The small intestine had entered a mesenteric hiatus, leading to ischemia. He was diagnosed with strangulated mesenteric hernia. Conclusion: In this case, four pitfalls led to delayed diagnosis. Factors for diagnostic errors specific to strangulated intestinal obstruction and intensive care should be noted.

Evaluating ChatGPT-4's Accuracy in Identifying Final Diagnoses Within Differential Diagnoses Compared With Those of Physicians: Experimental Study for Diagnostic Cases.

BACKGROUND: The potential of artificial intelligence (AI) chatbots, particularly ChatGPT with GPT-4 (OpenAI), in assisting with medical diagnosis is an emerging research area. However, it is not yet clear how well AI chatbots can evaluate whether the final diagnosis is included in differential diagnosis lists. OBJECTIVE: This study aims to assess the capability of GPT-4 in identifying the final diagnosis from differential-diagnosis lists and to compare its performance with that of physicians for case report series. METHODS: We used a database of differential-diagnosis lists from case reports in the American Journal of Case Reports, corresponding to final diagnoses. These lists were generated by 3 AI systems: GPT-4, Google Bard (currently Google Gemini), and Large Language Models by Meta AI 2 (LLaMA2). The primary outcome was focused on whether GPT-4's evaluations identified the final diagnosis within these lists. None of these AIs received additional medical training or reinforcement. For comparison, 2 independent physicians also evaluated the lists, with any inconsistencies resolved by another physician. RESULTS: The 3 AIs generated a total of 1176 differential diagnosis lists from 392 case descriptions. GPT-4's evaluations concurred with those of the physicians in 966 out of 1176 lists (82.1%). The Cohen κ coefficient was 0.63 (95% CI 0.56-0.69), indicating a fair to good agreement between GPT-4 and the physicians' evaluations. CONCLUSIONS: GPT-4 demonstrated a fair to good agreement in identifying the final diagnosis from differential-diagnosis lists, comparable to physicians for case report series. Its ability to compare differential diagnosis lists with final diagnoses suggests its potential to aid clinical decision-making support through diagnostic feedback. While GPT-4 showed a fair to good agreement for evaluation, its application in real-world scenarios and further validation in diverse clinical environments are essential to fully understand its utility in the diagnostic process.

Underdiagnosis of umbilical hernias in CT scans in a multicenter study - the radiologically neglected pathology and its surgical implications.

PURPOSE: Umbilical hernias (UH) have a higher prevalence than previously considered. With the high workload radiologists must endure, UH can be missed when interpreting a computed tomography scan (CT). The clinical implications of its misdiagnosis are yet to be determined. Unreporting could lead to content lesions in surgical approaches and other potential complications. The aim was to determine the prevalence of UH using CT scans, and the incidence of radiological reporting. METHODS: A multicenter, cross-sectional study was performed in four tertiary-level hospitals. CT scans were reviewed for abdominal wall defects at the umbilicus, and radiological reports were examined to compare findings. In the case of UH, transversal, anteroposterior, and craniocaudal lengths were obtained. RESULTS: A total of 1557 CTs were included, from which 971 (62.4%, 95% CI 0.59-0.64) had UH. Out of those, 629 (64.8%, 95% CI 0.61-0.67) of the defects were not included in the radiological report. Smaller UH (x̄: 7.7 × 6.0 mm) were more frequently missed. Of the reported UH, 187 (54.7%) included at least one axis measurement, 289 (84.5%) content description, and 146 (42.7%) whether or not there were complication signs. CONCLUSION: There is a high prevalence of UH, and a high incidence of under-reporting. This raises the question of whether this is a population-based finding or the norm worldwide. The reason of under-reporting and the clinical implications of these must be addressed in further studies.

Performance evaluation of ChatGPT in detecting diagnostic errors and their contributing factors: an analysis of 545 case reports of diagnostic errors.

BACKGROUND: Manual chart review using validated assessment tools is a standardised methodology for detecting diagnostic errors. However, this requires considerable human resources and time. ChatGPT, a recently developed artificial intelligence chatbot based on a large language model, can effectively classify text based on suitable prompts. Therefore, ChatGPT can assist manual chart reviews in detecting diagnostic errors. OBJECTIVE: This study aimed to clarify whether ChatGPT could correctly detect diagnostic errors and possible factors contributing to them based on case presentations. METHODS: We analysed 545 published case reports that included diagnostic errors. We imputed the texts of case presentations and the final diagnoses with some original prompts into ChatGPT (GPT-4) to generate responses, including the judgement of diagnostic errors and contributing factors of diagnostic errors. Factors contributing to diagnostic errors were coded according to the following three taxonomies: Diagnosis Error Evaluation and Research (DEER), Reliable Diagnosis Challenges (RDC) and Generic Diagnostic Pitfalls (GDP). The responses on the contributing factors from ChatGPT were compared with those from physicians. RESULTS: ChatGPT correctly detected diagnostic errors in 519/545 cases (95%) and coded statistically larger numbers of factors contributing to diagnostic errors per case than physicians: DEER (median 5 vs 1, p<0.001), RDC (median 4 vs 2, p<0.001) and GDP (median 4 vs 1, p<0.001). The most important contributing factors of diagnostic errors coded by ChatGPT were 'failure/delay in considering the diagnosis' (315, 57.8%) in DEER, 'atypical presentation' (365, 67.0%) in RDC, and 'atypical presentation' (264, 48.4%) in GDP. CONCLUSION: ChatGPT accurately detects diagnostic errors from case presentations. ChatGPT may be more sensitive than manual reviewing in detecting factors contributing to diagnostic errors, especially for 'atypical presentation'.

Determining diagnostic delays in Romanian multiple myeloma patients using the Aarhus statement.

Introduction: Multiple Myeloma (MM) is classified as one of the most challenging cancers to diagnose, and the hematological malignancy is associated with prolonged diagnostic delays. Although major steps have been made in the improvement of MM patient diagnosis and care, Romanian patients still face long diagnostic delays. Thus far, there have been no studies evaluating the factors associated with diagnostic errors in Romanian MM patients. Methods: Using the Aarhus statement, we prospectively determined the diagnostic intervals for 103 patients diagnosed with MM at Fundeni Clinical Institute, between January 2022 and March 2023. Results: Our data revealed that the main diagnostic delays are experienced during the "patient interval." Patients spend a median of 162 days from the first symptom onset until the first doctor appointment. Bone pain is the most frequently reported symptom by patients (78.64%), but it leads to a medical-seeking behavior in only half of the reporting patients and results in a median delay of 191 days. The changes in routine lab tests are considered most worrisome for patients, leading to a medical appointment after a median of only 25 days. The median primary care interval was 70 days, with patients having an average of 3.7 medical visits until MM suspicion was first raised. The secondary care interval did not contribute to the diagnostic delays. Discussion: Overall, the median diagnostic path for MM patients in Romania was more than 6 months, leading to a higher number of emergency presentations and myeloma-related end-organ damage.

Analysing false-positive errors when Australian radiographers use preliminary image evaluation.

INTRODUCTION: Diagnostic errors in the emergency departments can have major implications on patient outcomes. Preliminary Image Evaluation (PIE) is a brief comment written by a radiographer describing an acute or traumatic pathology on a radiograph and can be used to complement referrer's image interpretation in the absence of the radiologist report. Currently, no studies exist that focus their analysis on false-positive (FP) errors in PIE. The purpose of this study was to investigate the regions of the body that cause the most FP errors and recognise other areas in image interpretation that may need additional attention. METHODS: A longitudinal retrospective clinical audit was conducted to determine the accuracy of radiographer PIE's over 5 years from January 2016 to December 2020. PIE's were compared to the radiologist report to assess for diagnostic accuracy. FP and unsure errors were further categorised by anatomical region and age. RESULTS: Over this period, a sample size of 11,090 PIE audits were included in the study demonstrating an overall PIE accuracy of 87.7%. Foot, ankle and chest regions caused the most FP errors, while ankle, shoulder and elbow caused the most unsure cases. 76% of the unsure cases were negative for any pathology when compared to the radiologist report. The paediatric population accounted for 21.3% of FP cases and 33.6% of unsure cases. CONCLUSION: Findings in this study should be used to tailor education specific to radiographer image interpretation. Improving radiography image interpretation skills can assist in improving referrer diagnostic accuracy, thus improving patient outcomes.

What is diagnostic safety? A review of safety science paradigms and rethinking paths to improving diagnosis.

Diagnostic errors in health care are a global threat to patient safety. Researchers have traditionally focused diagnostic safety efforts on identifying errors and their causes with the goal of reducing diagnostic error rates. More recently, complementary approaches to diagnostic errors have focused on improving diagnostic performance drawn from the safety sciences. These approaches have been called Safety-II and Safety-III, which apply resilience engineering and system safety principles, respectively. This review explores the safety science paradigms and their implications for analyzing diagnostic errors, highlighting their distinct yet complementary perspectives. The integration of Safety-I, Safety-II, and Safety-III paradigms presents a promising pathway for improving diagnosis. Diagnostic researchers not yet familiar with the various approaches and potential paradigm shift in diagnostic safety research may use this review as a starting point for considering Safety-I, Safety-II, and Safety-III in their efforts to both reduce diagnostic errors and improve diagnostic performance.

Importance of clinico-pathologic correlation in rare, chronic infectious diseases: Actinomycetoma misdiagnosed as botryomycosis-A case report.

This case report explores the clinical journey of a patient initially diagnosed with botryomycosis, only to later reveal the underlying and rare condition of actinomycosis. The report highlights the challenges in getting to an accurate diagnosis, emphasizing the importance of considering uncommon pathologies, the utility of multi-disciplinary teams and clinico-pathologic correlation in clinical practice.

Fahr's disease with neuropsychiatric symptoms and intermittent course: a case report.

Fahr's disease is a rare neurodegenerative disorder with brain calcifications and neuropsychiatric symptoms. It can have variable phenotypic expression and intermittent symptomatology, making diagnosis challenging. In this report, we describe a young female patient presenting with symptoms of psychosis and confusion, which could be indicative of a delirium superimposed on the cerebral vulnerability associated with Fahr's disease. Notably, about two years prior, she experienced multiple episodes of tonic-clonic seizures that spontaneously resolved without pharmacological intervention. She had no previous psychiatric history. Following comprehensive investigations, other organic causes were ruled out, and Fahr's disease was diagnosed based on bilateral symmetrical brain calcifications seen on a head CT scan. Her treatment regimen encompassed antipsychotics and anticonvulsants. This case highlights the importance of considering Fahr's disease as a differential diagnosis in patients with new-onset neuropsychiatric symptoms. The case also explores the atypical early onset and intermittent nature of symptoms in the absence of a positive family history, highlighting the complexity of Fahr's disease. A multidisciplinary approach and regular follow-up are crucial for optimizing patient care and monitoring disease progression. Further research is needed to enhance our understanding of Fahr's disease and develop standardized treatment strategies for this rare condition.

Determining the psychometric properties of a written test to assess safe dental practice.

INTRODUCTION: Safe practice in medicine and dentistry has been a global priority area in which large knowledge gaps are present.Patient safety strategies aim at preventing unintended damage to patients that can be caused by healthcare practitioners. One of the components of patient safety is safe clinical practice. Patient safety efforts will help in ensuring safe dental practice for early detection and limiting non-preventable errors.A valid and reliable instrument is required to assess the knowledge of dental students regarding patient safety. OBJECTIVE: To determine the psychometric properties of a written test to assess safe dental practice in undergraduate dental students. MATERIAL AND METHODS: A test comprising 42 multiple-choice questions of one-best type was administered to final year students (52) of a private dental college. Items were developed according to National Board of Medical Examiners item writing guidelines. The content of the test was determined in consultation with dental experts (either professor or associate professor). These experts had to assess each item on the test for language clarity as A: clear, B: ambiguous and relevance as 1: essential, 2: useful, not necessary, 3: not essential. Ethical approval was taken from the concerned dental college. Statistical analysis was done in SPSS V.25 in which descriptive analysis, item analysis and Cronbach's alpha were measured. RESULT: The test scores had a reliability (calculated by Cronbach's alpha) of 0.722 before and 0.855 after removing 15 items. CONCLUSION: A reliable and valid test was developed which will help to assess the knowledge of dental students regarding safe dental practice. This can guide medical educationist to develop or improve patient safety curriculum to ensure safe dental practice.