Diagnostic Errors and the Bedside Clinical Examination

AI-powered techniques in anatomical imaging: Impacts on veterinary diagnostics and surgery

BACKGROUND

Artificial intelligence (AI) is rapidly transforming veterinary diagnostic imaging, offering improved accuracy, speed, and efficiency in analyzing complex anatomical structures. AI-powered systems, including deep learning and convolutional neural networks, show promise in interpreting medical images from various modalities like X-rays, ultrasounds, CT scans, and MRI/mammography.

STUDY DESIGN

Narrative review OBJECTIVE: This review aims to explore the innovations and challenges of AI-enabled imaging tools in veterinary diagnostics and surgery, highlighting their potential impact on diagnostic accuracy, surgical risk mitigation, and personalized veterinary healthcare.

METHODS

We reviewed recent literature on AI applications in veterinary diagnostic imaging, focusing on their benefits, limitations, and future directions.

CONCLUSION

AI-enabled imaging tools hold immense potential for revolutionizing veterinary diagnostics and surgery. By enhancing diagnostic accuracy, enabling precise surgical planning, and supporting personalized treatment strategies, AI can significantly improve animal health outcomes. However, addressing challenges related to data privacy, algorithm bias, and integration into clinical workflows is crucial for the widespread adoption and success of these transformative technologies.

Operational criteria application does not change clinicians' opinion on the diagnosis of mental disorder: a pre- and post-intervention validity study

Objective

Our objective was to check if the ICD-10 operational criteria application changes non-operational, prototype-based diagnoses obtained in a real-life scenario.

Methods

Psychiatry residents applied the diagnostic criteria of the ICD-10 as a "diagnostic test" to five outpatient patients they were already following who had a prototype-based diagnosis. Tests were used to ascertain whether changes in opinion were significant and if any of the diagnostic groups were more prone to change than others. The present paper is part of the study with UTN U1111-1260-1212.

Results

Seventeen residents reviewed their last five case files, retrieving 85 diagnostic pairs of non-operational-based vs. operational-based diagnoses. The Stuart-Maxwell test did not indicate a significant opinion change (χ2 = 5.25, p = 0.39; power = 0.94) besides 30% of diagnostic changes. Despite not being statistically significant, 20.2% of all evaluations resulted in a change that would affect treatment choices. Using ICD-10 operational criteria slightly increased the number of observed diagnoses, but probably without clinical relevance. None of the non-operational diagnoses have a higher tendency to change with operational criteria application (χ2 = 11.6, p = 0.07). The female gender was associated with a higher diagnostic change tendency.

Conclusion

Applying ICD-10 operational criteria as a diagnostic test does not induce a statistically significant diagnostic opinion change in residents and no diagnostic group seems more sensible to diagnostic change. Gender-related differences in diagnostic opinion changes might be evidence of sunk cost bias. Although not statistically significant, using operational criteria after diagnostic elaboration might help to deal with subjects without adequate treatment response.

The impact of automatic history-taking software on data quality in the cardiology outpatient clinic: Retrospective observational study

Background

Healthcare delivery now mandates shorter visits despite the need for more data entry, under-mining patient-provider interaction. Furthermore, enhancing access to the outcomes of prior tests and imaging conducted on the patient, along with accurately documenting medication history, will significantly elevate the quality of healthcare service delivery.

Objective

To enhance the efficiency of clinic visits, we have devised a patient-provider portal that systematically gathers symptom and clinical data from patients through a computer algorithm known as Automated Assessment of Cardiovascular Examination (AACE). We intended to assess the quality of computer-generated Electronic Health Records (EHRs) with those documented by physicians.

Methods

We conducted a cross-sectional study employing a paired-sample design, focusing on individuals seeking assessment for active cardiovascular symptoms at outpatient adult cardiovascular clinics. Participants initially completed the AACE, and subsequently, in the first protocol, patients were subjected to routine care without providing the AACE forms to examining physicians. In the second protocol, the AACE form was presented to the physician before the examination, and participants were subjected to routine care. We assessed the impact of AACE forms generated through computerized history-taking method on the examination, considering various clinical outcomes and satisfaction surveys.

Results

We included non-randomized eligible patients who visited seven general cardiology outpatient clinics between September 18, 2023, and October 27, 2023. These clinics were staffed by the same physicians who were unaware of the content and details of the study. A total of 762 patients (394 patients in protocol 1 and 368 patients in protocol 2) were included in the study. The mean overall impression score for computer-generated EHRs was higher versus physician EHRs (4.2 vs. 2.6; p < .001). Our study demonstrated that EHRs created by physicians' exhibit inaccuracies or deficiencies in various pieces of information. In the second protocol, in which the AACE form was presented to the physician before the examination, it was determined that the examination time was shorter, the number of tests requested, and the number of new drugs prescribed were less.

Conclusions

We observed that the patient-provider portal, systematically collecting symptom and clinical data from patients through a computer algorithm known as AACE, yielded records that were of higher quality, more comprehensive, better organized, and more relevant compared to those documented by physicians.

Risk assessment of vascular complications following manual therapy and exercise for the cervical region: diagnostic accuracy of the International Federation of Orthopaedic Manipulative Physical Therapists framework (The Go4Safe project)

QUESTION

What is the diagnostic accuracy of the International Federation of Orthopaedic Manipulative Physical Therapists (IFOMPT) framework to assess the risk of vascular complications in patients seeking physiotherapy care for neck pain and/or headache?

DESIGN

Cross-sectional diagnostic accuracy study.

PARTICIPANTS

One hundred and fifty patients seeking physiotherapy for neck pain and/or headache in primary care.

METHODS

Nineteen physiotherapists performed the index test according to the IFOMPT framework. Patients were classified as having a high, intermediate or low risk of vascular complications, following manual therapy and/or exercise, derived from the estimated risk of the presence of vascular pathology. The reference test was a consensus medical decision reached by a vascular neurologist and an interventional neurologist, with input from a neuroradiologist. The neurologists had access to clinical data and magnetic resonance imaging of the cervical spine, including an angiogram of the cervical arteries.

OUTCOME MEASURES

Diagnostic accuracy measures were calculated for 'no contraindication' (ie, the low-risk category) and 'contraindication' (ie, the high-risk and intermediate-risk categories) for manual therapy and/or exercise. Sensitivity, specificity, predictive values, likelihood ratios and the area under the curve were calculated.

RESULTS

Manual therapy and/or exercise were contraindicated in 54.7% of the patients. The sensitivity of the IFOMPT framework was low (0.50, 95% CI 0.39 to 0.61) and its specificity was moderate (0.63, 95% CI 0.51 to 0.75). The positive and negative likelihood ratios were weak at 1.36 (95% CI 0.93 to 1.99) and 0.79 (95% CI 0.60 to 1.05), respectively. The area under the curve was poor (0.57, 95% CI 0.49 to 0.65).

CONCLUSION

The IFOMPT framework has poor diagnostic accuracy when compared with a reference standard consisting of a consensus medical decision.

Improving decision accuracy using a clinical decision support system for medical students during history-taking: a randomized clinical trial

BACKGROUND

A clinical diagnostic support system (CDSS) can support medical students and physicians in providing evidence-based care. In this study, we investigate diagnostic accuracy based on the history of present illness between groups of medical students using a CDSS, Google, and neither (control). Further, the degree of diagnostic accuracy of medical students using a CDSS is compared with that of residents using neither a CDSS nor Google.

METHODS

This study is a randomized educational trial. The participants comprised 64 medical students and 13 residents who rotated in the Department of General Medicine at Chiba University Hospital from May to December 2020. The medical students were randomly divided into the CDSS group (n = 22), Google group (n = 22), and control group (n = 20). Participants were asked to provide the three most likely diagnoses for 20 cases, mainly a history of a present illness (10 common and 10 emergent diseases). Each correct diagnosis was awarded 1 point (maximum 20 points). The mean scores of the three medical student groups were compared using a one-way analysis of variance. Furthermore, the mean scores of the CDSS, Google, and residents' (without CDSS or Google) groups were compared.

RESULTS

The mean scores of the CDSS (12.0 ± 1.3) and Google (11.9 ± 1.1) groups were significantly higher than those of the control group (9.5 ± 1.7; p = 0.02 and p = 0.03, respectively). The residents' group's mean score (14.7 ± 1.4) was higher than the mean scores of the CDSS and Google groups (p = 0.01). Regarding common disease cases, the mean scores were 7.4 ± 0.7, 7.1 ± 0.7, and 8.2 ± 0.7 for the CDSS, Google, and residents' groups, respectively. There were no significant differences in mean scores (p = 0.1).

CONCLUSIONS

Medical students who used the CDSS and Google were able to list differential diagnoses more accurately than those using neither. Furthermore, they could make the same level of differential diagnoses as residents in the context of common diseases.

TRIAL REGISTRATION

This study was retrospectively registered with the University Hospital Medical Information Network Clinical Trials Registry on 24/12/2020 (unique trial number: UMIN000042831).

Will clinical signs become myth? Developing structured Signs Circuits to improve medical students' exposure to and confidence examining clinical signs

BACKGROUND

Correctly eliciting and interpreting physical examination (PEx) signs contributes to successful diagnosis and is fundamental to patient care. A significant decline in the time spent acquiring these skills by medical students, and the decreased ability to elicit and recognise signs is widely acknowledged. However, organising teaching to counteract this in the busy clinical environment is challenging. We evaluated the prior exposure to clinical signs, and experience of examination teaching among a cohort of final-year medical students. Following this, we assessed the utility of a structured circuit-based approach (Signs Circuits) using hospital inpatients and junior doctors to provide high-yield PEx teaching and overcome these limitations.

MATERIALS AND METHODS

Qualitative and quantitative survey feedback, including a standardised list of 62 clinical signs, was sought from final-year medical students during their rotations at a teaching hospital in London, UK, before and after the provision of Signs Circuits.

RESULTS

Prior to the course the 63 students reported limited exposure to even the most common clinical signs. For example, the murmurs of mitral and tricuspid regurgitation and the sound of lung crackles eluded 43%, 87%, and 32%, respectively. From qualitative feedback, the reasons for this included that much of their prior PEx experience had focused on the performance of appropriate examination steps and techniques in patients without pathology. During the course, students were exposed to an average of 4.4 new signs, and left with increased confidence examining and eliciting signs, and a firmer belief in their importance to diagnosis.

CONCLUSION

Medical students continue to have limited exposure to clinical signs in medical school. This signs-focused approach to PEx teaching is an effective and reproducible way to counter the deficiencies identified in signsexposure.

Education in the Pediatric Intensive Care Unit

This article addresses the latest data and ideas related to education in the pediatric intensive care unit, including traditional education methods with newer and technology-based methods. A review of adult learning theory is included with discussions regarding medical decision making and error prevention, bedside teaching, medical simulation, and electronic methods of education.

Difficulties with Learning Musculoskeletal Physical Examination Skills: Student Perspectives and General Lessons Learned for Curricular Design

Phenomenon: The development of foundational clinical skills, such as physical examination, is essential to becoming a competent clinician. Musculoskeletal medicine is often considered a specialized area of practice despite the high prevalence of musculoskeletal conditions in the general population and presenting to general clinical practices. Prior work has shown that medical learners and practicing clinicians have low confidence in these skills but understanding of the student perspective on why these skills are more difficult to acquire is unclear.Approach: Our study was guided by social constructivist learning theory to explore the learner experience and present their perspectives. Qualitative analysis investigated the difference between learning musculoskeletal physical examination versus other body systems, using the voices from 11 semi-structured focus group interviews. Participants included third-year medical students across two academic cohorts at one institution. Our analysis was grounded in the principles of phenomenology and used triangulation and reflexivity to provide rigorous analysis.Findings: Students provided rich and insightful perspectives regarding their experiences in learning musculoskeletal physical examination techniques. Four themes were developed from our data: a) the need for opportunities for both supervised and self-directed practice; b) assessment and competence as motivations for learning; c) the need for a different approach to the content and structure of musculoskeletal medicine and its associated examination techniques; and d) the need for distinct expertise and technical skill from musculoskeletal examination teachers.Insights: This study provides a valuable lens to critically reflect on existing curriculum and pedagogical approaches to musculoskeletal examination skills. Lessons from this study may be applicable to curriculum design in general, especially the teaching of physical examination skills, such as how it is taught and integrated with other content (including anatomy), how much practice is required, who teaches physical examination skills, and what faculty development is needed to standardize teaching. Promoting a learner-centered approach to the teaching and learning of these clinical skills will be beneficial to all stakeholders, especially to our future physicians and their patients.Supplemental data for this article is available online at https://doi.org/10.1080/10401334.2021.1954930 .

Sexual history taking: Doctors' clinical decision-making in primary care in the North West province, South Africa

BACKGROUND

Clinical reasoning is an important aspect of making a diagnosis for providing patient care. Sexual dysfunction can be as a result of cardiovascular or neurological complications of patients with chronic illness, and if a patient does not raise a sexual challenge, then the doctor should know that there is a possibility that one exists and enquire.

AIM

The aim of this research study was to assess doctors' clinical decision-making process with regards to the risk of sexual dysfunction and management of patients with chronic illness in primary care facilities of the North West province based on two hypothetical patient scenarios.

SETTING

This research study was carried out in 10 primary care facilities in Dr Kenneth Kaunda health district, North West province, a rural health district.

METHODS

This vignette study using two hypothetical patient scenarios formed part of a broader grounded theory study to determine whether sexual dysfunction as comorbidity formed part of the doctors' clinical reasoning and decision-making. After coding the answers, quantitative content analysis was performed. The questions and answers were then compared with standard answers of a reference group.

RESULTS

One of the doctors (5%) considered sexual dysfunction, but failed to follow through without considering further exploration, investigations or management. For the scenario of a female patient with diabetes, the reference group considered cervical health questions (p = 0.001) and compliance questions (p = 0.004) as standard enquiries, which the doctors from the North West province failed to consider. For the scenario of a male patient with hypertension and an ex-smoker, the reference group differed significantly by expecting screening for mental health and vision (both p = 0.001), as well as for HIV (p 0.001). The participating doctors did not meet the expectations of the reference group.

CONCLUSION

Good clinical reasoning and decision-making are not only based on knowledge, intuition and experience but also based on an awareness of human well-being as complex and multidimensional, to include sexual well-being.

Computerized history-taking improves data quality for clinical decision-making-Comparison of EHR and computer-acquired history data in patients with chest pain

Patients’ medical histories are the salient dataset for diagnosis. Prior work shows consistently, however, that medical history-taking by physicians generally is incomplete and not accurate. Such findings suggest that methods to improve the completeness and accuracy of medical history data could have clinical value. We address this issue with expert system software to enable automated history-taking by computers interacting directly with patients, i.e. computerized history-taking (CHT). Here we compare the completeness and accuracy of medical history data collected and recorded by physicians in electronic health records (EHR) with data collected by CHT for patients presenting to an emergency room with acute chest pain. Physician history-taking and CHT occurred at the same ED visit for all patients. CHT almost always preceded examination by a physician. Data fields analyzed were relevant to the differential diagnosis of chest pain and comprised information obtainable only by interviewing patients. Measures of data quality were completeness and consistency of negative and positive findings in EHR as compared with CHT datasets. Data significant for the differential of chest pain was missing randomly in all EHRs across all data items analyzed so that the dimensionality of EHR data was limited. CHT files were near complete for all data elements reviewed. Separate from the incompleteness of EHR data, there were frequent factual inconsistencies between EHR and CHT data across all data elements. EHR data did not contain representations of symptoms that were consistent with those reported by patients during CHT.

Trial registration: This study is registered at https://www.clinicaltrials.gov (unique identifier: NCT03439449).

Validation of an electronic trigger to measure missed diagnosis of stroke in emergency departments

OBJECTIVE

Diagnostic errors are major contributors to preventable patient harm. We validated the use of an electronic health record (EHR)-based trigger (e-trigger) to measure missed opportunities in stroke diagnosis in emergency departments (EDs).

METHODS

Using two frameworks, the Safer Dx Trigger Tools Framework and the Symptom-disease Pair Analysis of Diagnostic Error Framework, we applied a symptom-disease pair-based e-trigger to identify patients hospitalized for stroke who, in the preceding 30 days, were discharged from the ED with benign headache or dizziness diagnoses. The algorithm was applied to Veteran Affairs National Corporate Data Warehouse on patients seen between 1/1/2016 and 12/31/2017. Trained reviewers evaluated medical records for presence/absence of missed opportunities in stroke diagnosis and stroke-related red-flags, risk factors, neurological examination, and clinical interventions. Reviewers also estimated quality of clinical documentation at the index ED visit.

RESULTS

We applied the e-trigger to 7,752,326 unique patients and identified 46,931 stroke-related admissions, of which 398 records were flagged as trigger-positive and reviewed. Of these, 124 had missed opportunities (positive predictive value for "missed" = 31.2%), 93 (23.4%) had no missed opportunity (non-missed), 162 (40.7%) were miscoded, and 19 (4.7%) were inconclusive. Reviewer agreement was high (87.3%, Cohen's kappa = 0.81). Compared to the non-missed group, the missed group had more stroke risk factors (mean 3.2 vs 2.6), red flags (mean 0.5 vs 0.2), and a higher rate of inadequate documentation (66.9% vs 28.0%).

CONCLUSION

In a large national EHR repository, a symptom-disease pair-based e-trigger identified missed diagnoses of stroke with a modest positive predictive value, underscoring the need for chart review validation procedures to identify diagnostic errors in large data sets.

Diagnostic reasoning in internal medicine: a practical reappraisal

The practice of clinical medicine needs to be a very flexible discipline which can adapt promptly to continuously changing surrounding events. Despite the huge advances and progress made in recent decades, clinical reasoning to achieve an accurate diagnosis still seems to be the most appropriate and distinctive feature of clinical medicine. This is particularly evident in internal medicine where diagnostic boundaries are often blurred. Making a diagnosis is a multi-stage process which requires proper data collection, the formulation of an illness script and testing of the diagnostic hypothesis. To make sense of a number of variables, physicians may follow an analytical or an intuitive approach to clinical reasoning, depending on their personal experience and level of professionalism. Intuitive thinking is more typical of experienced physicians, but is not devoid of shortcomings. Particularly, the high risk of biases must be counteracted by de-biasing techniques, which require constant critical thinking. In this review, we discuss critically the current knowledge regarding diagnostic reasoning from an internal medicine perspective.

Typical RSV cough: myth or reality? A diagnostic accuracy study

Respiratory syncytial virus (RSV) is well known for causing a potentially severe course of bronchiolitis in infants. Many paediatric healthcare workers claim to be able to diagnose RSV based on cough sound, which was evaluated in this study. Parents of children < 1 year old admitted to the paediatric ward because of airway complaints were asked to record cough sounds of their child. In all children, MLPA analysis-a variation of PCR analysis-on nasopharyngeal swab was performed (golden standard). Sixteen cough fragments representing 4 different viral pathogens were selected and presented to paediatric healthcare workers. Thirty-two paediatric nurses, 16 residents and 16 senior staff members were asked to classify the audio files and state whether the cough was due to RSV infection or not. Senior staff, nurses and residents correctly identified RSV with a sensitivity of 76.2%, 73.1% and 51.3% respectively. Correct exclusion of RSV cases was performed with a specificity of 60.8%, 60.2% and 65.3% respectively. Sensitivity ranged from 0 to 100% between colleagues; no one correctly identified all negatives. Residents had significantly lower rates of sensitivity than senior staff and nurses. This was strongly related to work experience, in which more than 3.5 years of work experience was related to the best result.Conclusion: Senior staff and nurses were better in making a cough-based diagnosis of RSV compared to residents. Both groups were able to detect the same proportion of true RSV patients based on cough sounds compared to bedside tests but could not validly distinguish RSV from other pathogens based on cough sounds. What is Known: • Many paediatric healthcare workers claim to be capable of diagnosing RSV in infants based on cough sound • Up to now, no studies investigating the recognisability of RSV based on cough sound are published What is New: • Senior staff and paediatric nurses performed better than various other bedside tests in diagnosing RSV but could not replace MLPA analysis • Residents need at least 3.5 years of work experience to be able to make a RSV diagnosis based on cough sound.

Controversies in diagnosis: contemporary debates in the diagnostic safety literature

Since the 2015 publication of the National Academy of Medicine's (NAM) Improving Diagnosis in Health Care (Improving Diagnosis in Health Care. In: Balogh EP, Miller BT, Ball JR, editors. Improving Diagnosis in Health Care. Washington (DC): National Academies Press, 2015.), literature in diagnostic safety has grown rapidly. This update was presented at the annual international meeting of the Society to Improve Diagnosis in Medicine (SIDM). We focused our literature search on articles published between 2016 and 2018 using keywords in Pubmed and the Agency for Healthcare Research and Quality (AHRQ)'s Patient Safety Network's running bibliography of diagnostic error literature (Diagnostic Errors Patient Safety Network: Agency for Healthcare Research and Quality; Available from: https://psnet.ahrq.gov/search?topic=Diagnostic-Errors&f_topicIDs=407). Three key topics emerged from our review of recent abstracts in diagnostic safety. First, definitions of diagnostic error and related concepts are evolving since the NAM's report. Second, medical educators are grappling with new approaches to teaching clinical reasoning and diagnosis. Finally, the potential of artificial intelligence (AI) to advance diagnostic excellence is coming to fruition. Here we present contemporary debates around these three topics in a pro/con format.

The Modern Physical Exam - A Transatlantic Perspective from the Resident Level

Issue: The physical examination has been in decline for many years and poorer skills contribute to medical errors and adverse events. Diagnostic error is also increasing with the complexity of medicine. Comparing the physical examination in Ireland and the United States with a focus on education, assessment, culture, and health systems may provide insight into the decline of the physical exam in the United States, uncover possible strategies to improve clinical skills, and limit diagnostic error. Evidence: The physical exam is a core component of both undergraduate and postgraduate medical education in Ireland. This is reflected by the time and effort invested by medical schools and medical societies in Ireland in teaching and assessing skills. This high standard of skills results in the physical exam being a key component of the diagnostic process and a gatekeeper to expensive investigations essential in a resource-limited health system such as Ireland. Use of the physical exam in the United States is hindered by the high-tech transformation of healthcare and a more litigious society. Known strategies to highlight the role of the physical exam in clinical practice include creating an evidence base to show that better physical exam skills improve outcomes, identifying accurate physical exam maneuvers, stressing the therapeutic alliance the physical exam brings to the patient encounter, and the incorporation of technology into the bedside exam. Implications: Contrasting the education and clinical use of the physical examination in the United States with Ireland allowed us to identify a number of strategies which could be used to promote the physical exam among learners in both countries. Highlighting simple and pragmatic physical exam maneuvers combined with evidence-based physical exam diagnostic data may renew confidence in the physical exam as a core diagnostic tool. Use of the hypothesis-driven approach may streamline a clinician's physical exam during a patient encounter, focusing on the key examination components and avoiding unnecessary and low yield maneuvers. The absence of assessment of physical exam skills using real patients in United States licensing exams communicates to learners that these skills are not important. However, steps to introduce a culture of assessment to drive learning are being introduced. One area Ireland could learn from the United States is incorporating more technology into the bedside exam. Enhanced physical examination skills in both countries could reduce reliance on expensive investigations and improve diagnostic accuracy.

Acute Myeloid Leukemia: A "Head to Toe" Examination

Acute myeloid leukemia (AML) is a hematologic malignancy that can affect all blood cell lineages, the presentation varies, and infection is a common complication. This case involves a patient initially presenting with a necrotic foot ulcer and leukocytosis, ultimately leading to a diagnosis of osteomyelitis. After establishing adequate source control with serial debridements and intravenous antibiotics, the patient developed some knee swelling. On repeat assessment, he was discovered to have lymphadenopathy, and workup revealed AML. As indicated by this case, though it appeared a clear-cut diagnosis of osteomyelitis, there was an underlying malignancy that would have potentially gone unnoticed due to incomplete clinical examination.

Castleman flare or COPD exacerbation- can biomarkers override availability bias?

Effective treatments for human herpes virus 8 (HHV-8) associated multicentric Castleman disease (MCD) have led to prolonged survival for this complex systemic lymphoproliferative inflammatory disease. Nonetheless, significant challenges remain for the recognition of disease exacerbations, particularly when overlapping with common comorbid conditions. We present a case of a 60-year-old man with a 22-year history of MCD, current advanced COPD, and medication-controlled HIV. His recurrent presentations with flares of fatigue, worsening dyspnea, and productive cough were confusing to clinicians who were attempting to distinguish between exacerbations of MCD or COPD. Published biomarkers of MCD flare include HHV-8 and CRP, which were proposed by the patient to his clinicians as useful in guiding treatment. This case illustrates the value of patient insight as an antidote to the problem of availability bias.

Assessing physical examination skills using direct observation and volunteer patients

Background

Feedback based on direct observation of the physical examination (PE) is associated with enhanced educational outcomes, yet attending physicians do not frequently observe graduate trainees performing the PE.

Methods

We recruited volunteer patients (VPs), each with an abnormality of the cardiovascular, respiratory, or neurological system. Interns examined each VP, then presented a differential diagnosis and management plan to two clinician educators, who, themselves, had independently examined the VPs. The clinician educators assessed interns along five domains and provided post-examination feedback and teaching. We collected data on intern performance, faculty inter-rater reliability, correlation with a simulation-based measure of clinical skill, and resident and VP perceptions of the assessment.

Results

A total of 72 PGY-1 interns from a large academic training program participated. Performance on the cardiovascular and respiratory system was superior to performance on the neurologic exam. There was no correlation between results of an online test and directly observed cardiovascular skill. Interns preferred feedback from the direct observation sessions. VPs and faculty also rated the experience highly. Inter-rater reliability was good for the respiratory exam, but poor for the cardiovascular and neurologic exams.

Conclusions

Direct observation of trainees provides evidence about PE skill that cannot be obtained via simulation. Clinician educators’ ability to provide reliable PE assessment may depend on the portion of the PE being assessed. Our experience highlights the need for ongoing training of clinician educators in direct observation, standard setting, and assessment protocols. This assessment can inform summative or formative assessments of physical exam skill in graduate medical education.

Incoming Interns Recognize Inadequate Physical Examination as a Cause of Patient Harm

INTRODUCTION

As providers of a large portion of the care delivered at academic health centers, medical trainees have a unique perspective on medical error. Despite data suggesting that errors in physical examination (PE) can lead to adverse patient events, we are not aware of previous studies exploring medical trainee perceptions of the relationship between patient harm and inadequate PE. We investigated whether first-year residents at a large tertiary care academic medical center perceive inadequate PE as a cause of adverse patient events.

METHODS

As part of a larger survey given to incoming interns at Thomas Jefferson University Hospital orientation (2014-2018), the authors examined the perceptions of inadequate PE and adverse patient events. We also examined other details related to PE educational experiences and self-reported PE proficiency. The survey was developed a priori by the authors and assessed for face validity by expert faculty.

RESULTS

Ninety-eight percent of respondents (695/706) reported that inadequate PE leads to adverse patient events. Seventy percent (492/706) believe that inadequate PE causes adverse events in up to 10% of all patient encounters, and 30% (214/706) reported that inadequate PE causes adverse events in greater than 10% of patient encounters. Forty-five percent of surveyed interns (319/715) had witnessed a patient safety issue as a result of an inadequate PE. Only 2% of surveyed interns (11/706) did not think patients experience adverse events because of inadequate PEs. Ninety percent of surveyed interns (643/712) reported feeling proficient in performing PE. From 2015 to 2018, 80% (486/604) indicated that they received "just enough" PE education.

CONCLUSION

Nearly all incoming interns surveyed at our institution believe that inadequate PE leads to adverse patient events, and 45% have witnessed an adverse patient event due to inadequate PE. We urge clinicians, educators, and health care administrators to consider enhanced PE skills training as an important and viable approach to medical error reduction, and as such, we propose a 5-pronged intervention for improvement, including a redesign of PE curricula, development of checklist-based assessment methods, ongoing skills training and assessment of physicians-in-practice, rigorous study of PE maneuvers, and research into whether enhanced PE skills improve patient outcomes.

Reclaiming magical incantation in graduate medical education

Critical thinking relies upon conceptualization (what is the possible pathophysiology?), analysis (how do I relate an aberration in physiology to the lived experience of illness?), and synthesizing (how do I best intervene?). These cognitive skills are subsumed in the category of reflective competencies and are necessary for developing a differential diagnosis or a plan of care. A vulnerability of teaching medicine through the filter of heuristics is that it may simply recapitulate the teacher's style of cognitive shortcuts. Poorly calibrated heuristics may culminate in systematic errors of judgment. If the aim is to teach critical reasoning in the arena of clinical education, then a new paradigm is called for. Teaching critical reasoning as it applies to medical decision-making begins with recognizing decision scripts.Key Points• Medical heuristics are high-stakes endeavors.• The process of examining the choice of heuristics employed in any given clinical scenario is a meta-reasoning strategy.• Debiasing reduces cognitive errors due to motivated reasoning.

[Clinical intelligence and artificial intelligence: a question of nuance]

Current artificial intelligence (AI) in medicine has high performance, particularly in diagnostic and prognostic image analysis, but, in everyday clinical practice, evidence-based results of AI remain limited. In this forum, are analyzed the characteristics of clinical intelligence in medical practice, then the successes and promises of AI, as well as the limitations, reservations and criticisms brought to the introduction of AI in the front-line clinic. The importance of certain ethical and regulatory aspects is highlighted, including a "human guarantee" for AI, as suggested by the "Comité consultatif national d'éthique pour les sciences de la vie et de la santé" (National advisory committee on ethics for life and health sciences). Clinical intelligence could be this "human guarantee" of AI in medicine, and their complementarity could lead to a quality of decisions and, ultimately, of care, far higher than that provided, separately, by each of them.

Morning report innovation: Case Oriented Report and Exam Skills

Morning report is a valuable educational conference but is often a stand-alone classroom-based discussion which misses the opportunity for bedside education. In this report, we describe an innovative morning report structure - the Case Oriented Report and Exam Skills (CORES) - that addresses this pitfall of the traditional case conference format and brings learners to the bedside. The key components of CORES include highlighting concepts of clinical reasoning, emphasizing evidence-based and hypothesis-driven physical exam (HDPE), and integrating emerging bedside technologies such as point-of-care ultrasound (POCUS).

Elevating diagnostic skills with logical reasoning

NPs can use the abductive, deductive, and inductive forms of reasoning to adopt a rational and consistent approach to transforming effective data into accurate diagnoses. A case example is used throughout the article to illustrate how these classic logical reasoning skills may be combined with knowledge and experience to address issues of diagnostic accuracy and decrease diagnostic errors.