Complexities of perceived and actual performance in pathology interpretation: A comparison of cutaneous melanocytic skin and breast interpretations
Patricia A. CarneyFrederick A. SpencerLisa M. ReischLinda TitusStevan R. KnezevichMartin A. WeinstockMichael W. PiepkornRaymond L. BarnhillDavid E. ElderDonald L. WeaverJoann G. Elmore
2
Citation
20
Reference
10
Related Paper
Citation Trend
Abstract:
Little is known about how pathologists process differences between actual and perceived interpretations.In teaching clinical problem solving the probabilistic component should be emphasized. In the domain of empirical medical knowledge to be utilized in diagnostic strategies the prediction according to clinicians' subjective probabilities is used in medical decision making concerning patient problem definition and diagnosis. Two approaches in teaching differential diagnosing by decision models using short case studies are compared. The first is traditional one with the concept of one "true" diagnosis selected from proposed diagnoses group in a MCQ-test. The second approach reflects more the medical practice reality: the students have to predict quantified subjective probabilities of each diagnosis in a suggested diagnoses group considering all the diagnoses probable and not excluding each other. The advantages of the second procedure are evaluated.
Medical Decision Making
Cite
Citations (1)
Purpose Most diagnostic errors involve faulty diagnostic reasoning. Consequently, the authors assessed the effect of querying initial hypotheses on diagnostic performance. Method In 2007, the authors randomly assigned 67 first-year medical students from the University of Calgary to two groups and asked them to diagnose eight common problems. The authors presented the same primary data to both groups and asked students for their initial diagnosis. Then, after presenting secondary data that were either discordant or concordant with the primary data, they asked students for a final diagnosis. The authors noted changes in students' diagnoses and the accuracy of initial and final diagnoses for discordant and concordant cases. Results For concordant cases, students retained 84.2% of their initial diagnoses and were equally likely to move toward a correct as incorrect final diagnosis (6.9% versus 8.9%, P = .3); no difference existed in the accuracy of initial and final diagnoses: 85.9% versus 84.0% (P = .4). By contrast, for discordant cases, students retained only 23.3% of initial diagnoses, change was almost invariably from incorrect to correct (76.3% versus 0.4%, P < .001), and final diagnoses were more accurate than initial diagnoses: 80.7% versus 4.8% (P < .001). Overall, no difference existed in the accuracy of final diagnoses for concordant and discordant cases (P = .18). Conclusions These data suggest that querying an initial diagnostic hypothesis does not harm a correct diagnosis but instead allows students to rectify an incorrect diagnosis. Whether querying initial diagnoses reduces diagnostic error in clinical practice remains unknown.
Concordance
Cite
Citations (47)
Abstract Rationale, aims and objectives Diagnostic uncertainty is often encountered in a medical practice. Patients with ambiguous, uncertain, and undiagnosed problems are frequently referred for second opinions. Comparing referral diagnoses to final diagnoses provides an opportunity to determine how frequently final diagnoses vary and changes the direction of medical care. Methods A retrospective study was done at a single academic medical center using a sample of 286 patients referred by physician assistants, nurse practitioners, and physicians from primary care practices from January 1, 2009 to December 31, 2010. Patients' referral and final diagnoses were compared and classified into 1 of 3 categories: referral diagnosis and final diagnosis the same, referral diagnosis better defined/refined, and referral diagnosis distinctly different from final diagnosis. Episode costs for the respective categories were calculated for the referral visit and services that occurred at our facility within the first 30 days. Results In 12% (36/286) of cases, referral diagnoses were the same as final diagnoses. Final diagnoses were better defined/refined in 66% (188/286) of cases; but in 21% of cases (62/286), final diagnoses were distinctly different than referral diagnoses. Total costs for cases in category 3 (different final diagnoses) were significantly higher than costs for cases in category 1 ( P = .0001) and category 2 ( P = <.0001). Conclusion Referrals to advanced specialty care for undifferentiated problems are an essential component of patient care. Without adequate resources to handle undifferentiated diagnoses, a potential unintended consequence is misdiagnoses resulting in treatment delays and complications leading to more costly treatments.
Specialty
Cite
Citations (68)
The reliability of medical record information is of fundamental importance to the certainty with which a diagnosis can be made. 40 patients were chosen at random and each was examined by four clinicians. The information and a tentative diagnosis were written on a special record form. The results were judged by means of the coefficient kappa. The clinicians disagreed more on symptoms than on diagnoses. The diagnoses made by an automatic diagnosis system showed lower precision and lower accuracy than the clinicians’ diagnoses. The results of the study might explain why computer assistance in diagnostics is of limited value.
Cohen's kappa
Medical record
Medical History
Kappa
Acute abdominal pain
Cite
Citations (30)
Although making accurate diagnoses is a critical clinical skill, putting accurate diagnoses on the face sheet of a medical record does not have the same priority for most physicians. Even those who see recording a correct principal diagnosis as an important part of good clinical thinking are unlikely to be perfectionists regarding additional diagnoses because it is hard to see that they will have any effect on patient care. See also p 2197. In the last few years, however, major journals have published a growing number of articles in which recorded additional diagnoses play an important part in analysis, either to establish a clinical point or to assess additional diagnoses as a measure of clinical severity of illness. Investigators seek to secure information from additional diagnoses, which would otherwise require costly manual abstraction, about the occurrence of adverse events and the risk of outcomes such as death. In addition, hospital
Medical record
Cite
Citations (9)
Abstract Background Medical history contributes approximately 80% to a diagnosis, although physical examinations and laboratory investigations increase a physician’s confidence in the medical diagnosis. The concept of artificial intelligence (AI) was first proposed more than 70 years ago. Recently, its role in various fields of medicine has grown remarkably. However, no studies have evaluated the importance of patient history in AI-assisted medical diagnosis. Objective This study explored the contribution of patient history to AI-assisted medical diagnoses and assessed the accuracy of ChatGPT in reaching a clinical diagnosis based on the medical history provided. Methods Using clinical vignettes of 30 cases identified in The BMJ , we evaluated the accuracy of diagnoses generated by ChatGPT. We compared the diagnoses made by ChatGPT based solely on medical history with the correct diagnoses. We also compared the diagnoses made by ChatGPT after incorporating additional physical examination findings and laboratory data alongside history with the correct diagnoses. Results ChatGPT accurately diagnosed 76.6% (23/30) of the cases with only the medical history, consistent with previous research targeting physicians. We also found that this rate was 93.3% (28/30) when additional information was included. Conclusions Although adding additional information improves diagnostic accuracy, patient history remains a significant factor in AI-assisted medical diagnosis. Thus, when using AI in medical diagnosis, it is crucial to include pertinent and correct patient histories for an accurate diagnosis. Our findings emphasize the continued significance of patient history in clinical diagnoses in this age and highlight the need for its integration into AI-assisted medical diagnosis systems.
Medical History
Clinical history
Clinical Diagnosis
Medical record
Past medical history
Cite
Citations (6)
Background: Uncoded diagnoses in health insurance claims (HICs) may introduce bias into Japanese health statistics dependent on computerized HICs. This study's aim was to identify the causes and characteristics of uncoded diagnoses.Methods: Uncoded diagnoses from computerized HICs (outpatient, inpatient, and the diagnosis procedure-combination per-diem payment system [DPC/PDPS]) submitted to the National Health Insurance Organization of Kumamoto Prefecture in May 2010 were analyzed. The text documentation accompanying the uncoded diagnoses was used to classify diagnoses in accordance with the International Classification of Diseases-10 (ICD-10). The text documentation was also classified into four categories using the standard descriptions of diagnoses defined in the master files of the computerized HIC system: 1) standard descriptions of diagnoses, 2) standard descriptions with a modifier, 3) non-standard descriptions of diagnoses, and 4) unclassifiable text documentation. Using these classifications, the proportions of uncoded diagnoses by ICD-10 disease category were calculated.Results: Of the uncoded diagnoses analyzed (n = 363 753), non-standard descriptions of diagnoses for outpatient, inpatient, and DPC/PDPS HICs comprised 12.1%, 14.6%, and 1.0% of uncoded diagnoses, respectively. The proportion of uncoded diagnoses with standard descriptions with a modifier for Diseases of the eye and adnexa was significantly higher than the overall proportion of uncoded diagnoses among every HIC type.Conclusions: The pattern of uncoded diagnoses differed by HIC type and disease category. Evaluating the proportion of uncoded diagnoses in all medical facilities and developing effective coding methods for diagnoses with modifiers, prefixes, and suffixes should reduce number of uncoded diagnoses in computerized HICs and improve the quality of HIC databases.
Medical record
Gold standard (test)
Cite
Citations (4)
Abstract Background: Diagnostic error can lead to increased morbidity, mortality, healthcare utilization and cost. The 2015 National Academy of Medicine report “Improving Diagnosis in Healthcare” called for improving diagnostic accuracy by developing innovative electronic approaches to reduce medical errors, including missed or delayed diagnosis. The objective of this article was to develop a process to detect potential diagnostic discrepancy between pediatric emergency and inpatient discharge diagnosis using a computer-based tool facilitating expert review. Methods: Using a literature search and expert opinion, we identified 10 pediatric diagnoses with potential for serious consequences if missed or delayed. We then developed and applied a computerized tool to identify linked emergency department (ED) encounters and hospitalizations with these discharge diagnoses. The tool identified discordance between ED and hospital discharge diagnoses. Cases identified as discordant were manually reviewed by pediatric emergency medicine experts to confirm discordance. Results: Our computerized tool identified 55,233 ED encounters for hospitalized children over a 5-year period, of which 2161 (3.9%) had one of the 10 selected high-risk diagnoses. After expert record review, we identified 67 (3.1%) cases with discordance between ED and hospital discharge diagnoses. The most common discordant diagnoses were Kawasaki disease and pancreatitis. Conclusions: We successfully developed and applied a semi-automated process to screen a large volume of hospital encounters to identify discordant diagnoses for selected pediatric medical conditions. This process may be valuable for informing and improving ED diagnostic accuracy.
Pediatric emergency medicine
Diagnosis code
Pediatric hospital
Cite
Citations (8)
Cite
Citations (10)
A number of researchers from Johns Hopkins Hospital report that 40%-80% of chronic pain patients are misdiagnosed. Previous reports indicate an on-line questionnaire, The Diagnostic Paradigm and Treatment Algorithm, provides diagnoses with a 96.3% correlation with diagnoses of Johns Hopkins Hospital staff members, in patients with chronic back, neck or limb pain. This research was undertaken to determine if diagnoses generated by the Diagnostic Paradigm and Treatment Algorithm could be confirmed by irrefutable indications of pathology, i.e. intra-operative findings. Prior to surgery, the Diagnostic Paradigm and Treatment Algorithm was administered to ten patients. The Diagnostic Paradigm predicted 61/61 (100%) diagnoses which were confirmed intra-operatively. The Diagnostic Paradigm had 71 false positive diagnoses, but these were part of the differential diagnoses of the correct diagnoses. These differential diagnoses were refined by medical testing.
Cite
Citations (7)