This website uses cookies to enhance your experience in navigating the site, to personalize your academic tutoring, and to contribute to our research in education. By continuing, we understand that you accept the use of cookies. I Accept Cookies policy

NEURAL NETWORKS AND CLINICAL REASONING

Organization of knowledge and diagnostic quality: the neurocognitive track 

Longo analyzes how the organization of neural networks of knowledge affects diagnostic accuracy.

Background academic education and evidence-based medicine provide a firm foundation for building accurate diagnoses, but the foundation has much more: the “mortar” is how knowledge acquired at the neuronal level is structured and organized, allowing it to be recovered from long-term memory by clinicians at the appropriate time and circumstance. The latest findings in neurocognitive sciences play a key role in helping understand these phenomena.

Madrid - March 1st, 2019. Much research on the role of clinical reasoning in diagnosis development points to the importance of organized and available knowledge stored in a clinician’s memory to optimize the diagnostic process, avoiding possible errors. The question is: how is this knowledge organized? What concrete impact this organization has on diagnostic thinking? Moreover, what kinds of relationships connect stored knowledge with clinical judgment and professional expertise?

Cognitive theories have been increasingly pointed out as a support offering a better response to these questions, explaining the existing variability regarding quality and diagnostic accuracy. In this regard, the US National Academy of Sciences recently pointed out a need for a broader approach to be adopted in curricular strategies to improve medical diagnosis, underlining the importance of taking into account the impact of cognition in the decision-making process during clinical reasoning.

This recommendation is in line with the conclusions drawn by works such as the one carried out by Arthur S. Elstein, of the Department of Medical Education at the University of Illinois in Chicago (USA). For him, "It would be positive if doctors were as familiar with the principles of cognitive psychology as they are with those of physiopathology in a way that is clinically useful for their professional practice.”

Adopting this new approach and adapting it to the curricula of healthcare professionals implies closer ties between two fields: research in medical education and current neurocognitive science. To this end, the team of Palma J. Longo, an assistant professor in the Department of Health Sciences of the School of Health Professions at the University of Texas Southwestern in Dallas (Texas, USA), launched an investigation to verify how the degree of coherence in the organization of neural networks of knowledge impacts the fact of a diagnosis being weak or strong.

The meeting point of two methodologies
This work is the first attempt to correlate the functioning of the cognitive processes associated with clinical reasoning using the methodologies of Bordage and Anderson. The method by Georges Bordage, a professor emeritus in the Department of Medical Education of the School of Medicine at the University of Illinois (Chicago, USA), seeks to measure how knowledge is structured and organized based on a semantic analysis of clinical arguments. In contrast, the method by O. Roger Anderson, an expert in neurocognitive theory applied to science learning and professor of Natural Sciences at the University of Columbia (New York, USA), is based on the model of flow diagrams of expressed concepts.

The semantic qualifiers (SQs) described by Bordage are terms (nouns and adjectives) referring to the conceptualization and abstraction that clinicians make out of the signs and symptoms presented by the patient, allowing them to compare and contrast diagnoses. For example, “it started suddenly two weeks ago” translates into the SQ as an acute onset. Anderson, in turn, opts for a graphical representation of sequential links and multirelational links of recursive neural networks. That is, the number of times a clinical concept stored in memory is repeated during a narrative sequence to solve a problem and the associations between them.

For Bordage, the use of SQs allows for more precise and accurate diagnoses than the use of signs and symptoms alone. Anderson assumes that the interpretation of the behavior of neural networks sheds much light on the approach and solution of a scientific problem (a diagnosis, in the case of clinicians) and on how knowledge is organized in the memory.

The more structured, the better diagnosed
Longo's study included 12 postgraduate students who analyzed a clinical case of peptic ulcer and then answered structured questions (TAL). Their answers were coded and studied based on three competencies involved in clinical reasoning: semantics and diagnostic capacity, using the Bordage methodology, and organization of knowledge networks, according to the Anderson method.

The first two analyzed the number and quality of the diagnostic arguments by dividing the participants into two groups: on the one hand, students with poor semantic thinking, weak diagnostic thinking, and reduced or dispersed discourse, and on the other hand, those with a semantically rich, elaborated, or compiled discourse who manifested a diagnostic thought that was catalogued as strong. Subsequently, using the Anderson diagram, they analyzed the neural network (recursive) variables involved in knowledge organization, activated during the recall by the respondents.

By correlating the variables of the diagnostic competencies and those of the knowledge organization network, they found that the more organized and better structured this network was, the higher the quality of the diagnosis. This suggests that it is easier and more effective for clinicians to recall knowledge more coherently linked and ordered in long-term memory. The conclusions of the study also show that this structure improves the mobilization and interconnection between the neural networks involved in memory.

For Longo, the study has limitations (it was based on the analysis of a single clinical case, and the cohort of participants was small), but lays the foundation for future research aimed at achieving a closer link between neurocognitive science and research in medical education, so that both fields are aligned with the objective of better understanding the complex process of clinical reasoning.

Henceforth, this approach could be implemented in clinical skills training programs, since, as the study director says, "the ability to specifically identify where, when, and how a student uses acquired clinical knowledge can provide valuable information to detect the strengths and weaknesses in the processes of clinical reasoning.”

 

References

Longo P, Orcutt V, James K et al. Clinical Reasoning and Knowledge Organization: Bridding the Gap Between Medical Education and Neurocognitive Science. J Physician Assist Educ 2018; 29 (4): 230-235. doi: 10.1097/JPA.0000000000000224

Elstein A. Thinking about diagnostic thinking: a 30-year perspective. Adv in Health Sci Educ 2009; 14:7-18. doi: 10.1007/s10459-009-9184-0

Bordage G, Lemieux M. Cognitive Structures of Experts and Novices. Semantic Structures and Diagnostic Thinking of Experts and Novices. Acad Med 66, Suplement (September 1991):S70-S72

Anderson R, Demetrius O. A flow-Map Method of Representing Cognitive Structure Based on Respondent’s Narrative Using Science Content. Journal of Research in Science Teaching 1993; 30:953-969

Script Connect
Personalized help center

Our personalized help center enables you to obtain technical support and help for navigating through the site and using the program.

Frequently asked questions