The clinical reasoning process: A study of its development in medical students

Introduction: Clinical reasoning plays a crucial role in the provision of quality medical care. The consequences of poor clinical reasoning affect not only the patient concerned but the cost and effectiveness of the entire health care system. The aims of this thesis are three-fold: to monitor the development of clinical reasoning in students enrolled in a graduate-entry, problem-based learning program in a large medical school; to compare the clinical reasoning of these students at graduation with that of graduates of a traditional undergraduate course; and to gain insight into the nature of the clinical reasoning process which may be applicable to the teaching and evaluation of medical curricula. Methods: Two methods have been used to assess clinical reasoning. The first method, the Clinical Reasoning Problems (CRPs), was developed to meet specific criteria established for this study. A CRP is a short clinical scenario that provides information about a patients presentation, history and physical examination. It is designed to assess the first three steps in the clinical reasoning process identification and interpretation of pertinent data, and hypothesis generation. Subjects are required to nominate the two most likely diagnoses and to identify and interpret the clinical features that they considered in arriving at each diagnosis. Responses are marked using a scoring system based on, and derived from, those of a reference group of experienced general practitioners (GPs). The CRPs are presented as a set of ten problems, each taking about ten minutes to complete. The second method used was the Diagnostic Thinking Inventory (DTI). This is a questionnaire designed to probe clinical reasoning style by assessing two aspects of clinical reasoning: flexibility of thinking and the degree of structure of knowledge in memory. In this way, and because it is not tied to a specific clinical scenario, it provides an assessment of reasoning style that is independent of the content of a subjects knowledge base. Subjects: Subjects were volunteer medical students from the last cohort of the undergraduate course and the first three cohorts of the new, graduate-entry MBBS Program at the University of Queensland. The reference group consisted of twenty-two GPs voluntarily recruited from all areas of Queensland. Results: Evaluation of the CRPs showed them to be a reliable and valid method for the assessment of clinical reasoning skill over a range of competence levels. The reliability of the DTI was similar to that found in other studies. Analysis of data collected using these methods showed that scores on both the CRPs and the DTI increased significantly in all student cohorts throughout the MBBS Program. Factors which influence performance on one or other of these indicators include stage of progression through the MBBS, gender, academic background and pre-admission interview score. Analysis of the causes of poor performance on the CRPs strongly suggests that medical students in the latter part of the MBBS have reached an intermediate stage in the development of reasoning expertise. This stage is characterised by the possession of a large but poorly structured body of knowledge. Subsequent examination of the CRP responses of the 21 reference GPs led to the classification of six of the GPs as diagnostic experts on the basis of their response patterns. Comparison of these GPs with the remaining 15 in the reference group has provided valuable information about the clinical reasoning characteristics of diagnostic experts and the relationship of the clinical reasoning process to diagnostic expertise. The principal finding was that integration of clinical data - the ability to connect relevant, accurately interpreted clinical information in such a way as to lead to the correct diagnosis - is the most critical step in the clinical reasoning process. This step is commonly disregarded or under-emphasised in the literature. Conclusion: This thesis indicates that both the style, and level of skill, of clinical reasoning develops significantly in MBBS students throughout their undergraduate medical education. In addition, it has demonstrated that the ability to integrate clinical information is a distinguishing characteristic of the clinical reasoning process of diagnostic experts. Although it is accepted that further development into true expertise can only occur as a result of extensive independent clinical experience, these findings are highly relevant to both the teaching and evaluation of medical curricula