Problem Solving in the Clinical Interview: A Cognitive Analysis of the Performance of Physicians, Residents, and Studentsmx

This study evaluates differences between subjects at three levels of expertise in acquiring and using information obtained from a patient during the clinical interview. Endocrinologists, residents, and medical students interviewed a volunteer endocrine outpatient and provided differential diagnoses. The transcribed doctor-patient dialogues were analyzed using cognitive methods of analysis. The methods are based on a cognitive framework for studying doctor-patient interaction. The primary focus was on two units of knowledge acquisition: observations and findings. Observations are information units acquired from the patient's reports. Findings are more abstract units composed of clinically significant observations. Several variables based on these knowledge units were identified. Differences were found between groups in the accuracy of diagnoses and in the differential acquisition of information from the patient during the interview. These differences were manifested most clearly in terms of a series of efficiency measures designed to characterize the selective use of information in solving the problem. The endocrinologists elicited findings with significantly greater efficiency than the other two groups and used these findings to correctly diagnose the clinical case. In addition, several of the measures were significantly correlated with diagnostic accuracy. The implications for the assessment of clinical competency are discussed.