Computer analysis system of the physician‐patient consultation process

Abstract

Measurements of the quality of physician-patient communication are important in assessing patient outcomes, but the quality of communication is difficult to quantify. The aim of this paper is to develop a computer analysis system for the physician-patient consultation process (CASC), which will use a quantitative method to quantify and analyze communication exchanges between physicians and patients during the consultation process. CASC is based on the concept of narrative-based medicine using a computer-mediated communication (CMC) technique from a cognitive dialog processing system. Effective and ineffective consultation samples from the works of Saito and Kleinman were tested with CASC in order to establish the validity of CASC for use in clinical practice. After validity was confirmed, three researchers compared their assessments of consultation processes in a physician's office with CASCs. Consultations of 56 migraine patients were recorded with permission, and for this study consultations of 29 patients that included more than 50 words were used. Transcribed data from the 29 consultations input into CASC resulted in two diagrams of concept structure and concept space to assess the quality of consultation. The concordance rate between the assessments by CASC and the researchers was 75 percent. In this study, a computer-based communication analysis system was established that efficiently quantifies the quality of the physician-patient consultation process. The system is promising as an effective tool for evaluating the quality of physician-patient communication in clinical and educational settings.