The Impact of Judge Selection on Standard Setting for a Patient Survey of Physician Communication Skills

Abstract

Patient surveys such as the Communication Assessment Tool (CAT) are recommended to assess physicians. Optimal methods to select judges and set a minimum passing standard (MPS) for a patient survey are unknown. Fifty-eight judges in five groups provided item-based (Angoff) and group-based (Hofstee) judgments for the CAT on two occasions. Judges were communication experts, program directors, trainees, and patients (two groups). Ratings were used to calculate MPSs for the CAT. Interrater reliability and test-retest reliability (stability) were estimated. MPSs were applied to data from a CAT pilot study. Judges produced different MPSs depending on their background. Patients and communication experts were most stringent in their MPS decisions; program directors and trainees were most lenient. Failure rates from the pilot study ranged from 0% to 47%. Judge expertise and background dramatically influenced MPSs for a patient survey of physician communication skills. Judge selection is a key decision when setting an MPS for a clinical skills evaluation.