Medical student, nursing student, and non-health care respondents' implicit attitudes toward doctors and patients: Development and a pilot study of a new implicit attitudes test (IAT)

INTRODUCTION Medical educators have been concerned that medical students may decline in empathy for patients during the course of their training, based on studies measuring clinical empathy using psychometrically strong self-report measures. Clinical empathy is a complex construct, incorporating attitudes toward patients but also other components, such as professional detachment. Triangulation of extant measures with instruments based on nonreactive methods could provide a better understanding of whether and how physician attitudes toward patients may be changing during training. We sought to develop and pilot-test such a nonreactive method. METHODS We develop variations of an implicit association test (IAT) designed to measure attitudes toward physicians and patients based on speed of reaction to images of actors and positive and negative words. In the IATs, the same actors are photographed as doctors, clinic outpatients, hospitalized inpatients, and as a "general public" control. We examine preliminary evidence for their validity by collecting pilot data from internet participants (not involved in the health professions), medical students, and nursing students. RESULTS Internet participants (n = 314) and nursing students (n = 31) had more negative associations (IAT scores) with doctors than did medical students (n = 89); nursing students and female internet participants had more positive associations with hospitalized patients than did medical students and male internet participants. Medical students' associations with hospitalized patients varied by year of training. DISCUSSION This IAT may provide insight into implicit attitudes among those who enter training for the health profession and changes in those attitudes that may be inculcated during that training.