Background: Empathy is a human emotion that is important in the effective provision of health care and amenable to change through explicit and implicit experiences in an individual's life. This study measured levels of empathy in students pursuing doctoral degrees in physical therapy and compared the influence of professional education at different institutions on these levels.
Methods: Our cross-sectional, two-cohort, multisite study used a modified version of the Jefferson Scale of Physician Empathy, Student Version, to investigate empathy levels at enrollment, mid-curriculum, and end-of-curriculum. Statistical tests of differences were performed between institutions, within institutions for each cohort across the three time points, and within institutions between cohorts. Data were analyzed using descriptive statistics, ANOVA, and the least squared difference test. Alpha was set at 0.05 for main test of difference and 0.04 for all post-hoc tests.
Results: For both cohorts, empathy levels differed significantly between institutions at program entry (Cohort 1, p=0.0150; Cohort 2, p=0.0273); within institutions the two cohorts were similar at the beginning of the first semester. In Cohort 1, no significant changes occurred within any institution; students at the two institutions with higher entering scores maintained their higher scores at the end of the last didactic semester. Students in Cohort 2 showed significant differences in empathy levels at the end of the last didactic semester within and between institutions (p=0.0251; p<0.0001).
Conclusions: Empathy levels may differ at enrollment for PT students at different institutions even with similar recruitment approaches and no significant differences in student demographics between institutions. Despite uniform accreditation requirements for curriculum content, significant differences between institutions did exist in the last didactic semester in Cohort 2 but not Cohort 1. The direction and magnitude of such changes were not explained by institutional characteristics. This study challenges assumptions that measurements of empathy in students at one institution can be generalized to students at other institutions and that one cohort in the same institution can predict another cohort.