Testing two competitive models of empathic communication in cancer care encounters: A factorial analysis of the CARE measure.

Objective The mechanisms associating physician empathy (PE) with patient outcomes remain unclear. PE can be considered as a whole (one process) or three subcomponents can be identified (an establishing rapport process; an emotional process; a cognitive process). The objective was to test two competitive models of PE in cancer care: a three-process model adapted from Neumann's model versus a one-process model, with the use of the Consultation and Relational Empathy measure (CARE). Methods The CARE was completed by 488 oesogastric cancer patients from the national French database FREGAT. A confirmatory factor analysis (CFA) and a bifactor model were performed to test the two competitive models. Results The CFA revealed that the one-factor structure showed a moderate fit to the data whereas the three-factor structure showed a good fit. However, the bifactor model favoured unidimensionality. Conclusion We cannot provide a clear-cut conclusion about whether PE should be considered as on unique process or not. Further work is still needed. Meanwhile, one should not preclude the use of three subscores in cancer care if specific elements of the encounter need to be assessed.