Patient-centered health care delivery systems in the US: a Discrete-Choice Experiment

Abstract

Objectives: Patient-centered care, in which health services are customized on the basis of patients’ needs and values, is seen as a critical factor in a high-performance health care system. This project seeks to characterize patients’ needs and values for specific features of health care delivery systems. Methods: Quantitative data were obtained by means of a discrete choice experiment (DCE) using a logit model. Alternatives were described by 21 specific attributes that described certain features of a health care delivery system. Each set included six attributes with three specific levels. The DCE was divided into four blocks based on thematic mapping (DCE 1, patient involvement; DCE 2, point of care; DCE 3, personnel; DCE 4, organization). Each DCE included six attributes with three specific levels. Respondents were randomly assigned to an DCE. Results: N=3900 respondents (patients) have completed the survey at Duke HealthView. The feature “out-of-pocket costs” was of high importance across all 4 DCEs (DCE-1coefficient: 0.7050; DCE-2coefficient: 0.9057; DCE-3coefficient, 0.7330; DCE-4coefficient: 0.8384). In DCE 1 regarding patient involvement, “trust and respect” (0.6538) and “attention to personal situation” (0.4732) were of greatest importance. In DCE 2 addressing preferences at the point of care, “shared decision making” (0.7441) and “access to patient record” (0.4576) were nearly equally valuable to patients but of highest relevance. In DCE 3 focusing on personnel in health care delivery systems, “multidisciplinary care” (0.7308) was ranked highest. Lastly, in DCE 4 analyzing features of the organization of health care delivery systems, “medical devices and furnishings” (0.4082) and “treatment guidelines” (0.4702) were of almost equal value to patients.