Application of social network analytics to assessing different care coordination metrics

Abstract

Social network analytic approaches have been previously proposed to identifying key metrics of physician care coordination. Optimizing care coordination is a primary national concern for which yields significant cuts in medical care costs. However, the proposed metric-termed ‘care density’ for estimating care coordination is not completely accurate. Our objective is to compare the accuracy of the previously proposed ‘care density’, with our proposed ‘weighted care density’, ‘time varying care density’, and ‘time varying weighted care density’ in terms of predicting the cost of care. Our proposed metrics are based on the former care density, however, takes other variables into consideration, mainly patient hospitalization time frames and number of physician visitations. Our findings suggest that physicians coordinating over short time spans spike the cost of care above normal.