A Dynamic Analysis of Patient Continuity of Care: a Case of New Diabetes Mellitus Patients in Taiwan

Continuity of care (COC) is one of the pillars of primary. Studies found high COC was associated with lower probability of hospitalization and lower expenditures. Few studies have explored the dynamics of patient COC. This study aims to use new diabetes mellitus patients to analyze the dynamics of patient COC and its effect on medical expenditures. In particular, this study use regression analysis to understand how health shock would affect COC and how long the effect persist. This study follows Grossman's human capital model to explain the use of medical care, with a modification with regard to health shock which could increase the degree of patient risk aversion. Data are extracted from the National Health Insurance 200 million sample files. Panel OLS regressions are used to analyzed patient medical expenditures controlling for switching major providers over the 14-year sample period. Hospitalization is considered to be a health shock. Patients ever hospitalized are treated as the treatment group, and propensity score matching is used for choosing control group with similar characteristics at baseline. Regressions are used to test if past hospitalization would affect patient COC and for how long. Confidence intervals for estimated parameters are established to examine if the intervals expand with time.

Project ID:PF10607-1682
External Project ID:MOST106-2410-H182-023