Medicare DeSYNPuf analysis Gamma regression cost modeling & risk adjustment, and multiple factor analysis exposing factors associated with health care spending by Kyle Rasku and Dhara Kapoor
Dhara and I have improvements planned for this project including:
future Structural Equation Modeling (SEM) to examine relationships between factors, particularly theororized causal relationships
future modeling of both visit counts and cost data utilizing mixed effects modeling with CBSA as the fixed effect (initial GLMER models are in the "New" folder)
possible addition of SDOH data into the modeling to improve latent factor reliability
utilization of real Medicare data from a more recent time point to validate initial findings / factors discovered using synthetic, de-identified data