Innovative analytical methods for longitudinal medical cost data: Translating research to inform policy and practice

Rising health care cost is a major concern for health policy makers since 1960s. Sisko et al. (2009) projected that health care costs per person will increase from $8,160 in 2009 to $13,100 in 2018, and that total health care costs will account for over 20% of the gross domestic product by 2018. Statistical analysis of medical cost data is becoming increasingly important with heightened interests in containing healthcare cost. In order to better understand the factors associated with the growth in medical expenditures, it is important to study the longitudinal history of medical care cost data. The research provided in this book chapter is important for developing several new methods to enhance our understanding of how and why health care expenditures may change over time for an individual. It provides an excellent discussion of the contemporary issues in modeling medical costs, as well as new statistical models which are being developed for solving ongoing problems. The proposed research design is testing refinements/improvements to econometric models for analyzing longitudinal data on health care costs. The objective of this research is to develop and disseminate a number of models of medical costs in a longitudinal statistical framework that allows for the measurement of costs at regular intervals instead of only as a total (or cross-sectional cost). Additionally, this book chapter will emphasize practical application and policy of the methods and models presented. The specific aims are as follows: 1) to provide an overview of the currently available econometric and statistical models of medical costs to cross-sectional as well as longitudinal data; 2) to introduce the method of using more flexible functional forms of covariate specification (e.g., splines for non-linear temporal effects) in modeling longitudinal medical cost data.; 3) to describe an extension of the above flexible model to jointly analyze medical costs and multiple health outcomes (e.g., survival, or quality of life), and study the effect of risk factors on them simultaneously; 4) to discuss another extension that applies hierarchical models to address the clustering effect in modeling longitudinal medical cost data at different levels, e.g., health plans, families, and members; and 5) to present examples of practical application of the developed methods. The researchers will apply these new methods to three real-world databases: the Clinical Data Repository (CDR) at the University of Virginia Health System, the Medical Expenditure Panel Survey (MEPS), and the SEER-Medicare databases.