|
Abstract:
|
Health reform has focused on improving outcomes and containing costs. In patient-centered outcomes research, reform's impacts are typically estimated from administrative claims data, which reflect costs and certain clinical outcomes. For example, innovative programs under the Affordable Care Act are being tested and evaluated with Medicare claims databases (innovation.cms.gov). Yet, claims data, originally designed to pay providers, do not directly reflect the consumer's financial decisionmaking to access care. With the Consumer Expenditure Survey (CE), we holistically view health reform from patient's role as a consumer. Moving to consumer centeredness, we link the CE surveys and others that measure actual spending, such as the AHRQ Medical Expenditures Panel Survey, through geocodes and other mappings. Through blocking on respondent characteristics, we control for observed confounders. We apply Bayesian methods to pool estimates across respondent subgroups. Graphical posterior probabilities illustrate how the individual as both patient and consumer makes decisions based on spending and health outcomes.
|
