Quality and Financial Status of Critical Access Hospitals: A Preliminary Analysis
UMN will specify the quality measures to be used and extract the data from the Hospital Compare database. We will also assess the feasibility of obtaining and using other sources of quality data for the analyses. The two centers will also define and assemble other relevant community level data and measures. UMN and UNC will jointly develop hypotheses related to the finance-quality link and statistical models suitable for testing hypotheses. We will consider jointly modeling multiple outcomes and/or controlling for unobservable characteristics driving both quality and financial measures. Possible statistical techniques to be used include longitudinal, hierarchical, Bayesian, and spatial extensions. The analytic plan for the project will address the following questions:
- How are the indicators associated? We will begin by examining simple bivariate associations and extend into multiple regression. Potential non-linearity will also be explored because previous research in larger hospitals has found evidence of this type of relationship between financial performance and quality measures.
- Are the different measures capturing similar hospital characteristics? Factor analytic techniques will be used to discover common elements among the quality and financial indicator measures.
- What is the "timing" of association? That is, are the measures contemporaneously correlated, or does a change in one indicator lead to changes in subsequent years in another indicator?
- Is any empirical association due to correlation in factors unobserved by the analyst (such as managerial ability)? We will explore this by state-of-the-art methods accounting for unobserved heterogeneity.
- To what extent are relationships affected by hospital characteristics? We will explore whether CAH characteristics such as size influence the quality-finance relationship.
Staff from the two centers will perform preliminary analyses of data and present the results to the expert workgroup for their review. A report describing the results of the analyses and plans for continued analyses in Year 5 will be submitted to ORHP at the end of Year 4.
There may be products related to this project; please contact the lead researcher for more information.