|
Abstract:
|
Under the Affordable Care Act, there are increasing needs to identify and develop a core set of health care quality measures to improve the quality of care and evaluate the performance of health related entities across health care system. Such measure development projects typically request tremendous effort and coordination in planning, development, testing, implementation and validation, therefore it requires careful design at the planning stage to get adequate data to ensure the statistical validity of quality measures and reporting. One key task for measure developers is to reliably quantify the scientific soundness (such as data validity and measure reliability) of these measures using data from test sites and report the results to policy makers. Classic formulae using normal based assumptions from asymptotic theory may be either difficult to derive under a complex design, or inappropriate under unbalanced data structure in practice. In this talk, we will present a statistical framework and case studies on how Monte Carlo simulations, as an alternative tool, could provide a more transparent guidance on improving scientific rigor and facilitating decision making.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.