BACKGROUND/RATIONALE:
Large integrated health systems like the Veterans Health Administration (VHA) need methods of performance assessment that identify facilities with intermediate patient outcomes falling outside targeted ranges. While summary measures have been advocated for this purpose, those reflecting the management of chronic diseases such as diabetes are complex and need to rely on methods that are not based on achieving thresholds of care but rather acknowledge improvement across a continuum.
OBJECTIVE(S):
The short-term objectives of this proposal are to 1) establish the comparability of a threshold-based approach using administrative data to current performance measurement derived from EPRP chart abstraction, 2) compare facility-level performance using a threshold- and continuum-based approach derived from administrative data alone, and 3) assess the incremental informational value and potential tradeoffs of a simulation-based approach compared with one that is continuum-based. The long-term objective of this proposal is to provide managers, planners, and VISN and facility directors with a means of assessing population health and better focusing quality improvement efforts.
METHODS:
EPRP and DEpiC - a large epidemiological cohort of veterans with diabetes - will provide the data from which we will create measures to assess facility-level performance using three specific approaches (i.e., threshold-, continuum-, and simulation-based). Cross-sectional data samples will be obtained from both data sources for specified time periods (FY03 and FY04) and these will be used to evaluate facility-level performance. All veterans primarily receiving their health care within VHA with a diagnosis of diabetes will be included. We will identify all veteran clinical users during relevant timeframes with diabetes in DEpiC, and all veterans with diabetes in the EPRP data sets corresponding to the same timeframes. Based on recent estimates, we anticipate that the number of eligible persons with diabetes for the relevant timeframes will be around 875,000 and that we should have at least 135 facilities with 500 patients with diabetes at each. Facilities will then be ranked according to their performance based on the relevant methods and rankings will be compared using the kappa statistic. Assessment of within-facility performance will use the quadrant analysis methodology outlined by the Office of Quality and Performance.
FINDINGS/RESULTS:
None at this time.
IMPACT:
This proposal will have immediate and direct implications on the ability of VHA managers and leaders to more accurately monitor and document the quality of care delivered to veterans, using an approach that is sustainable.
PUBLICATIONS:
Journal Articles
- Terris DD, Litaker DG. Data quality bias: an underrecognized source of misclassification in pay-for-performance reporting? Quality Management in Health Care. 2008; 17(1): 19-26.
