Performance measures are the specific care choices that are evaluated when judging the quality of care a doctor provides. Current performance measures have known limitations, including failure to personalize care, creating a burden for doctors, and guiding care towards overtreatment. There is currently a push to summarize multiple performance measures into composite scores that measure a clinic's overall care. The best way to accomplish this, however, is unknown. Prevention of cardiovascular disease (CVD) is an ideal case for improving performance measures. CVD has an exceptionally strong clinical evidence base; it is the leading cause of morbidity and mortality in VA; and there has already been extensive work on developing individual performance measures.
In this project, we propose examining alternative approaches to constructing composite performance measures for the medicines of CVD prevention, including statin, blood pressure, and anti-platelet therapies like aspirin. Is it feasible to create a composite that incorporates more clinical nuance, is preferred by VA leadership and frontline Patient Aligned Care Team (PACT) members, and can be communicated in ways that help users improve their performance? This project focuses on CVD prevention, but the model being explored could be generalized to other conditions.
The first part Aim 1 creates the individual and composite measures. As part of honing the measures, the second part of Aim 1 looks at the assessment implications of potentially changing measures when patients have certain specific complicated clinical situations, such as nonadherence. To accomplish these, we will create a dataset of all VHA patients and their risk factor and comorbidity profiles linked to their providers and care facilities and clinical outcomes. This database will provide the data for how care provided would be evaluated by each of the proposed performance measures.
Aim 2 uses the VHA electronic health record (EHR) and the simulation model to learn how different approaches to constructing CVD composite measures might alter reliability, validity, and clinical incentives. The robustness of all results in Aim 2 will be assessed with sensitivity analyses that look at the potential impact of variations in most clinical variables, including the benefit of different treatments, different treatment thresholds, and flaws in the estimates of Atherosclerotic Cardiovascular Disease (ASCVD) risk.
In Aim 3 we will improve the measures created in Aims 1 and 2; examine their content validity and face validity; and create an audit and feedback tool to improve the use of the measures. We will evaluate content validity with a team of performance measure stakeholders - specifically national policy leaders and national and regional clinical leaders. We will improve and test face validity together with practicing clinicians. As part of this work, we will create and evaluate an automated audit and feedback system that will help providers understand and improve their CVD preventive care and performance measures. The work in Aim 3 will occur throughout the project, but will be weighted towards the beginning so it can influence what measures are created and tested in the other aims.
We are still in the development stages, but expect to generate new findings in the forthcoming months once we have begun our qualitative interviews and have completed the assembly of our quantitative cohort dataset.
We are in the final stages of the development stages and thus have not started the quantitative and qualitative components of this project.
- Vance MC, Wiitala WL, Sussman JB, Pfeiffer P, Hayward RA. Increased Cardiovascular Disease Risk in Veterans With Mental Illness. Circulation. Cardiovascular quality and outcomes. 2019 Oct 1; 12(10):e005563.
- DeJonckheere M, Robinson CH, Evans L, Lowery J, Youles B, Tremblay A, Kelley C, Sussman JB. Designing for Clinical Change: Creating an Intervention to Implement New Statin Guidelines in a Primary Care Clinic. JMIR human factors. 2018 Apr 24; 5(2):e19.