Contact PD/PI: Johnston, Kenton James Project Summary The Medicare program is implementing value-based payment (VBP), in which outpatient clinicians are paid based on the quality and cost of the care they deliver. However, there is increasing concern these VBP programs may not accurately measure performance among clinicians who serve high-risk patient populations, particularly those with Alzheimer’s disease and related dementias (hereafter “dementia”). Prior research finds that dementia is independently associated with poorer patient outcomes and markedly higher costs of care. Furthermore, patients with dementia tend to be clustered among certain types of clinicians, such as geriatricians. However, although Medicare adjusts clinicians’ performance under VBP for the medical risk of their patients using the Centers for Medicare and Medicaid Services’ hierarchical condition category (CMS- HCC) risk model, it does not adjust for dementia. This could result in systematic under-estimation of performance for clinicians that disproportionately serve patients with dementia. Therefore, there is a critical need for research on the effect of patient dementia on measures of clinician performance under VBP, how this impacts providers who disproportionately serve such patients, and whether measures of dementia can be integrated into the CMS-HCC Medicare risk adjustment model. Our long-term goal is to incorporate measures of patient dementia into the risk adjustment system that Medicare uses in implementing VBP. The objectives of this R21 application are to: 1) validate the adequacy of claims-based measures of dementia; 2) evaluate the performance of a claims-based costs of care model that includes dementia by adding it to the standard CMS- HCC risk model; and 3) demonstrate the effects of our new model versus the standard CMS-HCC model on clinicians’ relative performance on cost measures in MIPS and the program more broadly. Our rationale for this project is that most clinicians in the U.S. will be held accountable for their performance on cost measures under VBP, and there is a significant risk that clinicians who disproportionately serve patients with dementia may be negatively impacted if these factors are not accounted for by risk adjustment. This research study will pursue three specific aims. For aim #1, we will conduct a retrospective cohort study using the Medicare Current Beneficiary Survey (MCBS) linked to prescription drug and claims data to validate the use of claims- based measures of dementia against patient self-report and prescription drug data. For aim #2, we will use the most recently available Medicare claims data from the CMS Virtual Research Data Center (VRDC) to identify the portion of Medicare costs attributable to dementia that are currently unexplained by the CMS-HCC risk model, and evaluate a model that improves prediction by adding risk points for dementia patients. For aim #3, we will use the CMS VRDC and Physician Compare data to as...