Abstract Body (Do not enter title and authors here): Introduction
Veterans Affairs (VA) facilities evaluate cardiology clinicians using three performance measures for outpatient care: anticoagulation for atrial fibrillation or atrial flutter (AF) with elevated stroke risk, statin therapy for coronary artery disease (CAD), and renin-angiotensin-aldosterone system (RAAS) inhibition for heart failure (HF). We sought to assess these metrics and their differences by clinician characteristics using VA health record data.

Methods
We assessed AF, CAD, and HF performance measures for cardiology clinicians with at least 50 annual outpatient encounters from 2017-2023 in VA. A clinician-patient relationship was defined as ≥2 encounters within 1 year or ≥3 encounters within 3 years. Performance measure achievement was calculated separately for each clinician as well as an overall average. If more than one encounter occurred per year for a clinician-patient dyad, one was chosen at random for evaluation. Clinician performance distributions were represented as histograms, and differences in unadjusted performance by clinician training status and gender were assessed using Welch two sample t-tests.

Results
Of 456,071 unique cardiology encounters, 179,476 included AF with elevated stroke risk, 313,603 included CAD, and 225,476 included HF with EF ≤40%. There were 1,046 cardiology providers including 53% physicians and 52% women. Overall mean performance measure achievement was 87.7%. Mean performance scores for AF, CAD, and HF were 86.3%, 94.2%, and 78.9% respectively, and distributions are shown in the Figure. There was a modest difference in unadjusted overall performance between physicians and non-physician providers (88.0 vs. 87.2%, p = 0.005) and no difference between female and male clinicians (87.5% vs. 87.8%, p = 0.4)

Conclusion
There was high achievement of three established VA outpatient cardiology performance measures with minimal differences by clinician training or gender. This may represent a ceiling effect. However, significant variation remained at the clinician level, indicating potential opportunities for quality improvement even for these well-established treatments.