Heart failure with preserved ejection fraction (HFpEF) affects over 3 million adults in the US with increasing prevalence each year.1 HFpEF is a heterogeneous clinical syndrome with hallmark symptoms of breathlessness and poor cardiac pump reserve despite having a left ventricular ejection fraction (EF) of ≥ 50%.2 Further, it is associated with ageing, hypertension, and obesity, and other comorbidities.3 The ability to accurately identify HFpEF from the electronic health record (EHR) may improve the understanding of the epidemiology, outcomes, and treatment response of representative patient populations with HFpEF, and can facilitate the planning and conduct of clinical research.

The reference standard for diagnosing HFpEF is invasive hemodynamic testing,4, 5, 6 which may not be performed during an initial diagnostic evaluation due to risk and availability. Several clinical diagnostic scores, including H2FpEF and HFA-PEFF scores, have been applied using EHR-derived parameters in the diagnosis of HFpEF among at-risk populations.7, 8, 9, 10, 11 These scores have yielded varying patient characteristics with varying short-term outcomes among hospitalized and ambulatory populations.4,9,10 The addition of concomitant loop diuretic use and natriuretic peptide level to these screening measures have also offered uncertain diagnostic potential when applied to a large population undergoing both echocardiography and cardiac catheterization.12

In the present study, we sought to characterize HFpEF cohorts identified by different EHR-based screening strategies and associated long-term outcomes across a large, diverse, and unrestricted population undergoing echocardiography, as well as those who underwent echocardiography within six months of invasive hemodynamics. We also aimed to understand how the addition of loop diuretics and natriuretic peptide levels may influence the agreement of HFpEF screening algorithms to HFpEF diagnosed by invasive hemodynamics.