Aim

Heart failure with preserved ejection fraction (HFpEF) is a frequent disease. Nevertheless, diagnosing HFpEF remains challenging. Recently, different algorithms were developed to predict the likelihood of HFpEF. Our objective was to provide an in-depth comparison of the ESC 2016 algorithm, the H₂FPEF- and the HFA-PEFF score for diagnosing and characterising HFpEF in the general population.

Methods and Results

The study included 5,613 participants of the population-based Hamburg City Health Study (HCHS), aged 61.9 ± 8.4 years (51.1 % women), that were enrolled between 2016 and 2019. Exclusion criteria were other common causes of dyspnea. Unexplained dyspnea or self-reported heart failure was present in 407 subjects. In those, the H₂FPEF-score showed the highest applicability (97.1%), followed by the ESC 2016 algorithm (92.1%) and the HFA-PEFF score (88%). The estimated prevalence of HFpEF was 21.1% (ESC 2016), 12.3% (H₂FPEF), and 7.6% (HFA-PEFF) (Figure). In 57.5% and 59.2% of individuals, HFpEF was “not excludable” according to the HFA-PEFF and the H₂FPEF score. In subjects without dyspnea, HFpEF was predicted in 3.7% (HFA-PEFF), 0.5% (H₂FPEF) and 0.1% (ESC 2016). For all algorithms, subjects diagnosed with HFpEF showed a high prevalence of atrial fibrillation as well as the classical cardiovascular risk factors elevated age, body mass index, arterial hypertension and diabetes. The concordance between the three algorithms for diagnosing or excluding HFpEF ranged from poor to fair.

Conclusion

The ESC 2016 algorithm, the H₂FPEF- and the HFA-PEFF score differentially diagnosed HFpEF with the highest prevalence applying the ESC 2016 algorithm and the highest feasibility applying the H2FPEF score. Most individuals were classified as HFpEF not excludable according to the HFA-PEFF and the H₂FPEF score challenging their clinical applicability. Further evidence is needed for consistently defining and diagnosing HFpEF. 

Figure. Prevalence and concordance of the three HFpEF algorithms in subjects with unexplained dyspnea. 
Of the 407 subjects with unexplained dyspnea or self-reported heart failure, the prevalence ranged from 21.1% (n = 86, ESC 2016 guideline) to 12.3% (n = 50, H₂FPEF score) and 7.6% (n = 31, HFA-PEFF score). The concordance was highest between the ESC 2016 guidelines and the HFA-PEFF score reflected by a kappa coefficient of 0.39 and a reclassification rate of 16.4%. RecR = reclassification rate.