Bridging sex differences in HFpEF diagnosis: enhanced predictive performance of the H2FPEF score by incorporating BNP and stepwise E/e′

Background Heart failure with preserved ejection fraction (HFpEF) is more prevalent in women due to biological and hormonal factors. The diagnosis of HFpEF remains challenging, as current diagnostic criteria are often derived from mixed cohorts, resulting in variability in interpretation. The European Society of Cardiology (ESC) Consensus Criteria is regarded as the gold standard but is complex and not easily applicable in routine clinical practice. The H2FPEF-Score is easier to use but has limited predictive capacity. A gender-specific scoring system incorporating laboratory parameters, comorbidities and echocardiographic values is needed. Methods 1,881 patients who underwent standard transthoracic echocardiographic assessment were included in this study. Data from the CentraXX database was extracted from the department of Cardiology, Pneumology and Angiology at a university hospital. Chemotherapy patients were excluded. The final analysis included 1,376 patients, of whom 783 were male and 593 female. Patients were stratified according to the different HFpEF scoring systems. The diagnostic classifications were then compared across male and female patients to analyse positive predictive value according to gender. ROC analysis evaluated the predictive capacity of the modified composite score in relation to ESC criteria. The H2FPEF-Score was extended by incorporating BNP levels and a stepwise E/e′ assessment (H2FPEF++-Score). A six-month follow-up was conducted and multidimensional radar plot analysis used to determine the parameter with the highest discriminatory capacity for HFpEF progression. Results The diagnosis of HFpEF varied between different scorings and sexes. Female patients showed higher rates of H2FPEF-Score >6 (p=0.014). Standard diastolic dysfunction determination was equivalent. H2FPEF-Score was better at diagnosing female patients with HFpEF according to ESC criteria, indicating that comorbidities are more significant in females (p = 0.002). The H2FPEF++-Score was able to detect male and female patients with true HFpEF more precisely according to ESC criteria (false negative rate 11,11% in male and 0% in female patients). Predictive capacity measured by ROC showed the highest predictive capacity for the H2FPEF++ in women (AUC: 0.899), and even more pronounced in women with atrial fibrillation (AUC: 0.997). In longitudinal analysis of untreated HFpEF patients, radar plot analysis identified E/e′ as the parameter demonstrating the greatest change over time both in female and male patients. Conclusion The study showed the H2FPEF Score++ is highly accurate. It uses BNP and E/e', which are more practical and easier to calculate than the ESC criteria for HFpEF. E/e' also shows the most changes over time in non-treated HFpEF and can be used to assess progression. Our results show the importance of incorporating clinical and functional parameters and NTproBNP into the diagnosis of HFpEF.