Simple echocardiographic algorithm incorporating left atrial strain for estimation of left atrial pressure: comparison with the 2025 ASE/EACVI guideline
H Ahn, M Pak, J Park, H Choi, I Hwang, Y Yoon, G ChoAbstract
Background
Accurate noninvasive estimation of left atrial pressure (LAP) is essential for clinical decision-making. While the 2025 ASE/EACVI guideline updated its algorithm, it remains complex for routine practice. Left atrial strain (LAS), a sensitive marker of LA function, is a promising parameter for LAP estimation but is currently limited to a supplementary role in existing guidelines.
Objective
This study aimed to evaluate a simple echocardiographic algorithm incorporating LA reservoir strain (LASr) as a primary criterion for LAP estimation. Its diagnostic accuracy was compared with the 2025 ASE/EACVI algorithm, using invasive direct LAP measurement as the reference standard.
Methods
We retrospectively analyzed 114 patients in sinus rhythm (SR) who underwent simultaneous echocardiography and invasive LAP measurement (LAP ≥ 15 mmHg defined as increased). The 2025 ASE/EACVI algorithm was applied using its stepwise approach. The novel algorithm utilized two primary criteria: (1) E/e’ > 15 and (2) LASr < 18%. If criteria were discordant, supplementary criteria (TR velocity > 2.8 m/s or LAVi-4C > 34 mL/m²) determined the final classification.
Results
Among 114 analyzable NSR patients (mean LAP 12.3 ± 4.9 mmHg), 31 (27.2%) had invasive LAP ≥ 15 mmHg. Mean E/e’ was 10.6 ± 3.2, mean LASr was 20.2 ± 6.7%, mean LAVI was 46.2 ± 12.0 mL/m², and mean EF was 62.1 ± 6.0%. The 2025 guideline classified 91 patients (79.8%) as having increased LAP, yielding a sensitivity of 87.1% but a specificity of only 22.9%, with an overall accuracy of 40.4%. In contrast, the novel algorithm classified 40 patients (35.1%) as having increased LAP, achieving an accuracy of 67.5%, sensitivity of 54.8%, and specificity of 72.3%. The PPV improved from 29.7% to 42.5%, while the NPV remained comparable (82.6% vs. 81.1%). The improvement in overall diagnostic accuracy was statistically significant by the McNemar test (p < 0.001), with 41 patients correctly reclassified by the new algorithm versus 10 patients misclassified.
Conclusion
The 2025 ASE/EACVI algorithm showed high sensitivity but low specificity due to prevalent elevated LAVi. A novel algorithm elevating LASr to a primary criterion alongside E/e` significantly improved diagnostic accuracy (67.5% vs. 40.4%, p < 0.001) and specificity (72.3% vs. 22.9%). These findings suggest that incorporating LA strain as a core parameter enhances noninvasive LAP estimation, especially in patients with LA structural remodeling.