Introduction Transthoracic echocardiography (TTE) is an established screening tool used in the assessment of suspected Pulmonary Hypertension (PH). There is evidence to suggest Doppler estimates of Pulmonary Artery Systolic Pressure (PASP) are inaccurate compared to Right Heart Catheter (RHC) measures. Therefore, the ESC and BSE recommend a multi parameter assessment of TTE probability of PH. This retrospective audit sought to evaluate the effectiveness of the BSE/ESC TTE probability algorithm in detecting PH in a real world cohort referred to a shared care PH centre.
Method Between 2010 and 2019, a total of 310 patients referred for initial assessment of PH underwent TTE followed by RHC (median interval 31±30 days). PH TTE probabilities were calculated following BSE/ESC guidance. Bland-Altman analysis was used to investigate the accuracy of Doppler estimates of PASP compared to RHC.
Results The mean sample age was 67±14years (62% female). Seventy-six percent (n=235) had PH (mPAP ≥25mmHg) (average mPAP 42.8±11.7mmHg), 26% (n=80) having CTEPH. Doppler estimates of PASP tended to underestimate true PASP (bias -3.7±15.2mmHg) with wide limits of agreement (95% limits of agreement -33.5-26.1mmHg) (figure 1); highlighting the imprecision of Doppler estimates alone. Only 44% of Doppler TTE PASP estimates were within 10mmHg of RHC PASP readings. Underestimation occurred more frequently accounting for 66% of inaccurate TTE PASP estimates.
Ninety-six percent of those with a high TTE probability (average RHC mPAP 45.3±12.1mmHg) and 78% with an intermediate probability of PH (average RHC mPAP 32.8±11.2mmHg) had PH confirmed at RHC (figure 2). The sensitivity and specificity for detecting PH for the whole cohort was 89% and 68%, respectively. However, 35% (n=27) of low TTE probability had PH (average RHC mPAP 31.9±6.6mmHg) with PH secondary to LHD (n=17), CTEPH (n=5), PAH (n=4) and multifactorial/unclear PH (n=1) observed in this group. The 4 PAH patients all had PH secondary to connective tissue disease (systemic sclerosis). NT-proBNP values were significantly higher in those with LHD compared to those with pre-capillary PH (462.6±428.1ng/L vs 148.4±107.7ng/L, p<0.01).
There were 38 documented all-cause mortalities (median FU 2 years, IQR 2), 37 in those with PH. Eighty-four percent (n=31) of these had a high TTE probability of PH. Those that died were significantly older, had lower cardiac index, PA sats, 6MWT distance, higher NT-proBNP, mPAP, PVR, TRV and more frequently had TTE evidence of raised RAP and pericardial effusion (p<0.05) (table 1).
Conclusion The PH TTE probability algorithm provides high sensitivity and moderate specificity for screening individuals at risk of PH. However, 35% of patients with low TTE probability had PH confirmed by RHC. Alternative strategies are needed to improve the sensitivity to detect PH in those with a low TTE probability of PH.
Conflict of Interest None