Abstract
Guidelines recommend different approaches to investigate for pulmonary hypertension (PH) by transthoracic echocardiography (TTE). We used data from the CIPHER study (NCT04193046) to prospectively evaluate TTE detection of PH. Participants newly referred to PH clinics who underwent right heart catheterization (RHC) within 6 weeks and TTE within 60 days of enrolment (blinded central TTE reading) were classified by TTE probability of PH applying (i) the 2015 European Society of Cardiology (ESC)/European Respiratory Society (ERS) TTE algorithm or (ii) right ventricular systolic pressure (RVSP) > 40 mmHg. For calculation of sensitivity and specificity, 'non-assessable' patients (peak tricuspid regurgitation velocity [TRV] missing or ≤ 2.8 m/s with missing data on other echocardiographic signs) and patients with missing RVSP were counted as PH-negative. Performance was measured against RHC-confirmed diagnosis of mean pulmonary artery pressure > 20 mmHg. Of 475 patients included, 345 (73%) had PH. Using the ESC/ERS algorithm, PH probability was high, intermediate, low and non-assessable for 198, 104, 22 and 151 patients and PH prevalence was 98%, 75%, 23%, and 44%, respectively. Seventy-three patients were missing RVSP and 292 had RVSP > 40 mmHg. The two TTE approaches achieved similar results: sensitivity was 79%-77%, specificity was 78%-79%. This prospective study of patients newly referred to PH centres for RHC found similar sensitivity and specificity when using either RVSP > 40 mmHg or the 2015 ESC/ERS TTE algorithm. Among patients who were low-probability or non-assessable by ESC/ERS algorithm, 42% had PH, highlighting the persistent need for additional non-invasive investigative tools.