Search

Background A mean pulmonary artery pressure >20 mm Hg now defines pulmonary hypertension. We hypothesize that echocardiographic thresholds must be adjusted. Research Question Should tricuspid regurgitation velocity thresholds to screen for pulmonary hypertension be revised, given the new hemodynamic definition? Study Design and Methods This multicenter retrospective study included 1,608 patients who underwent both echocardiography and right heart catherization within 4 weeks. The discovery cohort consisted of 1,081 individuals; the validation cohort included 527. Screening criteria for pulmonary hypertension were derived with the use of receiver operating characteristic analysis and the Youden index, assuming equal cost for false-positive and -negative classification. A lower threshold was calculated with the use of a predefined sensitivity: 95%. Results In the discovery cohort, echocardiographic tricuspid regurgitation velocity had a good discrimination for pulmonary hypertension: area under the curve, 88.4 (95% CI, 85.3-91.5). A 3.4-m/s threshold provided a 78% sensitivity, 87% specificity, and 6.13 positive likelihood ratio to detect pulmonary hypertension; 2.7 m/s had a 95% sensitivity and 0.12 negative likelihood ratio to exclude pulmonary hypertension. In the validation cohort, the discovery threshold of 2.7 m/s provided sensitivity and negative likelihood ratios of 80% and 0.31, respectively. Right cardiac size improved detection of pulmonary hypertension in the lower tricuspid regurgitation velocity groups. Interpretation Our data support a lower tricuspid regurgitation velocity of approximately 2.7 m/s for screening pulmonary hypertension, with a high sensitivity in tertiary referral centers. Right heart chamber measurements improve the diagnostic yield of echocardiography.

Introduction: Infective endocarditis (IE) is associated with high morbidity and mortality. Conventionally, a repeat transesophageal echocardiogram (TEE) is necessary if the first TEE is negative, and there is high clinical suspicion. We aimed to investigate the diagnostic performance of contemporary TEE for IE. Hypothesis: We hypothesize that with advancements in TEE imaging, including three-dimensional (3D) imaging, the diagnostic performance of TEE for IE would be improved. Methods: Patients who had two or more TEEs at our center, within 6 months, for evaluation of IE in 2011 (pre-3D imaging) and 2019 (post-3D imaging) were included. Patients not meeting the Duke criteria for IE (n=899) were excluded. The primary endpoint was the sensitivity of TEE to detect IE. TEE sensitivity was also compared among the different subtypes of IE (native valve, prosthetic valve, device-related, central line-related, and aortic prosthetic graft-related). Results: 242 patients were included: 70 in 2011 and 172 in 2019. In 2011, there were more smokers and a lower rate of pacing devices; otherwise, there were no significant differences in baseline characteristics, including intravenous drug use. The sensitivity of the initial TEE for IE was 85.7% in 2011, improving significantly to 95.3% in 2019 (p=0.01). The improved diagnostic performance in 2019 was mainly driven by the improved detection of prosthetic valve IE (70.8% vs. 93.7%, p=0.009). There were no significant differences for the other subtypes of IE (Figure). In 2019, patients had a higher rate of IE diagnosis within 15 days of index admission, which did not reach statistical significance (n=161, 93.6% vs. n=62, 88.6% in 2011, p=0.19). Kaplan Meier analysis demonstrated that patients in 2019 had a higher rate of procedures for endocarditis (p=0.01). Conclusions: Contemporary TEE imaging including 3D technology improved the detection of IE on the initial TEE, mainly driven by improved detection of prosthetic valve IE.