External validation of a simple non-invasive algorithm to rule out chronic thromboembolic pulmonary hypertension after acute pulmonary embolism.
Thromb Res. 2014 Dec 13;
Authors: Klok FA, Tesche C, Rappold L, Dellas C, Hasenfuß G, Huisman MV, Konstantinides S, Lankeit M
PURPOSE: International guidelines do not provide strong recommendations on the duration and intensity of follow-up after acute pulmonary embolism (PE), nor on screening-programs for chronic thromboembolic pulmonary hypertension (CTEPH). We aimed to address this gab by performing an external validation of the easy "CTEPH rule-out-criteria" based on a normal NT-proBNP level and the absence of 3 ECG characteristics.
METHODS: 134 patients underwent clinical follow-up 6months after PE. Predefined transthoracic echocardiographic (TTE) criteria were used to categorize patients as "PH unlikely" or "PH possible/likely". The latter patients underwent further (invasive) diagnostic procedures to confirm and classify the diagnosis of pulmonary hypertension. NT-proBNP and ECGs, both assessed at the day of echocardiography, were evaluated post-hoc.
RESULTS: Sixty-three patients (47%) scored none of the "CTEPH rule-out criteria" positive, of whom 61 had normal TTE (97%). Twenty-five patients (19%) were categorized by TTE as "PH possible/likely"; of those, 6 were diagnosed with CTEPH. The sensitivity of rule-out criteria for CTEPH was 100% (95%CI 56-100%; 6/6 patients identified), and for "PH possible/likely" on TTE 92% (95%CI 74-99%; 23/25 patients identified): 2 asymptomatic patients with estimated systolic pulmonary arterial pressure of 36mmHg and 38mmHg, respectively, who remained stable during further 2-year follow-up, were not identified. Inter-observer agreement for the adjudication of the ECG characteristics was excellent (kappa-statistic 0.97).
CONCLUSIONS: In this external validation cohort, we confirmed the diagnostic accuracy and reproducibility of the "CTEPH rule-out criteria". These results provide a solid ground for future outcome trials applying this algorithm.
PMID: 25746363 [PubMed - as supplied by publisher]