Validation of Echocardiographic Measurements in Patients with Pulmonary Embolism in the RIETE Registry.
Mads Dam LyhneBehnood BikdeliDavid M DudzinskiAlfonso Muriel-GarcíaChristopher KabrhelTeresa Sancho-BuesoEsther Pérez-DavidJosé Luis LoboÁngel Alonso-GómezDavid JiménezManuel Monrealnull nullPublished in: TH open : companion journal to thrombosis and haemostasis (2024)
Background In acute pulmonary embolism (PE), echocardiographic identification of right ventricular (RV) dysfunction will inform prognostication and clinical decision-making. Registro Informatizado Enfermedad TromboEmbolica (RIETE) is the world's largest registry of patients with objectively confirmed PE. The reliability of site-reported RV echocardiographic measurements is unknown. We aimed to validate site-reported key RV echocardiographic measurements in the RIETE registry. Methods Fifty-one randomly chosen patients in RIETE who had transthoracic echocardiogram (TTE) performed for acute PE were included. TTEs were de-identified and analyzed by a core laboratory of two independent observers blinded to site-reported data. To investigate reliability, intraclass correlation coefficients (ICCs) and Bland-Altman plots between the two observers, and between an average of the two observers and the RIETE site-reported data were obtained. Results Core laboratory interobserver variations were very limited with correlation coefficients >0.8 for all TTE parameters. Agreement was substantial between core laboratory observers and site-reported data for key parameters including tricuspid annular plane systolic excursion (ICC 0.728; 95% confidence interval [CI], 0.594-0.862) and pulmonary arterial systolic pressure (ICC 0.726; 95% CI, 0.601-0.852). Agreement on right-to-left ventricular diameter ratio (ICC 0.739; 95% CI, 0.443-1.000) was validated, although missing data limited the precision of the estimates. Bland-Altman plots showed differences close to zero. Conclusion We showed substantial reliability of key RV site-reported measurements in the RIETE registry. Ascertaining the validity of such data adds confidence and reliability for subsequent investigations.
Keyphrases
- pulmonary embolism
- left ventricular
- mycobacterium tuberculosis
- ejection fraction
- electronic health record
- mitral valve
- pulmonary hypertension
- big data
- left atrial
- inferior vena cava
- heart failure
- aortic stenosis
- blood pressure
- hypertrophic cardiomyopathy
- liver failure
- decision making
- acute myocardial infarction
- randomized controlled trial
- cardiac resynchronization therapy
- data analysis
- coronary artery disease
- physical activity
- prognostic factors
- respiratory failure
- newly diagnosed
- machine learning
- study protocol
- aortic valve
- percutaneous coronary intervention
- artificial intelligence
- patient reported outcomes
- aortic dissection
- deep learning