25 Visual grading of systemic right ventricles: are we any good?. (2nd February 2018)

• Record Type: Journal Article
• Authors: Perez, M; Green, J; Edwards, N; Hudsmith, L
• Is Part Of: Heart. Volume 104(2018)Supplement 2
• Page Start: A9
• Page End: A9
• Publication Date: 2018-02-02
• Subjects: Heart -- Diseases -- Treatment -- Periodicals; Cardiology -- Periodicals; 616.12
• DOI: 10.1136/heartjnl-2017-BCCA.25
• Languages: English
• ISSNs: 1355-6037

Abstract

Abstract : Purpose: Accurate and reproducible quantification of ventricular function is essential for serial monitoring and management of complex congenital patients. Systemic right ventricular functional analysis can be challenging. We sought to investigate visual grading and quantification in systemic right ventricles with transthoracic echo and CMR and impact of training and experience on reproducibility. Methods: 10 anonymous apical '4-chamber' transthoracic (Epic, Phillips) images, duplicated twice, were viewed by 20 participants of varying experience and accreditation and 10 anonymous '4-chamber' CMR (1.5 T Avanto, Siemens) HLA images of the same patients, duplicated twice, were analysed by 12 participants. Participants were asked to provide an overall visual grading of systemic RV function, an 'eyeball' estimate of systemic RV function and a judgement of image quality (%). Experience and accreditation were noted. Results: 10 patients, 6 female, mean age 28 years, range 20–48 years: 8 patients with single ventricle (Fontan), 1 atrial switch (Senning) and 1 ccTGA. 16 echo viewers were accredited, 5 regularly performing congenital echo and 6 CMR viewers were accredited. CMR image quality ranged between 50–100% and echo image quality between 15–90%. Repeat blinded visual estimates by the same observer were in agreement between 45–100% in echo and 25–83% with CMR with less reproducibility in those observers without CMR accreditation. Conclusion: Visual grading and quantification of complex congenital heart disease is difficult with limited reproducibility. There is a variation in visual grading by observers, particularly in CMR. Formal quantification by experienced congenital imagers is recommended to optimise reliability of serial measurements with both echo and CMR