Focused cardiac ultrasound to guide the diagnosis of heart failure in pregnant women in India.
Alsharqi M., Ismavel VA., Arnold L., Choudhury SS., Solomi V C., Rao S., Nath T., Rani A., Goel I., Kakoty SD., Mahanta P., Roy I., Deka R., Opondo C., Baigent C., Leeson P., Nair M., MaatHRI collaboration None.
BACKGROUND: Cardiac complications are a leading cause of maternal death. Cardiac imaging with echocardiography is important for prompt diagnosis, but it is not available in many low-resource settings. The aim of our study was to determine whether focused cardiac ultrasound performed by trained obstetricians and interpreted remotely by experts can identify cardiac abnormalities in pregnant women in low-resource settings. METHODS: Cross-sectional study of 301 pregnant and postpartum women recruited from ten hospitals across three States in India. Twenty-two obstetricians were trained in image acquisition using a portable cardiac ultrasound device following a simplified protocol adapted from the Focus-Assessed Transthoracic Echocardiography. It included parasternal long axis (PLAX), parasternal short axis (PSAX), and apical four-chamber views in 2D and colour Doppler. Independent image interpretation was performed remotely by two experts, in the UK and India, using a standard semi-quantitative assessment protocol. Interrater agreement between the experts was examined using Cohen's Kappa (κ). Diagnostic accuracy of the method was examined in a sub-sample for whom both focused and conventional scans were available. RESULTS: Cardiac abnormalities identified using the focused method included valvular abnormalities(27%), rheumatic heart disease(6.6%), derangements in left ventricular size(4.7%) and function(22%), atrial dilatation(19.5%) and pericardial effusion(30%). There was substantial agreement on the cardiac parameters between the two experts ranging from 93.6% (k=0.84) for left ventricular ejection fraction to 100% (k=1) for valvular disease. Image quality was graded as good in 79% of PLAX, 77% PSAX and 64% apical four-chamber views. The chance corrected k coefficients indicated fair to moderate agreement (k 0.28 to 0.51) for the image quality parameters. There was good agreement on diagnosis between the focused method and standard echocardiography (78% agreement) compared in 36 participants. CONCLUSION: The focused method accurately identified cardiac abnormalities in pregnant women and could be used for screening cardiac problems in obstetric settings.