Validation of thoracic impedance cardiography by echocardiography in healthy late pregnancy

Jordan PR McIntyre, Kevin M Ellyett, Edwin A Mitchell, Gina M Quill, John MD Thompson, Alistair W Stewart, Robert N Doughty, Peter R Stone, the Maternal Sleep in Pregnancy Study Group

Assessment of stroke volume (SV) is often necessary in clinical and research settings. The clinically established method for SV assessment in pregnancy is echocardiography, but given its limitations, it is not always an appropriate measurement tool. Thoracic impedance cardiography (ICG) allows continuous, non-invasive SV assessment. However, SV determination relies on assumptions regarding the thoracic shape that may mean the algorithm is not valid in pregnancy. The available data regarding the validity of ICG against an established reference standard using modern SV algorithms are both limited and conflicting. We aimed to test the validity of ICG in a clinically realistic setting in late pregnancy using echocardiography. Twenty-nine women in late pregnancy underwent standard echocardiography assessments with simultaneous ICG measurement. Agreement between devices was tested using Bland-Altman analysis. Bland-Altman analysis of the relationship between ICG and echocardiography demonstrated that the 95% limits of agreement exceeded acceptable or expected ranges. Measures of maternal and fetal anthropometry do not account for the lack of agreement. Absolute values of SV as determined by ICG are not valid in pregnancy. Further work is required to examine the ability of ICG to assess relative changes in maternal haemodynamics in late pregnancy.