Left ventricular function evaluation of the fetal heart: reference intervals and inter-observer variability of 2D speckle-tracking echocardiography measurements.

Abstract

Measurement of myocardial strain using 2D speckle-tracking echocardiography can successfully quantify ventricular function, being considered superior to conventional echocardiography. This study aimed to establish reference intervals, interobserver agreements and reliability of two fetal echocardiographic parameters which reflect left ventricular myocardial function, left ventricular apical 4 chamber endo peak strain (AP4pLS) and ejection fraction (EF). We conducted a prospective study on 103 healthy fetuses. In each case, cardiac ultrasound images obtained were stored and afterwards were subject to offline 2D speckle-tracking echocardiographic analyses. In 15 randomly chosen subjects a second examiner also carried out an offline analysis of the 4-chamber view and the archived images, in order to assess inter-observer reproducibility and agreement level. Our study group was sub-divided into four different gestational age groups. Reference ranges were established for the two measured parameters, AP4pLS and EF, which did not differ significantly between four different gestational age groups (p=0.98 and p=0.64) and neither correlated with gestational age progression (p=0.37 and p=0.08). An excellent level of agreement between the two examiners was found for the echocardiographic measurements, expressed through an intra-class correlation coefficient (ICC) value of 0.85 (0.62-0.94 for 95%CI) for AP4pLS and 0.78 (0.47- 0.92 for 95% CI) for EF. Speckle tracking AP4pLS and EF parameters are useful for assessment of ventricular myocardial function in healthy fetuses and can be reliably reproduced by two different skilled examiners. Further studies conducted on larger populations are required to standardize reference values of fetal speckle-tracking measurements.