ESTIMATION OF PULMONARY ARTERIAL-PRESSURE IN THE NEWBORN - STUDY OF THE REPEATABILITY OF 4 DOPPLER-ECHOCARDIOGRAPHIC TECHNIQUES

Despite the increasing use of Doppler echocardiographic (DE) technique s to determine pulmonary arterial pressure in the neonate undergoing i ntensive care, there have been no studies comparing their repeatabilit y in this population. Our objective was to compare the repeatability o f four such techniques in neonates. The study was conducted in two reg ional neonatal units serving the North East of England. Group A (repea tability between observers): Two experienced observers performed detai led DE examinations, one directly after the other. Group B (within obs erver repeatability/temporal variability): One observer performed two examinations 1 hour apart. Group A comprised 15 preterm babies (26-36 weeks' gestation, 975-2915 g), most with mild respiratory failure: 4 h ealthy term babies; and 7 with congenital heart disease, in whom tricu spid regurgitation (TR) only was measured. Their ages were 18 hours to 12 days. Group B comprised 11 babies aged 12-64 hours with moderate t o severe respiratory failure; 10 were preterm (26-36 weeks, 785-2800 g ). We recorded four measurements: (1) Peak velocity of TR in m/s; (2) peak left-to-right ductal flow velocity (PDAmax in m/s); (3) TPV/RVET ratio; and (4) PEP/RVET ratio, where TPV = time to peak velocity at th e pulmonary valve, PEP = right ventricular preejection period, and RVE T = right ventricular ejection time. The Bland-Altman analysis was use d to produce the coefficient of repeatability (CR: 95% confidence limi ts of repeatability), also expressed as a repeatability index (CR/mean value) and as a number of ''confidence steps''-a measure of sensitivi ty of the technique to hemodynamic change (range of values within the population/CR). Between-observer and within-observer repeatabilities w ere similar, Within-observer CR and index (%) results were for TR +/- 0.26 m/s (9%); for PDAmax, +/- 0.48 m/s (39%); TPV/RVET 0.1:1.0 (34%), PEP/RVET 0.12:1.00 (36%), TR and PDAmax had the largest number of con fidence steps in the expected range of values (TR 8.5; PDA max 6.5; TP V/RVET 3.2; PEP/RVET 3.2). The most repeatable technique was TR, but P DAmax would also be useful for a serial study owing to the potential f or large change. Systolic time interval ratios were less repeatable an d Likely to be less sensitive indicators of hemodynamic change.