Non-invasive determination of pulmonary arterial pressure in the newborn.

In the newborn preterm baby with respiratory failure, pulmonary arterial pressure can fluctuate dramatically, yet invasive monitoring is not usually practical in this population. This thesis studies the use of Doppler echocardiography in the non-invasive assessment of pulmonary arterial (PA) pressure in the newborn, concentrating on measuring peak velocity of tricuspid regurgitation and applying the modified Bernoulli equation (the "TR method"), but also assessing the alternatives; analysis of ductal flow and two right ventricular systolic time interval ratios. The TR method was found to be reasonably accurate when compared to direct cardiac catheter measurement (95% confidence limits -12 mmHg to +8 mmHg) and repeatable (repeatability index 10%) but feasibility of measurement was limited, particularly in babies with bronchopulmonary dysplasia. Ductal flow and right ventricular systolic time intervals (TPV/RVET ratio (time to peak velocity at the pulmonary valve/right ventricular ejection time) and PEP/RVET ratio (right ventricular pre-ejection period/RVET) were assessed as alternative techniques by comparing values with those obtained simultaneously using the TR method. 741 Doppler echocardiograms from 203 babies in the first month of life were analysed. The pattern of ductal flow during the cardiac cycle was shown to reflect both the pulmonary: systemic arterial pressure ratio, and the quantity of left-to-right shunting. Measurement of the velocity of flow across the duct was useful in studying serial haemodynamic change within a baby, particularly in persistent transitional circulation, but application of the modified Bernoulli equation to the velocity, to determine pressure drop across the duct, was unreliable. Systolic time intervals and ratios were subject to considerable error in measurement, and the relationship to pulmonary arterial pressure was not consistent between babies of different gestation and was influenced by other factors such as ductal patency. Using a combination of the TR method and analysis of ductal flow, longitudinal studies of systolic PA pressure in healthy newborns and babies ventilated for HMD were performed in the first days of life. The rate of fall in the pulmonary : systemic (PA:Ao) arterial pressure ratio was similar in healthy term and preterm babies. In the babies with HMD, PA pressure fell more slowly, Ao pressure rose more slowly and ductal closure was delayed. PA pressure did not correlate with disease severity. Babies of lower gestation had lower PA pressure values, and, in general, a more rapid fall in the PA:Ao arterial pressure ratio than more mature babies. The haemodynamic effects of altering inspired oxygen levels was investigated using all four methods of PA pressure estimation. This was done by expressing the haemodynamic changes in terms of multiples of confidence limits of repeatability of each technique (termed "confidence steps") a technique which shows considerable promise for further study. Amongst babies with persistent fetal circulation, ductal flow velocities were particularly sensitive in detecting haemodynamic change, but none of the initial measurements of PA pressure correlated with eventual outcome. Doppler echocardiography now permits reasonably accurate and clinically useful non-invasive determination of pulmonary arterial pressure in most newborn babies with cardiorespiratory distress. Four techniques were evaluated in this thesis, and while each was found to have advantages and disadvantages, the TR method, limited by feasibility in some patients, was found to be the most reliable.