Subjects

Infants were eligible for inclusion in the study if they were intubated and ventilated from birth, if they had indwelling arterial lines for direct measurement of blood pressure, and if they had not been exposed to opiates in pregnancy or during delivery. Seventeen infants (eight boys) were studied. Their median gestational age was 27.0 weeks (range 24–32) and median birth weight 1.03 kg (range 0.56–1.80). At the time the morphine was given, the median age was 3.1 hours (range 2.3–5.3). Nine had respiratory distress syndrome, of whom eight received their first dose of surfactant before the study. The remaining infants were ventilated for immaturity alone. The median Fio ₂ was 0.25 (range 0.21–0.97); seven infants were breathing air at the time of study. The median mean airway pressure was 10 mm Hg (range 6–13). Infants were studied with the written consent of a parent and the study was approved by the Central Sydney Area Health Service ethics review committee.

Statistical analysis

The physiological data were normally distributed. The value of each variable at 10 and 60 minutes was compared with the baseline value using a paired t test.

Results

Table 1 gives the mean values for the physiological variables, and table 2 gives the mean percentage change at 10 and 60 minutes from baseline values.

There was a small but statistically significant fall in heart rate at 10 and 60 minutes. Changes in systolic, diastolic, and mean blood pressure were examined in three ways. When point measurements at 10 and 60 minutes were compared with baseline values in all 17 infants, there were no significant changes. In the 10 infants for whom serial measurements were taken, blood pressure was averaged over 10 minute periods leading up to each of the three measurement points; again no significant changes were shown. In each of these 10 infants, the time plot of systolic, diastolic, and mean blood pressure was examined visually. No change could be seen in response to the bolus injection of morphine.

RVO fell by 6% at 10 minutes and was 3% below baseline values at 60 minutes, but these falls were not statistically significant. No significant changes were seen in SVCF. Overall values of RVO and SVCF remained within our normal range for preterm infants of this age.

There was a highly significant reduction in duct diameter in the hour after the morphine was given (mean 16% fall from baseline value; p < 0.005). In all cases the shunt direction was either solely or predominantly from left to right and unaffected by morphine.

Discussion

Intravenous morphine given as a bolus to healthy adults has a vasodilator effect and acts particularly on the venous system.11 ,12 It has little or no effect on mean arterial blood pressure but improves systemic blood flow by reducing vascular resistance. It does, however, impair the normal compensatory mechanisms that maintain arterial blood pressure in the upright position and in the presence of hypovolaemia.11 The venodilator action is thought to be through the central nervous system,12 ,13although there may be a peripheral vascular action caused directly by the opiate or indirectly by histamine release. In adults with cardiovascular disease, there may be a transient fall in blood pressure, but cardiac output is either improved or unchanged, not reduced.14 ,15

Previous studies of intravenous morphine in the preterm ventilated newborn have shown a small fall in mean arterial blood pressure (from 2 to 4 mm Hg) when the drug is given as an infusion over two hours in a dose of 100 or 200 μg/kg.1-3 Sabatinoet al 2 studied the haemodynamic effects of a 100 μg/kg dose of morphine infused over two hours in 30 preterm infants. They measured cerebral blood flow velocity and left ventricular output before and during the infusion, and found no significant changes. Left ventricular output, however, is an unreliable measure of cardiac output as most ventilated preterm infants on the first day of life will have a ductal shunt that is mainly or wholly left to right.8

This is the first study of the effects of a bolus injection of opiate in the newborn. We have shown a small but statistically significant reduction in heart rate following an intravenous bolus of morphine, consistent with its sedative effect. We have failed to show any other cardiovascular action. The effect on systemic arterial blood pressure (systolic, diastolic, and mean) was examined in a number of ways, but no significant effect of the morphine bolus was shown. There was no evidence of a fall in systemic blood flow. RVO showed a small non-significant fall but values were well within the normal range for infants of this gestation and age. SVCF (to which cerebral blood flow is the major contributor16) was unchanged.

This study was uncontrolled—that is, there was no parallel untreated group for comparison of the cardiovascular variables. It is possible therefore that a hypotensive or cardiodepressant action of morphine was masked by the effect of time because blood pressure and cardiac output in preterm infants tend to increase in the early newborn period. We think, however, that this is unlikely. The morphine was given as a rapid bolus over a few seconds and the effects were sought 10 and 60 minutes later. It is within this period that any cardiovascular effect of morphine would be expected to occur. Vasoactive drugs given as an intravenous bolus should exert their effects very quickly. In several respects, this study represents the worst case situation for demonstrating cardiovascular side effects of an intravenous opiate. The drug was given rapidly to a very immature group of infants who were studied very soon after birth at a time of cardiovascular adaptation, yet no effects could be seen. The number of infants studied was not large, so a small cardiovascular effect or an uncommon idiosyncratic reaction to the morphine injection could not be ruled out.

It is theoretically possible that intravenous morphine impairs closure of the ductus arteriosus by a direct action on the muscle of the ductal wall. This has not been examined before but we found no evidence of it. There was a highly significant reduction in ductal width of 16% over the one hour study period with no alteration in the nature of the ductal shunt, which remained wholly or mainly from aorta to pulmonary artery. This is consistent with the natural history of spontaneous duct closure in this group of infants.17

The study does not support the recommendation that a 100 μg/kg loading dose of morphine has to be infused slowly over two hours. There may be no particular advantage in giving it either slowly or quickly in an infant who is already being ventilated. However, if sedation and analgesia is needed instantly, for example during semielective intubation in an infant with worsening respiratory distress or recurrent apnoea, then a bolus injection seems to be an appropriate and safe way of providing it.