The definition, incidence and clinical importance of hypotension in the newborn period have recently been reviewed.1 ,2 Although there is clearly an association between hypotension and adverse outcomes in preterm infants3^–6 this association is not necessarily causative, and there is no evidence that treating hypotension improves outcome.2 Volume of flow in the superior vena cava (SVC) is a repeatable measure of true systemic perfusion7 ,8 and low SVC flow has been associated with periventricular haemorrhage9 and adverse long-term neurodevelopmental outcome.10 Volume of blood flow is considered to be a clinically more important variable than blood pressure in the newborn.11 However, the ability to measure flow is predominantly restricted to research studies, such that blood pressure remains the more frequently monitored indicator of neonatal circulatory status.1

Blood pressure is the product of flow and resistance. Thus systemic arterial pressure values potentially could be normal or high in the presence of low cardiac output if the vascular resistance is high.11 ,12 During the transition from the in-utero to the ex-utero circulation there is evidence that systemic vascular resistance is the principal determinant of blood pressure in preterm infants.13 ,14 In persistent pulmonary hypertension of the newborn high pulmonary arterial pressure is associated with low pulmonary blood flow, yet paradoxically, for many years a normal or high systemic blood pressure has been taken to indicate normal or high systemic flow.12

Studies in newborn infants have variously suggested weakly positive,15 weakly negative13 or no14 ,16 association between arterial blood pressure and left ventricular output (LVO). It has been suggested that the volume of flow in the SVC has a weakly positive association with blood pressure.9 There are no published data on the association between blood pressure and right ventricular output (RVO), which may be a more reliable marker of systemic perfusion than LVO in the presence of ductal shunting17 though RVO is confounded by atrial shunting. Similarly, there are no published data on the relationship between blood pressure and volume of descending aortic (DAo) blood flow, another potential marker of systemic perfusion.8

The aim of this study was to examine the relationship between invasively monitored arterial blood pressure and four measurements of systemic perfusion, LVO, RVO, SVC flow and DAo flow, in the first 48 h of postnatal life in preterm neonates.

METHODS

Infants enrolled in this study were part of a larger cohort of 80 infants of less than 31 weeks’ completed gestation studied at National Women’s Hospital, Auckland, New Zealand, between 1 December 2002 and 1 May 2004. Thirty-four of these infants had indwelling arterial lines. They underwent echocardiography as close as possible to 5, 12, 24 and 48 h postnatal age. All scans were performed with an ATL 3000 ultrasound scanner (Advanced Technological Laboratories, Bothell, Washington, USA) equipped with a 7 MHz phased-array probe. Images were recorded on videotape and measurements undertaken away from the cotside by one investigator (AMG).

We assessed SVC and DAo flow volume as described previously.8 LVO was assessed with internal aortic diameter measured from the parasternal long axis view immediately distal to the valve orifice at end systole using the M mode trailing edge-leading edge technique, and aortic flow velocity assessed from a modified apical view.17 RVO was assessed with pulmonary diameter measured from either the parasternal short axis or the tilted parasternal long axis view at the hinge-points of the pulmonary valve by frame-by-frame analysis of the two-dimensional image during cardiac systole, and pulmonary flow velocity assessed from the parasternal short axis view at the level of the tips of the pulmonary valve leaflets.18

The measurement of the scans is described elsewhere.8 All measurements were averaged over five consecutive cardiac cycles19 except for SVC flow velocity, where measurements were averaged over 10 consecutive cycles to minimise the impact of respiratory movements on flow.

Systolic, diastolic and mean blood pressure values taken from indwelling arterial lines were downloaded every 60 s using Marquette Solar 8000 monitors (GE Medical Systems, Wisconsin, USA) and Bedmaster V1.3 software (Excel Medical Electronics Inc, Florida, USA). The downloaded blood pressure value is an average of the blood pressure monitored over the previous 6 s. These measurements of blood pressure were then averaged over the precise duration of the echocardiogram.

The level of circulatory support provided to the infants was determined by the attending clinicians, who were unaware of the results of the echocardiograms. In general, clinicians followed unit guidelines of using fluid boluses and inotropes to maintain mean arterial blood pressure greater than gestational age in weeks.1

At each scan, we compared LVO, RVO, SVC flow and DAo flow with blood pressure by using univariate analysis. We examined the impact of potential confounding factors such as birth weight and gestation by adjusting for these variables using multiple linear regression analysis. All data were analysed using Statview software (SAS Institute, Cary, NC, USA). In all cases statistical significance was taken where p<0.05.

RESULTS

Thirty-four infants were studied, with a median (range) gestation of 27 weeks (24–29) and a median birth weight of 885 g (510–1430 g). Sixteen (47%) infants were boys. The mothers of 32 (94%) infants received antenatal corticosteroids. Thirty-one (91%) infants required mechanical ventilation, three (9%) required continuous positive airway pressure (CPAP) only. Seven (21%) infants developed periventricular haemorrhage and four (12%) died before hospital discharge.

Twenty (59%) infants received at least one bolus of intravenous fluid for circulatory support; 17 (50%) received a bolus prior to their first echocardiogram. Seven (21%) infants received inotropic support with dopamine, and no other inotropic agents were used. Two (6%) infants were receiving dopamine (5–12 μg/kg/min) at the time of the 5-h scan, six (18%) were receiving dopamine (6–15 μg/kg/min) at each of the 12-h and 24-h scans and three (9%) were receiving dopamine (6–9 μg/kg/min) at the 48-h scan. Continuously monitored invasive measures of blood pressure were available in 29–34 infants at the time of each echocardiogram.

5-h scan

At 5 h postnatal age there was a wide scatter of SVC flow values for any given blood pressure. However, there was a weak but significant inverse correlation between volume of SVC flow and mean arterial blood pressure (R2 = 0.141, p = 0.04) (fig 1). For each 1 mm Hg increase in mean blood pressure, SVC flow tended to decrease by an average of 2 ml/kg/min.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Inverse relationship between superior vena cava (SVC) flow and continuously monitored mean blood pressure (BP) in 30 preterm infants at 5 h postnatal age.

The inverse correlation between mean blood pressure and SVC flow volume persisted when correcting for birth weight (p = 0.04) and gestation (p = 0.03) on multiple regression analysis. The correlation also persisted when the two infants receiving inotropic support at the time of echocardiography were excluded from the analysis (R2 = 0.168, p = 0.02). Mean blood pressure was not significantly associated with volume of DAo flow (p = 0.23), LVO (p = 0.27) or RVO (p = 0.15) at 5 h postnatal age.

A weak but significant inverse correlation was also seen between volume of SVC flow and diastolic blood pressure (R2 = 0.135, p = 0.05) but not between volume of SVC flow and systolic blood pressure (p = 0.39). Neither the systolic nor the diastolic blood pressure was significantly associated with volume of DAo flow, LVO or RVO at 5 h postnatal age.

12-h scan

At 12 h postnatal age the volume of SVC decreased as mean blood pressure increased (fig 2). However this association was not statistically significant (R2 = 0.113, p = 0.06). The inverse correlation between blood pressure and volume of SVC flow became significant when the six infants receiving inotropic support at the time of echocardiography were excluded from the analysis (R2 = 0.180, p = 0.03). Mean blood pressure was not significantly associated with volume of DAo flow (p = 0.21), LVO (p = 0.12) or RVO (p = 0.22) at 12 h postnatal age.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Inverse relationship between superior vena cava (SVC) flow and continuously monitored mean blood pressure (BP) in 32 preterm infants at 12 h postnatal age.

Neither the systolic nor the diastolic blood pressure was significantly associated with volume of SVC flow, DAo flow, LVO or RVO at 12 h postnatal age.

DISCUSSION

We evaluated the relationship between blood pressure and echocardiographic markers of systemic perfusion. Critically, at no time in the first 48 h of postnatal life was there evidence of a positive association between blood pressure and volume of blood flow in any of the four vessels studied in this cohort of sick preterm infants. By including only blood pressure readings downloaded from indwelling arterial catheters over the exact duration of echocardiography we minimised the risk of inaccuracies in blood pressure measurement masking a true positive association between pressure and flow.

Our study suggests that if there is any association between blood pressure and volume of SVC flow (as a marker of systemic perfusion) in the first 12 h of postnatal life it is an inverse one. The data linking low SVC flow to high blood pressure in our cohort must be interpreted cautiously since the association was weak and was no longer significant at 12 h postnatal age. However, if genuine, an inverse relationship may be of critical importance as it is systemic hypotension in the immediate postnatal period that most frequently triggers intervention.20

Blood pressure is determined by a complex interplay of circulating blood volume and venous return (preload), myocardial contractility, systemic vascular resistance (afterload) and, in the presence of a patent ductus arteriosus, pulmonary vascular resistance.21 Given this interplay it is perhaps not surprising that the relationship between blood pressure and measures of tissue and cerebral blood flow in the transitional circulation is so variable in published reports.9 ,13–16 22–24 Specifically an inverse relationship between blood pressure and LVO in the first postnatal day in preterm infants has been noted previously.13 Systemic vascular resistance, which increases dramatically following clamping of the umbilical cord, may be the principal determinant of blood pressure in the transitional circulation of preterm infants.1 ,13 ,14 In addition contractility in the immature preterm heart may be particularly poor in the presence of high afterload.25 If resistance is dominant, and contractility vulnerable, an inverse association between pressure and flow would be expected. Therapeutic intervention may then need to target vasodilator inotropes such as milrinone26 rather than vasoconstrictors such as dopamine.27 It is interesting that diastolic blood pressure (the stronger marker of vascular resistance) rather than systolic blood pressure showed an inverse association with volume of SVC flow in our cohort.

A previous study of SVC flow volume in preterm infants demonstrated a weakly positive association between blood pressure and SVC flow at 5 h postnatal age.9 Similar techniques were used in the two studies, so the disparity in results is most likely due to differences in the cohorts rather than methodological differences. For optimal accuracy of blood pressure measurements both studies required an indwelling arterial line to have been inserted by the clinical team. However, in Kluckow and Evans’ study, arterial lines were inserted in a higher proportion of infants admitted to the NICU (personal communication, N Evans, 2007) suggesting that a relatively healthier cohort of infants may have been studied. Healthy preterm infants have a more normal circulatory adaptation28 and tend to have both higher blood pressure and higher SVC flow in early extrauterine life than sicker infants. A cohort which includes both healthy and sick preterm infants may therefore show a positive association between blood pressure and flow in the cohort as a whole, whereas a subgroup of sicker infants may show no association or an inverse relationship. The discrepancy in the relationships between blood pressure and flow in the two studies may also be due to differences in clinical management, including use of volume support, inotropes, sedation or indometacin.29

Studies have demonstrated that at least a proportion of sick preterm infants have deficiencies in cerebrovascular autoregulation,22 ,30 with increases in blood pressure being associated with increases in cerebral blood flow. Therefore the potential inverse relation between SVC flow (which includes cerebral venous return) and blood pressure described in our cohort initially seems counterintuitive. However, it is important to differentiate between cross-sectional studies which give a guide to the relationship between blood pressure and blood flow in a population at a single point in time and longitudinal studies which report changes in individual infants over time. Cross-sectional studies such as ours and others24 demonstrate that interindividual variation in circulatory status, particularly peripheral vascular tone, makes blood pressure a poor predictor of blood flow in the transitional circulation. Longitudinal studies demonstrate that some preterm infants cannot modulate their cerebral vascular tone in response to changes in blood pressure,30 that is they have deficient autoregulation. However, in these circumstances blood pressure may remain a poor screening tool for low cerebral perfusion on a cross-sectional basis as the level of overall vascular tone varies greatly between infants.

We have previously described a weak positive association between electroencephalographic (EEG) activity and measures of both blood flow and blood pressure in preterm infants.31 Although we have stressed that this association was weak, and may not be causal, the findings of the two studies do appear somewhat contradictory. That only 24 of the infants included in the cohort described in this paper were included in the larger cohort of infants studied with EEG may explain some of this inconsistency. The positive associations between blood pressure and EEG activity were predominantly seen at 12–24 h in our previous study, whereas a significant inverse association between blood pressure and SVC flow was only seen at 5 h in this study. Given the rapidly changing nature of the transitional circulation it is possible that behaviour of the systemic vasculature changes over this time period. In addition, in our previous study measurements of blood pressure were predominantly associated with trends in EEG continuity, whereas measurements of blood flow were predominantly associated with trends in EEG amplitude, reinforcing the complex nature of any link between blood pressure and cerebral perfusion.

Although there was no evidence of an association between mean blood pressure and left or right ventricular output in the first 48 h of postnatal life in our cohort there was a weak inverse association between diastolic blood pressure and both measurements at 48 h. This association was presumably produced by those infants with a large ductal and atrial shunts having lower blood pressure32 but higher LVO and RVO33 at this postnatal age. Further analysis of the impact of ductal shunt on cardiac output and systemic perfusion in this cohort will be reported in a subsequent paper.

Our study has a number of limitations. Importantly, echocardiographic measurements of cardiac output are prone to a degree of variability18 ,34 such that true associations can be masked.35 Similarly, erroneous associations may be generated, raising the possibility that the inverse relation of SVC flow and blood pressure seen in our cohort is a chance association. That an inverse relation between blood pressure and SVC flow was no longer seen at 24 and 48 h could suggest a chance finding. Data on other markers of circulatory health such as plasma lactate levels or mixed venous saturations could have assisted in the interpretation of results, but were not available at the time of the study.

Despite these limitations it seems reasonable to conclude that there is no evidence of a consistent positive correlation between systemic blood pressure and any echocardiographic parameter of systemic blood flow in this population of sick preterm infants, such that monitoring blood pressure cannot be a substitute for monitoring blood flow. To assist the study of novel interventions aimed at improving circulatory support and improving neurological outcome26 ,36 ,37 further examination of the transitional circulation in the preterm neonate is clearly required. In our view there is increasing evidence to support the judicious use of echocardiographic markers of circulatory status in addition to traditional cotside markers to guide optimisation of circulatory support.