During the period of perinatal transition, preterm newborns may experience a transient circulatory deterioration characterised by low systemic blood flow (LSBF).1,2 The immaturity of the myocardium of preterm newborns may contribute to the development of low systemic blood flow, along with other factors such as inflammation (chorioamnionitis, sepsis), hypoxemia or changes in preload (mechanical ventilation, patent ductus arteriosus [PDA], etc.). Techniques like functional echocardiography or measurement of regional oxygen saturation allow the detection of abnormalities in systemic and cerebral blood flow that would otherwise go unnoticed.3–5 The clinical signs used routinely for monitoring (heart rate, blood pressure, etc.) are not sensitive enough to detect changes in systemic blood flow.2,5–8 Furthermore, other indirect markers of perfusion, such as serum lactate levels or urinary output, are of little usefulness in extremely preterm newborns during the transitional period.9 Low systemic blood flow tends to be transient and resolve spontaneously in the first 24–48h of life.10 However, there is evidence that this period of relative hypoperfusion may be involved in the aetiology of brain injury, with a demonstrated association between a superior vena cava flow (SVCf) of less than 41mL/kg/min and intraventricular haemorrhage (IVH), ischaemia-induced white matter injury, long-term neurodevelopmental impairment and death.1,3,11–15

There are interventions, such as delayed cord clamping or umbilical cord milking, that may reduce the risk of IVH.1,16–19 Some authors have proposed the use of inotropic drugs for prevention and treatment of LSBF during the transitional period, although there is no clear evidence of their impact on mortality or long-term developmental outcomes. Unfortunately, none of the studies conducted to date was sufficient powered, so no recommendations can be made on this subject.20–22

There is published evidence on the use of functional echocardiography in some neonatal units in Spain, but we do not know whether routine implementation of this method in everyday practice would have an impact on clinical outcomes.23

The aim of this study was to assess the impact of implementing a protocol for screening of LSBF and selective treatment with dobutamine (DB) during the transitional period on the incidence of severe IVH and/or mortality in low-birth-weight preterm newborns.

In the first 18 months following the introduction of the protocol, it was applied to 30 patients, whom we matched to 60 controls. We excluded 1 case and 6 controls for different reasons: prospective exclusion of 3 patients due to diagnosis of IVH in the first 6h (1 case and 2 controls), and exclusion of 4 eligible patients due to missing clinical data. Thus, the final analysis included 29 cases and 54 controls.

In the total sample, the GA at birth was 27.5 weeks (26.2–29.1) and the weight 990g (800–1180). Table 1 compares the characteristics of the patients in the two groups. We found no significant differences between groups in the variables under study except in the prenatal administration of magnesium sulphate, the proportion of pregnancies resulting from in vitro fertilisation and the number of transfusions received.

In the period that the protocol was implemented, 3/29 patients (10.2%) developed severe IVH or died, compared to 17/54 (31.5%) in the pre-intervention period (P=.032). There was a statistical trend to decreased incidence of severe IVH in the protocol period (3/29 [10.3%] vs 11/53 [20.7%]; P=.100), although we did not find significant differences in the incidence of overall IVH of any grade. A total of 13 patients died (3 in the protocol period and 10 in the pre-intervention period). In 3 patients (2 cases and 1 control), the cause of death was withdrawal of treatment due to severe IVH.

We found a decreasing linear trend in severe IVH and/or death that coincided with the implementation of the protocol (P=.071) (Fig. 3). We did not find this trend in the incidence of retinopathy, haemodynamically significant PDA, moderate-to-severe bronchopulmonary dysplasia or nosocomial sepsis, although we did find a decreasing trend in the incidence of enterocolitis (P=.093).

Figure 3.

Changes in the incidence of severe IVH and/or death in the period under study. The dotted line marks the introduction of the protocol for the detection and selective treatment of low systemic blood flow. The P-value refers to linear association assessed by the χ2 test.

(0.06MB).

In the multivariate logistic regression analysis, the implementation of the protocol was associated with a significant reduction in the incidence of severe IVH and/or death (OR, 0.11 [95% CI, 0.01–0.65], P=.015) after adjusting for potential confounders (Table 2). When we performed sensitivity analysis using inverse probability treatment weighting, we obtained a sample with an adequate balance on all the covariates included in the model (range of standardised differences, −8.6% to +8%). In the analysis of the weighted sample, the LSBF protocol was associated with a reduction in the incidence of severe IVH and/or death (OR, 0.23 [95% CI, 0.09–0.56]; P=.001).

The intraclass correlation coefficient comparing the investigators that measured MPAVmax by means of echocardiography was 0.992 (95% CI, 0.988–0.996; P<.001). During the intervention period, systemic blood flow was assessed by measurement of the MPAVmax in all patients, supplemented by measurement of the SVCf in 15 patients and the MPAf in 9 patients. Fig. 4 shows the linear correlation between the MPAVmax and the SVCf.

Figure 4.

Linear correlation between the values of MPAVmax and SVCf in the 15 patients in which both markers were measured simultaneously. We present the Pearson correlation coefficient and the P-value. MPA, main pulmonary artery; SVC, superior vena cava.

(0.06MB).

In the protocol period, 10/29 patients (34.5%) received a diagnosis of LSBF, which was made at a median of 4h post birth (IQR, 2–10), and were treated with DB starting with a dose of 6μg/kg/min (IQR, 5–10) (Table 3). We found no difference between these patients compared to patients that did not receive DB in GA, birth weight, Apgar score at 5min, incidence of respiratory distress syndrome (RSD) or duration of mechanical ventilation in days; however, they did have a significantly higher CRIB score at 12h post birth (7 [4–9.2] vs 1 [1–5]; P=.03). After initiation of DB, there was a significant increase in systemic blood flow (MPAVmax pre-DB, 0.40 m/s [0.32–0.42] vs post-DB, 0.60 m/s [0.53–0.60]; P=.018). None of the patients treated with DB developed severe IVH or died. Conversely, 3 patients in the group that did not receive DB developed severe IVH. The values of MPAVmax in these patients ranged between 0.47 and 0.53 m/s, compared to a median of 0.60 m/s (IQR, 0.52–0.80) in patients that did not develop IVH. Out of the patients that did not receive DB, functional echocardiography was performed to assess changes in systemic blood flow in only 10 (data not available for the 3 patients with severe IVH).

In this study, we assessed the impact of the implementation of a protocol for screening for LSBF by means of functional echocardiography and its selective treatment with DB in the first 18 months from its introduction. We found a decrease in the incidence of severe IVH and/or death that was not associated with any adverse events.

The gold standard for diagnosis of LSBF is measurement of blood flow in the superior vena cava by means of echocardiography.4,34,35 However, this is a technically complex measurement, and thus may be difficult to implement in everyday clinical practice. The main source of error is the measurement of the diameter of the vena cava, as this value is squared to calculate the flow, and some authors have questioned the validity of this method for the purpose of clinical decision-making.31,35–38 Measurement of the MPAVmax has been proposed as an alternative for estimating systemic blood flow. The MPAVmax is proportional to cardiac output and is strongly associated with systemic blood flow in the first days of life, when the volume diverted through the foramen ovale is small. It is much easier to obtain than the SVCf and does not require any calculations or measurement of dimensions that increase the risk of error. A peak velocity of less than 0.35 m/s is indicative of LSBF in most patients, while about half of patients with a peak velocity of less than 0.45 m/s may have LSBF.39

Despite the mounting evidence on the association of LSBF with brain injury and mortality, its prevention and treatment through the use of inotropic agents remains controversial. In 2007, Osborn et al.13 demonstrated that treatment with DB is more efficacious in increasing systemic blood flow and reducing the incidence of IVH compared to dopamine, while a different clinical trial did not find significant increases in systemic blood flow with the use of milrinone.21 However, there is no evidence that the use of inotropes is associated with an improvement of long-term neurologic outcomes. Unfortunately, none of the studies conducted to date had sufficient power to detect differences in morbidity and mortality.20 In 2015, Bravo et al.40 conducted a pilot randomised-controlled trial where they assigned preterm newborns with LSBF (SVCf<41mL/kg/min) to either DB (n=16) or placebo (n=12). The study also included a control group of 98 preterm newborns with normal SVCf values. They found that LSBF was associated with higher mortality and an increased incidence of severe ischaemic events. The incidence of IVH was 12% in the DB group compared to 33% in the placebo group. Although the differences were not statistically significant due to the low statistical power of the study, its results were very similar to our own findings. We believe that a larger study could find significant differences in the incidence of this important clinical event.

Our results were promising but must be interpreted taking into account the limitations of our study. First of all, it was a single-centre study with a retrospective control group. We cannot be completely sure that the decrease in the incidence of severe IVH resulted solely from the implementation of the LSBF management protocol. Factors like additional improvements in clinical practice during the period under study or the possibility that more attention or care was given to the patients in the intervention group (Hawthorne event) may have influenced the outcomes. We attempted to minimise these effects intrinsic to the study design by selecting matched controls and analysing potential confounders. We did not find statistically significant differences in most of the baseline characteristics or in morbidity between the groups, which suggests that the 2 cohorts were comparable. The incidence of other morbidity, which could indicate an overall improvement in care, did not decrease during the period under study, whereas we did observe a decreasing trend in the incidence of IVH that coincided with the implementation of the protocol. Lastly, the multivariate regression analysis and the model using inverse probability treatment weighting found an independent association between the protocol and the decrease in severe IVH and/or death after adjusting for potential confounding factors.

The diagnosis of LSBF and the prescription of DB were mainly based on the measured value of the MPAVmax. There are no validation studies of MPAVmax against SVCf, which is considered the gold standard for diagnosis. Nevertheless, we found a strong linear correlation between the MPAVmax and the SVCf, which needs confirmation by studies specifically focused on this aspect. In addition we found an excellent interrater agreement in the measurement of the MPAVmax. Our protocol did not include monitoring of changes in blood flow in patients that did not receive DB. In fact, the three IVH events occurred in patients in this group. These newborns had initial values of MPAVmax ranging between 0.47 and 0.53m/s, which were barely above the threshold we had established for indication of DB. It is possible that these patients had LSBF that went undetected. For this reason, we consider that monitoring changes in blood flow with functional echocardiography is crucial in these patients, especially when the initial values are near the threshold for treatment.

Performance of prospective multicentre studies, preferably clinical trials, with sufficient statistical power is necessary to establish whether interventions in the management of LSBF such as administration of inotropes have a favourable impact on the prevention of IVH and long-term neurologic outcomes.

In this study with retrospective controls, the implementation of a protocol for the detection and selective treatment of LSBF during the transitional period was associated with a reduction in the incidence of severe IVH/death in preterm newborns delivered before 30 weeks’ gestation. Due to the study design, we were unable to rule out the potential contribution of other factors to the observed improvement.

The authors have no conflicts of interest to declare.