Premature infants with a patent ductus arteriosus (PDA) are at increased risk of more prolonged and more severe respiratory distress syndrome, bronchopulmonary dysplasia, and death than similar infants whose ductuses have closed. In an attempt to improve the outcome for these infants, three broad strategies have been proposed: prophylactic “closure”, closure of the asymptomatic but clinically detected PDA, and closure of the symptomatic PDA. Interventions, some of which have been subjected to more rigorous scrutiny than others, used as part of these strategies include fluid restriction, the use of diuretics, the use of cyclo-oxygenase inhibitors, and surgical closure using a variety of surgical techniques. There are now seven completed systematic reviews in the Cochrane library looking at such interventions.¹ The bulk of the evidence available assesses the impact on short term outcomes, with some of the strategies resulting in, for example, an apparent decrease in the risk of developing a symptomatic PDA or a reduction in duration of oxygen dependence. These short term benefits have to be balanced against adverse effects such as temporary renal impairment and a possible increase in the risk of necrotising enterocolitis. Importantly, there is no evidence available that any of the strategies examined result in long term benefit, in particular, disability-free survival, and for some strategies—for example, prophylactic intravenous indometacin—clear evidence exists that there is no long term benefit. Much uncertainty about “best practice” therefore remains.²

In this month’s journal, Brooks et al report an observational study that explores the outcome for preterm infants born at the turn of the century whose ductuses failed to close.³ Using statistical modelling and regression analyses to correct for potential confounding variables, they suggest that infants with a persistent PDA are at increased risk of dying compared with those infants in whom the ductus closes spontaneously or in whom it is closed medically (OR 4.02 (95% CI 1.12 to 14.51)). This seemingly dramatic result might naturally lead clinicians to wonder whether or not closing the ductus surgically in these babies would improve survival. It should be noted, however, that none of the strategies tested in randomised clinical trials to date that effectively close the duct seem to result in improved survival, and such a hypothesis would need to be tested more rigorously.

Brooks et al suggest that a robust assessment of this issue might best be achieved by conducting a randomised controlled trial presumably comparing surgical closure with non-closure in infants whose ductuses remain patent in the neonatal period despite attempt(s) to close them medically. The design of such a study would need very careful consideration, however. Cardiac services are now highly centralised in most western countries, and very few infants will be in a situation where they can have their patent ductus ligated surgically on the neonatal unit. In most cases, any consideration of surgery is likely to depend on transporting the neonate at the very least a few hundred yards down a corridor and more likely many miles, sometimes hundreds of miles, to the nearest cardiac centre. Although in some parts of the world neonatal transport services are highly developed, in others they remain ad hoc, and the quality of transport care along with its impact on neonatal outcome is relatively unstudied. What is clear, however, is that this aspect of any infant’s care—the transfer—would have to be considered an integral part of the intervention being tested. This would be analogous to the seminal UK Collaborative ECMO Trial⁴ where a coordinated retrieval service was established for the trial, and babies were randomised to either travel to an extracorporeal membrane oxygenation (ECMO) centre or to “stay put”. In addition, long term neurodevelopmental follow up was an integral feature of the trial.⁵ The whole episode of care, whether or not the infant ended up receiving ECMO or not, was then reported including an intention to treat analysis⁶ providing a true picture of the impact such a service might have on both short and long term infant outcomes. A similar study design may be appropriate to test this current hypothesis.

In the meantime, neonatologists will have to consider the options available. Are the short term gains in actively treating PDA worth the potential adverse effects when there is scant evidence of longer term benefit? And in relation to surgical closure when the duct remains open, particularly when this might involve interhospital transfer of an unwell baby, what might be the impact of the transfer itself? Certainly much more evidence is needed before this can be recommended as “best practice”.