The answer to the first question is traditionally ‘twice as long as half a piece of string’ which is supposed to be funny.
Equally difficult is the second question, when is a patent ductus arteriosus (PDA) hemodynamically significant? We could answer ‘when there is too big a shunt’ but that also begs the question, how big is too big, and how do we measure it accurately. But is that what is really important about a PDA, is the size of the shunt the most important factor? So what is a clinically significant PDA? In the past we have treated the ductus if we thought by doing so we could reduce the risk of adverse outcomes, in particular BPD. The problem with that is that although there is a correlation between the development of BPD and having had a PDA, there is no good evidence that any treatment of the PDA reduces the risk of BPD.
A systematic review of definitions of what is a hemodynamically significant PDA was published in 2011. Those authors found that was no consistency in definitions of hemodynamic significance, sometimes it was based on clinical findings, sometimes on echocardiographic features and those echo features were very variable. Even when the same index was used (most commonly the LA:Ao ratio) the threshold for significance was inconsistent. This means of course that you cannot compare the results of the various trials. Most often in fact, no definition was given in the articles they reviewed.
A new study from Monash in Melbourne has tried to answer some of these questions. The authors compared the echocardiographic data they have recorded, from an extensive standardised evaluation of the PDA performed within the 48 hours prior to the first treatment with ibuprofen, between infants who did and did not develop BPD. (Sehgal A, Paul E, Menahem S: Functional echocardiography in staging for ductal disease severity : Role in predicting outcomes. Eur J Pediatr 2013, 172(2):179-184). Their scoring system includes items for size of the duct, size of the shunt, and impact on left ventricular function.
All of the individual items in the scoring system were significantly different between those infants who developed BPD and those who did not, except for left ventricle to Aortic root ratio.
The authors showed a progressively increasing risk of developing BPD with increasing scores, and a pretty good predictive function of the score. I think this is really useful information; if you have a bigger PDA with more shunt, and the left ventricular function is more affected by it, then you are more likely to develop BPD. It has never been clear whether the statistical association between PDA and BPD was causative in any way, this new data suggests that it might be, and perhaps only among those with the PDAs that have larger shunts.
We could use this sort of assessment as the basis for clinical trials of PDA treatment.