Statistics from Altmetric.com
I read with interest the article by Broughton et al1 and wish to offer the following comments.
(1) The duration of oxygen therapy (in both table 1 and the text) ranges from 30 to 107 weeks, thus qualifying every baby in the cohort as having bronchopulmonary dysplasia (BPD). Even if this was in days, it would make every baby oxygen dependent 28 days after birth compared with 19% in other studies.2
(2) The rate of BPD in this cohort was 40%, which is much higher than has been reported for babies of similar gestation in recent years.2,3 In view of this high rate of BPD and prolonged oxygen dependency, the studied cohort might have consisted of rather sick infants.
(3) Since discharge home on oxygen is mentioned as one of the explanatory variables, it would be helpful to know its frequency and significance in relation to the risk of hospital admission and RSV infection.
(4) It is not clear if the healthcare utilisation includes all events after discharge from the neonatal unit or only those after an RSV infection. The authors have elsewhere suggested that babies with lung function deficits at discharge from the neonatal unit are more likely to sustain symptomatic RSV infections.4 If the healthcare utilisation includes all post-discharge events, then is it possible that the excessive healthcare utilisation of RSV infected infants is a manifestation of their underlying lung deficit rather than an effect of RSV?
(5) The authors suggest consideration of palivizumab for the two risk factors of RSV lower respiratory tract infections—maternal smoking during pregnancy and the presence of school aged siblings. However, palivizumab has been shown to be effective in “reducing hospitalisations from RSV” rather than “preventing RSV infection” itself,5 and the validity of this extrapolation remains to be tested. Indeed, the indicator for healthcare utilisation (GP attendances) just reached significance among non-admitted RSV infected infants compared with infants with no LRTI, the use of reliever medications being comparable in all three groups. I am not sure that widening the indications for palivizumab to the suggested groups will prove to be cost effective.
We thank Dr Deshpande for his letter regarding our paper published recently in Thorax.1 Dr Deshpande is concerned about the cost effectiveness of palivizumab and states this as a conflict of interest. We wish to emphasise that it was not an objective of our paper to assess the cost effectiveness of palivizumab but rather to assess any respiratory morbidity following RSV infection in prematurely born infants.
In response to his specific comments:
(1) We apologise for the fact that the duration of oxygen therapy was given in a confusing fashion as postmenstrual age (weeks), not as number of weeks since birth.
(2) The 40% rate of BPD at 36 weeks postmenstrual age is very similar to the 46% rate reported recently.2
(3) We have previously reported that discharge on home oxygen increases subsequent healthcare utilisation.3
(4) We have recently reported4 that diminished lung function is a risk factor for RSV infection and subsequent respiratory morbidity, but in that paper we also found that RSV infection was an independent risk factor for days of cough and wheeze.
(5) We agree it is very important to find effective ways to stop antenatal women and householders of premature babies smoking. As our data show, current methods are clearly ineffective. From the results of all studies, hypotheses are generated and need to be tested—hence our comments regarding consideration of giving palivizumab to infants who have siblings and whose mothers smoked during pregnancy. We hope our comments will encourage researchers to undertake an appropriately designed study to test this.
Competing interests: The author has previously questioned the cost effectiveness of palivizumab in relation to its recommended indications.
Competing interests: none declared.