Alternative consent processes in a neonatal resuscitation trial (SAIL): secondary analysis of an open-label, international, multicentre, randomised trial.

backgroundDelivery room trials face ethical and logistical enrolment challenges, including requirements for prospective, antenatal, parental consent. The Sustained Aeration of Infant Lungs (SAIL) study was a randomised controlled trial of two resuscitation strategies at birth, enrolling infants using both prospective antenatal consent and deferred postnatal (consent-to-continue) pathways. We aimed to compare recruitment and outcomes between SAIL trial centres with deferred consent available versus centres only using prospective antenatal consent.

methodThis study is a secondary analysis of the data from the open-label, international, multicentre, randomised SAIL trial. Infants born at 23 to less than 27 weeks' gestation at 18 centres across nine countries who were deemed to require intermittent positive pressure ventilation due to inadequate respiratory efforts or bradycardia, had no known major congenital anomalies, and were neither stillborn nor considered to be non-viable by their clinician were eligible for inclusion. Centres were compared by consent mode availability: antenatal consent only versus centres with deferred consent available. The primary outcome was the combined rate of death or bronchopulmonary dysplasia compared between consent type groups. People with lived experience did not contribute to the design or implementation of the SAIL trial. The SAIL trial was registered with ClinicalTrials.gov (NCT02139800).

findingsThe SAIL trial recruited 426 infants (222 [52%] were male and 204 [48%] were female) born from 371 mothers (218 [59%] were White) between Aug 27, 2014, and Sept 14, 2017. In 12 centres using only antenatal consent, 197 (41%) of 479 eligible infants were recruited and included in analysis compared with 229 (73%) of 313 eligible infants from six centres where deferred consent was available (with or without antenatal consent; absolute difference 31·78% [95% CI 24·46-39·10; p<0·0001]). Deferred consent was not obtained from 34 (15%) of 225 infants randomly assigned via the deferred pathway, who were excluded from this analysis. Compared with centres using only antenatal consent, infants recruited at deferred consent centres had less exposure to antenatal corticosteroids (166 [84%] of 197 vs 167 [73%] of 229 infants, p=0·012); were more likely to be male (88 [45%] vs 134 [59%], p=0·0065); were heavier (median birthweight 687 g [IQR 558-817] vs 767 g [661-870], p<0·0001); were less likely to be growth restricted (34 [27%] vs 19 [8%], p=0·0075); and less frequently intubated at birth (125 [63%] vs 105 [46%], p=0·0005). The rate of death or bronchopulmonary dysplasia at deferred consent centres was 65% (149/229) versus 57% (113/197) at antenatal consent centres (adjusted relative risk 1·18 [95% CI 0·98-1·42]).

interpretationCentres where deferred consent was available recruited a higher proportion of eligible infants at high risk who had more morbidities, reflecting a broader, more generalisable sample of the population compared with centres that only recruited using antenatal consent.

fundingEunice Kennedy Shriver National Institute of Child Health and Human Development.