The study by Olivier Baud and colleagues provides additional insight into the use of hydrocortisone in the prevention of bronchopulmonary dysplasia in premature infants. The large, multicentre, double-blind randomized controlled trial showed that, in extremely preterm infants born before 28 weeks, the rate of survival without bronchopulmonary dysplasia (defined as a requirement for respiratory support and oxygen requirement at 36 weeks of postmenstrual age) was significantly increased by prophylactic low-dose hydrocortisone given in the first 10 days of life. The authors suggested the need for further studies to evaluate the use of low-dose hydrocortisone in premature infants between 24 weeks and 25 weeks of gestation, who had a significantly increased incidence of late-onset sepsis in the hydrocortisone group versus the placebo group (30 [40%] of 83 vs 21 [23%] of 90 infants; sub-hazard ratio 1·87, 95% CI 1·09–3·21, p=0·02).

This study by Anne Monique Nuyta and Bernard Thébaud, which praised the quality of the study and highlighted the importance of additional findings that low-dose hydrocortisone reduced the need for persistent ductus arteriosus ligation and did not increase the rate of intestinal perforation unlike in previous similar studies.

The finding that low-dose hydrocortisone was associated with an increase in late-onset sepsis in infants at 24–25 weeks gestational age poses substantial ethical challenges to further research in this gestational age group. Future researchers will need to justify to human research ethics committees and parents that the benefits of carrying out further studies outweigh the harm (of late-onset sepsis). On one hand, researchers might argue that they are in equipoise—ie, they do not know which steroids and in what doses are most suitable to prevent and reduce the mortality and morbidity from bronchopulmonary dysplasia. On the other hand, in human research, “the risk of harm must be no greater than that encountered in ordinary life”.4 In this case, late-onset sepsis is potentially life threatening. Unfortunately, this gestational group happens to be the one most at risk of bronchopulmonary dysplasia.

The second ethical challenge is that justification of the use of placebo is difficult when early treatment with other corticosteroids, such as dexamethasone, have been shown to help extubation and reduce the risk of chronic lung disease, although some short-term complications and long-term neurodevelopmental complications might occur including cerebral palsy.5 Comparison of low-dose hydrocortisone to an existing treatment, such as low-dose dexamethasone, might be more suitable, because dexamethasone is widely used in neonatology at present.

The long-term follow-up of infants from the study by Baud and colleagues will certainly be useful to establish whether hydrocortisone has any adverse neurodevelopmental consequences. Ethical considerations are, however, likely to hamper further research on low-dose hydrocortisone use for the prevention of bronchopulmonary dysplasia in the infants most in need—the most premature infants.