Some historical aspects

Corticosteroids, in the form of hydrocortisone, were first used in neonatal medicine in the 1970s to try to modify the course of respiratory distress syndrome.1 No clinical benefit was shown but the treatment appeared to be without detrimental effects until two follow-up studies were published within 2 years.2 3 These studies showed increased risks of intraventricular haemorrhage2 and neurosensory and encephalographic abnormalities.3 A decade later, corticosteroids, in the form of high-dose dexamethasone (0.5 mg/kg/day), were tried again this time to treat preterm infants with bronchopulmonary dysplasia (BPD).4 5 These two small studies used sequential analysis and appeared to show benefit as regards weaning from ventilation but apart from infection4 adverse effects were not reported. In the next decade or so high-dose dexamethasone was studied in many randomised trials given early (in the first week) to try to prevent BPD or later (after the first week) to treat it.6 Meta-analyses of randomised controlled trials (RCTS) showed that dexamethasone facilitated extubation from ventilation and reduced BPD, although adverse effects of hypertension, hyperglycaemia and growth failure were considered to be both tolerable and reversible on stopping steroids. In 1998, Yeh and his colleagues from Taiwan first reported longer term adverse effects of dexamethasone on neurodevelopment,7 later confirmed by Shinwell and coworkers from Israel.8 Hence, the dilemma for the neonatologist (and parents): choose between reduced risk of BPD but an increased risk of cerebral palsy as enunciated in the early 2000s.9

A number of questions remained unanswered in the early 2000s, for example:

• Was early and late dexamethasone treatment equally harmful?

• Were dexamethasone and other steroids such as hydrocortisone similarly harmful?

• Was the high dose of dexamethasone (0.5 mg/kg/day) problematic?

• Would lower doses of steroids be equally effective but less harmful?

• What is the optimal duration of …