Delivery room research: when does poor quality evidence become an ethical issue?

As Drs Wootton, Arnold, and Tyson acknowledged, many emergency therapies have never been rigorously studied. Neonatal resuscitation medicine is especially bereft of highlevel evidence. Only 23 of 157 (15%) publications cited in the 2010 ILCOR statement on neonatal resuscitation were randomized trials or metaanalyses of trials (level of evidence [LOE] 1) (Table 1). In contrast, 114 (73%) studies lacked a control group (LOE 4), or worse, were performed in a different population, in animals, or in a mechanical model (LOE 5). We urgently need well-designed studies of the safety and efficacy of delivery room interventions in newborns.