In the absence of well designed paediatric studies we need to judiciously apply what we can extrapolate from adult studies to each child we are treating; selecting interventions and approaches carefully, adjusting them when there is no evidence of individual benefit.