Children are mostly little adults

I know we get told for most of the first 18 months in paediatrics how children are importantly different than adults, and neonates are different than children, and teenagers are not quite the same as Martians but have strong similarities, but… perhaps we might be protesting a bit too much. When we are dealing with evidence for use in clinical practice, we need to think about not only justifying our existence as paediatricians, but also how much value we have in extracting information from adult studies.

We have spoken before about using common sense and proper science—like understanding pathophysiology and pharmacokinetics—in making a good guess about how much we can believe adult evidence.1 We can go beyond that too, in some cases, where we have got trials in both grown-ups and child patients in comparing just how similar or different they are.2 There are different mathematical techniques we can suggest for this: do we look at each group differently and just eyeball them?; do we work out the average extra benefit (or disadvantage) you get from being a child treated with the same drug?; do we look at doing a regression analysis with age as a potentially predictive variable?

Which you choose depends loads on exactly what you want the information for and how much data and time you have. Researchers who are setting up new trials or policymakers deciding to commission a new monoclonal antibody might well need something more precise and robust than a team looking to use a fourth-line therapy in an exceptional 10-year-old. At the moment, it is probably reasonable to use educated guesswork about when children are and are not little adults—but the time is hopefully coming where we can use statistically enabled guesswork to make things even better.