It would be a wonderful thing if every treatment we used had been tested in trials where the populations matched ours exactly. Sadly, this isn’t the case. In paediatrics, the evidence we have may be in the “wrong” population: including lots of adults or children of the wrong age. Or the outcomes recorded may only be surrogates, rather than clinically important changes. In order to use the best evidence in practice, we need to consider how far we can take the results “beyond the evidence”.

Fortunately, as with many aspects of critical appraisal, there are guides as to how to think about the issues related to using studies which don’t directly apply to our population.¹ It is suggested that first, we should ask if there are biological differences between the populations. For example, does the same process produce cradle cap in babies as seborrhoeic dermatitis in adults? Here it may be relevant to look at pathological data, or compare the results of studies of alternative treatments in the populations under suspicion.

Second, it is appropriate to consider whether differences in psychology, social setting, or economy will stop the data being applicable. When we turn to psychological differences it is clearly inappropriate to use a cognitive-behavioural therapy in infants, but how should we appraise a trial which shows improved quality of life in adults? If there are significant differences in economic or social setting, it may strongly affect the family’s adherence to a therapy.

If the treatment seems to be feasible and sensible, we are suggested to address issues of risk and co-morbidity. If COX-2 inhibitors do reduce the chance of gastrointestinal bleeding,² should we be using them in children with juvenile arthritis? We need to know the basic risk of GI bleeding in our population, in order to estimate the benefit they may gain from using the new drugs.

The last issue to consider is that of outcomes. What is the information on side effects? Is there any information about adverse events in children? Are the outcomes we are given directly relevant to our patients (such as improved function in JIA) or surrogate outcomes (such as reduced serum CRP)?

As with everything in evidence based practice, these guides don’t give you the rules to act on, but tools to think through. When considering if you can go “beyond the evidence”, look at biological and psychological differences, consider the inherent risk and co-morbidities, and examine all the outcomes closely. Then you’ll have a better idea of how far you can apply “best evidence” to your practice.