Can you do evidence based medicine without evidence?

Where does evidence based medicine go without evidence? I have recently been working with lots of folk who manage children with disability and long-term conditions, a massive group of heterogenously named conditions with sometimes similar and sometimes different problems. The teams working with them are passionate and committed and frequently want to use research. Then they go looking for it and frequently find nothing. Now—the experience of not finding evidence to help your question is not unique to disabilities—there are plenty of uncertainties in children's cancer, neonatology and diabetes. How can you move on without evidence in an evidence based way? We can take this forward in two ways. First, acknowledging the lack of really good evidence, we should use indirect, analogous and ‘lower level’ evidence.1 Responding to the clinical problem will be combining what data we do and do not have with discussions with the parents and child/young person. Second, we might use this to generate a clinical study question and seek folks to answer it. Evidence-based medicine is certainly not clinical research, but many really important clinical trials develop from the gaps identified by this sort of approach. Do not be upset when you find nothing—it's not that uncommon2 to go searching and find very little—but remember that evidence based medicine is possible, even without obvious evidence.