Sometimes, EBM is accused of being slavishly devoted to the Randomised Controlled Trial. This is clearly garbage if you look to answer a question outside of therapeutics: see our Archi posts on diagnosis and prognostication, for example. But even within the setting of picking the right treatment for the patients you see, the RCT is not the only way of seeking after the Truth. (‘Heresy’ I hear cried – but it’s not – read on and be delighted or frightened.)
Jeremy Howick, Paul Glazsiou and Jeffrey Aronson have captured a concept that all practising EB paediatricians have used for some time, and set it solidly within an unassailable tradition of critical appraisal. They have revisited the Revised Bradford-Hill Guidelines (not criteria) for Causation and developed the parachutes of evidence: a trio of the main ‘chute (direct evidence), and two subsidiary ‘chutes (mechanistic and parallel evidence).
Direct Evidence is the stuff one you rely on most securely. This is when the RCT gives results, or an exceptionally effective intervention is impossibly better than all others (like when insulin was first used in DKA). Parallel Evidence is the stuff we use in paediatrics a lot: where similarities exist between the direct evidence in other populations and the ones we treat, and by replication of results across different areas. Mechanistic Evidence I still feel the shakiest relying on: coherence with other theories of how things work, and a plausible biological explanation. The fear of finding out our theory of how illness and the body works is wrong – again – haunts me in the guise of the inability of bacteria to live in the acidic stomach, and the nightmare of SIDS caused by babies choking on their own vomit. There are times though, when we can use nothing else. (How else would you work out how to treat pancreatoblastoma, for example?)
This clear conceptualisation of how the various guises that evidence exists in can link to help us work out what works, what works best, and what’s likely to work for this patient goes even further to helping the practicioners of EBM assist those who don’t currently work within this framework to see it’s not all smoke, mirrors and NNTs.
Acknowledgement: o0bsessed, under a CreativeCommons 2.0 licence