EBM Episode 1: Setting expectations

This is the first episode in our series on EBM. In it we explore what can evidence based medicine do and what it cannot.

In EBM, large cohorts of patients are used to predict likelihoods of, say, a drug to cure a disease. These cohorts come from clinical trials, systematic reviews and several other types of studies.

EBM is a decision aide for clinicians to provide patient-centred care. It is compatible with a shared decision model.

EBM does not:

• replace empathy, listening and caring and so automation of EBM is not replacing clinicians
• predict outcomes beyond probabilities
• take all available data into account (it ignore data for which there is no evidence)
• account for changing environments yet (e.g. antibiotic resistance is reducing the efficacy of drugs, or when the efficacy placebo is rising)
• come free.

Some of these are well beyond the realm of EBM. The others are temporary limitations that will be met in the future through better technologies, more data, and incremental improvements to evidence gathering and dissemination processes.

We will address these in future episodes of this series.

For more reading (paywall, sorry) see the classic: Evidence based medicine: what it is and what it isn't