On evidence, embellishment and efficacy

Evidence‐based medicine (EBM) — to use a vogue word — is the contemporary mantra of many academic physicians. Unfortunately, the evidence they consider, although superficially convincing, is often slanted, occasionally deliberately, but more often as a result of carelessness and defects in study design. These biases and confounding factors are often hidden and arcane and seriously impair objective assessment. Such defects can only be detected when all investigators agree to go along with Bradford Hill's dictum, namely ‘It is the essence of science to disclose both the data upon which a conclusion is based and the methods by which the conclusion is obtained’; a cardinal rule more often honoured in the breach than the observance.