The paucity of scientific data underpinning medical practice

Ideally, medical decisions should be backed up by underlying scientific data, preferably obtained from randomised clinical trials, to ensure that medical decisions are valid and reliable and lead to more benefit than harm. This is seldom the case. In this paper, we argue that only 10% of medical decisions are supported by scientifically sound data obtained from clinical trials without important limitations. We base this on how often medical doctors adhere to clinical guidelines, how often guideline recommendations are supported by clinical trials, and how often randomised clinical trials have important defects rendering their results invalid. We must acknowledge and embrace this uncertainty related to clinical practice and use it as a catalyst to conduct more – and better – randomised clinical trials to reduce the 90% knowledge gap that dominates our healthcare system.