Don't trust statistics: Why «significant» is far from important
«Survived one week longer – statistically significant!» Headlines like this sound like medical progress. But what does this actually mean for patients? Many statistically verifiable differences in medicine are measurable, but hardly noticeable in the everyday lives of those affected.
Statistical significance merely shows that a difference exists, not whether it is relevant to people. For years, the research team led by Prof. Milo Puhan (University of Zurich) and Prof. Pierre-Alain Clavien (University of Zurich and Swiss Medical Network, Privatklinik Bethanien, Zurich) has been committed to a more patient-centred evaluation of medical outcomes. An international consensus conference on this topic was held in Zurich in 2022 (published in Nature Medicine). Objective quality markers such as the Clavien-Dindo classification (CD) and the Comprehensive Complication Index (CCI®), now established worldwide as instruments for assessing the quality of outcomes of medical interventions, are important milestones on this path.
In a new study, the team has now determined the Minimal Important Difference (MID) of the CCI® for the first time, i.e. the threshold at which differences in the extent of complications are actually noticeable for patients. The result: a change of 12 points in the CCI® marks the point at which the burden of complications becomes clearly noticeable for those affected.
This finding is groundbreaking for the planning and interpretation of clinical studies. It helps to evaluate medical treatments not only according to statistical indicators, but also according to their actual benefit for patients, a decisive step towards a form of medicine that is both scientifically precise and humanely relevant.
