Improvements in life expectancy are associated with the education level of the population. Both kinds of in-dicators are also associated with socio-economic status. However, an independent, direct effect of educationstatus on health status can also be documented. Some issues that may explain this relationship are the abilityto recognise disease as an objective process that can be influenced through practical measures and the abilityto operationalise the probabilistic notion of risk and its implications. Physician performance is related to theability to follow probabilistic logic and statistical arguments, as most scientific results on which clinical practiceis based are probabilistic. We review a range of widespread errors of judgement regarding the objective natureof organic diseases, the principles of pharmacologic effect and the ability to monitor its effects, atitudes towardsmedical errors, common (mis)perceptions of causality among health professionals and the general population,and issues in the comprehension and application of published research results. Probabilistic logic and scientificeducation in general need to be improved at all levels, and in particular in medical training and research, in orderto achieve a further improvement of the general health status and life expectation. Research fields and thematicsthat need to be encouraged include cultural anthropology studies about the concept of disease and treatment,the study of medication errors, of the logic of clinical arguments and their perception in the patient population,as well as interdisciplinary approaches aiming to improve probabilistic models of disease and treatment such aspharmacometry, epidemiologic modelling or systems biology.