Contenu du sommaire

Paolo Vineis: Causality in medicine: theoritical models and practical problems. In the period when the dissection of corpses was the main basis of medical knowledge, doctors' interest was focussed primarily on the way in which disease developed and took hold in the body of a single individual. Subsequently, the successes of microbiology shifted attention to the conceptual framework of an infective agent and its host, implying a simple, linear notion of causality. Then, from the 1950's on, epidemiological studies of degenerative illnesses like cancer and heart disease raised new problems. Researchers became more aware that one disease may be due to a number of causes, and that the same exposure can induce a number of diseases, and they started to conceive of the contraction of an illness on the basis of a probabilistic model. This paper reflects on the methodological difficulties introduced by this complicating of the idea of cause. Today, rather than a single cause, we are more likely to talk about a grid or «network of causation» - a concept which takes into account the effects of a number of agents and of the interaction between them. Also the criteria of disease classification need to be considered critically. In the case of cancer, we are still using a «manifestational» classification (based on clinical and pathological features). However, both the définition and classification of cancer seem to fit a «polytethic» rather than a «monothetic» model similarly to Wittgenstein's «family resemblances» or to the «taxa polytethica» used in zoology. Finally, the paper describes the role of analogies and metaphors in the génération of new hypotheses, in the context of a complex inferential process in which the use of statistical methods is only one of the tools.

Anne Chevalier, Marcel Goldberg: Absenteeism: social indicator or heaith indicator? On the basis of a defined context - the company Electricité de France - Gaz de France - in which the medical reasons for absenteeism are recorded, the authors attempt to distinguish between absenteeism for medical reasons and that due to socio-professional factors. Various measures of time off work were calculated and analyzed as a function of individual and collective parameters. Among these measures, the duration of absence from work emerged as an indicator of morbidity, whereas the percentage of the personnel absent during a given period was influenced by socio-professional factors. It is as if, in addition to a «hard nucleus» of absenteeism due to illness, more immediate factors related to the individual himself or to the work environment generate short-term absenteeism which, in turn, could be the first step in a process of health deterioration.

Dominique Boullier, Jacques D. de Certaines: The art of socio-technical compromise in hospital innovation: the case of Picture and Archiving Communication Systems (PACS). The history of certain pioneers in Picture and Archiving Communication Systems (PACS) provides useful information on the contradictory objectives and interests that are at play when innovations are introduced in the hospital environment. The Rennes project on which this paper is based was highly ambitious, and also generated extreme conflict during its development. The participants in the project were marginal vis-à-vis both the personnel in the radiology department and their industry partners, and it is not therefore surprising that no marketable product was produced. The main resuit was an increase in their power and notoriety within the hospital, which may, in itself, form a type of innovation, in which the technical considerations simply served as a lever for institutional change which, in turn, would lead to «technical» improvements. The authors propose indicators for evaluating technical choices adapted to medical imaging networks.