[Appearances deceive in investigations of cancer clusters].

Many a doctor from time to time encounters people concerned about environmental causes of disease, whom he or she cannot answer properly because of lack of knowledge concerning the effects of chemical or physical exposure of the human body (nuclear plants, environmental pollution, electromagnetic radiation). Usually post hoc cluster investigation is very unrewarding especially when there is no clearcut hypothesis or evidence of a causal relation, and when the relative risk is well below 8. From an epidemiological point of view it is surprising that an unexpectedly low frequency of a certain disease in a particular region does not attract this kind of attention. Concerned people will most probably benefit more from risk communication by environmental epidemiologists than from cluster investigation or extensive case-control studies ('fishing expeditions') in case little is known of the etiology.