[Research in epidemiology and causality: the example of the Basel Multiple Sclerosis Study 1980 to 1986]

Statistic evaluations on multiple sclerosis 1980 to 1986 in two Swiss cantons have been analyzed, using stochastic mathematical models in order to investigate significance of and possible reasons for variations in incidence. Fluctuation rates of up to 20% could hereby be demonstrated to be due to chance. This could be validated by computer simulations and comparisons with published data. The results show that such apparently large fluctuations in incidence rates should thus be attributed to possible causes with utmost care and prompt for more detailed rules for collection of data.