Soz.- Pr~ventivmed.43 (1998) 162-166 0303-8408/98/030162-05 $1.50 + 0.20/0 9 BirkhauserVerlag, Basel, 1998 Hans L. Rieder 1, Marcel Zwahlen 2, Hanspeter Zimmermann 2 7 International Union Against Tuberculosis and Lung Disease, Paris 2 Federal Office of Public Health, Bern-Liebefeld Mortality from respiratory tuberculosis in Switzerland The epidemiologic situation regard- ing tuberculosis in Switzerland has improved remarkably in this cen- tury. Since 1945, the average annual risk of infection with Mycobacteri- um tuberculosis has decreased each year by an estimated average of 11.4%, and tuberculosis notifica- tions by 4.9 % from 1945 to 1973, and by 5.7% from 1975 to 1988. Tuber- culosis mortality decreased on aver- age 7.5% per year from 1952 to 19861 . Before the introduction of systema- tic tuberculosis case notification systems, mortality data on tubercu- losis were usually more readily available than morbidity data. They also provided a reasonably reliable measurement of the extent of the tuberculosis epidemic in a commu- nity, because the natural history of tuberculosis without intervention resulted in a morbidity-to-mortality ratio that was fairly constant 2. To ascertain the course of the tuber- culosis epidemic in Switzerland using records of mortality, publish- ed data on deaths from respiratory tuberculosis were scrutinized to portray more precisely how tuber- culosis has receded. Materials and methods The Federal Office of Statistics pub- lishes health statistics for Switzer- land. Since 1901 this institution has regularly published tuberculosis mortality data by age and sex, sepa- rately for respiratory and other forms of tuberculosis. For the pre- sent analysis, only data on mortality from respiratory tuberculosis were utilized. The analysis was limited to respiratory tuberculosis because of the generally more reliable diagno- stic accuracy, and because overall tuberculosis mortality is largely determined by deaths from respira- tory rather than extra-respiratory tuberculosis, as the incidence of the former considerably exceeds that of the latter 3. Deaths among per- sons aged 80 years and older were omitted from the analyses of rates, because the small numbers and the resulting large variation in mortali- ty rates in the 80- to 89-year-old and 90- to 99-year-old age groups precluded meaningful evaluation. Mortality rates were used as pub- lished, or recalculated from numera- tor and denominator data. Mortali- ty rates in children less than 10 years of age had to be aggregated for easier presentation. Utilizing infor- mation on tuberculosis mortality for the seven 10-year age groups from age 0 through 79 years and the ten cross-sectional reports from 1901 to 1991 allowed the evaluation of deaths from respiratory tuberculo- sis for birth cohorts from 1836 162