In 1969-1970 our random sample was drawn in the population around Stockholm, Sweden. The population aged 18-65 years old was stratified into three age strata and a random sample was drawn. Using sample fractions in the proportions 3 : 2 : 1 the highest fraction used for the youngest stratum. This sample comprised 32186 individuals and constituted our cohort in the present study. Their records in the in-patient register and a postal questionnaire was used for substratification into four strata; one of expected great needs for services, one of expected lesser needs, one expected to be free from needs for services and lastly one substratum of individuals not having been in-patients and not having filled up the questionnaire. In these four strata a random sample was drawn of 3064 exposed for health screening in the present study. They received a questionnaire survey which provided data from 30 questions about somatic, social and health history including three questions about alcohol use. The health survey focused specially on social, physical and psychical functions and diseases. The examinations was made by social workers, physicians and psychiatrists during one day. Laboratory tests etc was taken.
A 25 year follow-up of this population has now started at the Family medicine department at the Karolinska institute, Novum, Huddinge hospital, Stockholm in cooperation with the Social medicine department at Umeå university, Sweden. The latest article accepted for publishing is an artikel in the subject: Is one general health screening effective on changing total mortality?
The article has the title: "Effects of an assessment of needs for
medical and social services on long-term mortality. A randomized
controlled study" and has be printed in the
International Journal of
Epidemiology 1998;27:194-198
Background: The aim of this study was to investigate the long-term effects of one general health screening on mortality.
Method: After stratification and randomization of a population of 450 000 inhabitants two groups were formed, an intervention group of 3064 persons and a control group of 29 122 persons. From the National Cause of Death Register, data were collected as regards death and causes of deaths from 1970 to 1990.
Results: Multivariate analysis was used to correct for known confounders. We then found no differences between the groups regarding deaths from all causes, cardiovascular diseases, cancer or accidents and poisoning.
Conclusions: One general health screening seems to have little, if any value in preventing fatal diseases.