[Response rate and analysis of non-responses in a cohort study].

In 1987, a national survey on psychotropic substances and life styles was carried out by the Swiss Institute for Alcohol and Drug Problems (SIPAD). Out of 1910 participants, 1568 did agree to take part in a further study. In 1995, in order to execute a cohort type survey with minimal cost, the co-ordinates of 1472 of them were retrieved and a new questionnaire was sent to them. After 2 written reminders and another one by phone, the response rate reached 72.6%. It is therefore possible, at least in Switzerland, to lead a cohort type survey using a cross-sectional study as the starting point. Significant differences arise between the respondents and non-respondents at the level of linguistic region, age, income, civil status, educational level and alcohol or drugs consumption. On the other hand, neither sex, nor tobacco or cannabis consumption had an influence on the response rate.