Methodological issues in adolescent health surveys: the case of the Swiss Multicenter-adolescent Survey on Health.

In many countries, health care and prevention professionals face serious lack of data regarding health status, health care utilisation and lifestyles of adolescents. With reference to a survey on health and lifestyles of Swiss teenagers 15-20 years old, this paper reviews the different methodological issues linked with the conception and the realization of such a study. The main objective of the so-called "SMASH" project (Swiss Multicentric Adolescent Survey on Health) is, like other similar researches, to gather health indicators; that is, to measure attitudes and behaviour regarding different aspects of health and lifestyles, to identify the perceived health needs of the respondents and to describe their utilisation of health care services. The main issues which are addressed in the paper are related to: 1) the choice of the channel and the instrument to be used (telephone, vs. face-to-face interview vs. self-administered questionnaire; 2) the sampling procedures (pure random vs. cluster school-based samples; size of the sample); 3) the design and the content of the questionnaire (choice of the areas to be covered, selection and wording of the questions); 4) the ethical considerations linked with the collection of data. One of the specificities of SMASH is the inclusion of youth participation at various steps and levels: the design of the questionnaire, the strategies used to gather the data, and the analysis of the results. Two important issues emerge from this review. First, the importance in the future of being able to rely on common indicators that could be used in different countries and settings through the construction of a validated instrument. Second, the fact that most of the surveys conducted in this field neglect several sub-populations of adolescents with special and important needs: handicapped adolescents, drop-outs and recent immigrants coming either as unskilled workers or as refugees. Surveys in such special in-need groups will have to be planned with specific designs, both in terms of content and of approach.