Self-rated health in public health evaluation

Sammanfattning: There is still a debate concerning the evidence base for community interventions. The randomised clinical trial design (RCT) is increasingly challenged as a gold standard for their evaluation. This thesis takes the Norsjö health programme in Västerbotten as the starting point for a discussion about the ethical platform of community interventions and for exploring the role of self-rated health. The specific objectives are: 1) to better understand barriers to community participation and to assess the role of ethical premises among decision-makers, 2) to explore how health related norms and attitudes interact with self-rated health and the risk factor outcome of an intervention and 3) to analyse the gender and socio-cultural interplay of self-rated health with biomedical risk factors for cardiovascular disease.The participation and views of different actors in the planning and implementation phases of the intervention were studied by contrasting information between official documents, interviews with decision makers and professionals and questionnaires to community members. The role of basic values in setting priorities and choosing intervention strategies utilised a questionnaire design with hypothetical scenarios sent to a representative sample of Swedish health care politicians. Qualitative research interviews were used to explore health related norms and attitudes. Health examination measurements and questionnaire data formed the basis for analysis of the development of self-rated health and risk factor load during a 10-year follow-up of the intervention. Access to a stroke registry enabled a case-referent approach for studying the interaction between bio-medical risk factors, socio-demographic factors and self-rated health. Data from the Västerbotten Intervention Programme (VIP) could be utilised for a cross-country comparison with a “sister project” in Otsego, U.S.A.The results point to both strengths and limitations of the efforts made to involve people in the intervention. The problem definition mainly remained with the professionals and participation as a goal in itself, strengthening local democracy was felt to be an exaggerated ambition. However, there was an overall agreement about the seriousness of the health problem, the need to intervene and about the implementation mode. Self-rated health and reported behavioural change were important indicators of participation and young men with bad health seemed to have been least involved. Among Swedish health care politicians there was an overall agreement to allocate resources for prevention directed towards communities when there are serious health problems. The majority preferred an intervention strategy that involved primary health care. The risk of harm by creating some degree of anxiety or stigma was for many considered an acceptable drawback of a successful intervention. The follow-up study revealed a positive risk factor reduction accompanied by a positive development of self-rated health, especially for men. Additional support for an intervention effect was given through a comparison with a reference area. The interaction pattern between risk reduction and self-rated health was more polarised for men than for women, with a corresponding pattern for the lower compared to the higher educated. These results could be linked to a transition in the health related norm system and to “ideal types” representing attitudinal sets towards the intervention. The case-referent analysis suggested an interaction effect between self-rated health and bio-medical risk factor load in predicting stroke that was greater for men than for women. The cross-country comparison revealed a stronger influence of education in the U.S.A. The lower educated, with a high risk load, had a greater risk of self-rated poor health than their Swedish counterparts.The thesis suggests that self-rated health is an unexplored indicator, potentially important for understanding the complexity of community interventions. Self-rated health may predict disease development as well as modify the impact of established risk factors.