Marcel Slockers

116 Chapter 8 and unhealthy lifestyle and living conditions before the introduction of the Action Plan for Social Support. This may have led to irreversible health damage (for example atheroscle- rosis) which cannot be influenced sufficiently by social policy. After the introduction of the Action Plan for Social Support, we did not find a decrease in total mortality, but we did find a shift in the pattern of causes of death. Just as in international studies, we observed, after implementation of social policy, a smaller share of unnatural causes of death and a larger share of cardiovascular diseases in mortality within our cohort. It may be that social policies only have effects on mortality from acute unnatural causes of death and are then replaced by mortality from chronic diseases, on which these policies have little or no effect. As part of the Major Cities Policy (an official Dutch policy to improve living and working conditions in themajor cities),Amsterdam, Rotterdam, The Hague andUtrecht implemented the same Action Plan for Social Support. As a result, the social situation and safety of homeless people improved, but only a relatively small part of this group found a stable mix of housing, income, and contact with care workers. 3,4 It therefore seems plausible that the effects of policy were also diluted in the other major cities and that the Action Plan for Social Support has not, in any of the cities, led to a decrease in total mortality. Our research on the influence of health insurance policy changes on the accessibility of care for homeless people (see chapters 6 and 7) has a limitation as it was conducted with small convenience samples, of which the response rate was rather low. This research has nevertheless identified several factors that may impede homeless people's access to care. Because of the limitations mentioned above, we do not present a complete and represen- tative picture of barriers and facilitators , but some first indications of possible causes of limited accessibility of care have been obtained. These first indications are relevant and form the basis for policy recommendations and a flow chart for care providers. 3. Results compared with literature How do our observed results on mortality and health problems among the homeless population in Rotterdam compare with the existing literature? Studies indicating higher mortality and lower life expectancy Mortality among homeless people in Rotterdam is 3.5 times higher than among other inhabitants of Rotterdam. This result is in line with international reviews which show high excess mortality. 5,6 Several Western countries in and outside Europe have previously shown that homeless people everywhere die much earlier than the general population, despite differences in healthcare systems, culture, and populations. 7,8,9,10 However, our reported