SALUD - SpatiAL microsimUlation for Decision support

Government policies have a significant impact on certain population groups. For instance, when looking at the issue of smoking, an increase in cigarette prices will probably affect different population groups compared to introducing a smoking ban in public areas. Therefore, it is important to know what types of people are affected most by policy changes and this can be best analyzed at microlevel, i.e. it is necessary to have microdata, which provide information at the individual or household level. Such data are often not available, e.g. due to confidentiality reasons, and therefore, we have to estimate so called "missing data". This estimation can be done with microsimulation, which is a method to build up large-scale datasets based on attributes of individuals or households by linking different datasets. To date, microsimulation has been applied principally to understand the impacts of new economic tax/income policies. The challenge in this project is to incorporate the spatial component (e.g. geographical data) by building a spatial microsimulation model that best addresses policy issues at the local level. For this research, the potential of spatial microsimulation will be investigated by examining the impacts health policy scenarios have, not only on different population groups but also on the locations where these groups live. In Austria, spatial microsimulation models for health could make a major contribution to policy analysis as these models are not widely established and distributed yet. One powerful advantage of spatial microsimulation is the possibility to model various what-if scenarios, which allows policy makers to explore the impact that certain policies would have on certain population groups. Application areas of what-if scenarios can address topics that are high on the Austrian policy agenda, including smoking, obesity, nutrition, fertility or cancer related diseases, where missing data for small areas can be estimated and spatial variations can be found that are otherwise hidden when looking at large scale surveys only. The overall aim of this project is to build a spatial microsimulation model focusing on relevant health issues in Austria by linking different data sources to provide missing data. This will raise technical and empirical issues: for instance, how many variables can or should be incorporated in the model to produce reliable results? In the end, the results will be validated and the process optimized when necessary and presented to health policy makers to support future regional planning and in turn reduce health inequalities. Cooperation with two international partners in the UK (who are experts in this research field) has been agreed. This will allow the adoption of the simulation model between two different countries to address problems related to spatial scales and population compositions and this was not done before in this context. Finally, a foundation can be laid and the findings can be used to serve as interface between research and government or companies.

Regional analyses are important for future planning purposes. In Austria, and many other countries, no health related data are available for small areas such as municipalities. Such data are essential for effective regional planning purposes. Therefore, the SALUD project used spatial microsimulation modelling to synthesize data for small-area municipalities in Austria. SALUD focused on simulating the prevalence of smoking, heavy smokers, diabetes, high blood pressure, and obesity at municipality level to identify hot spot areas. The results showed spatial variations at municipality level, where smoking prevalence was higher in the west of Austria, and diabetes, high blood pressure, and obesity rates were higher in the east. One main advantage of spatial microsimulation modelling is the possibility to simulate specific target groups. For example, it is possible to simulate trends for men in the 30-39 age group who have lower educational attainment at municipality level, as this cohort is known to have higher smoking rates. Having such detailed information is useful for planning the location of health resources, but also to help target hard to reach groups more effectively. In addition, it is possible to link this information with other spatial data and models to identify gaps in service provision, such as location, opening hours, and accessibility to service provision with public or private transport.Spatial microsimulation models use algorithms to link different datasets, such as the Austrian Health Interview Survey and the Registered-based Census. Currently, no standard software exists to produce such simulations, although some packages are publicly available. However, these packages normally require expert knowledge. Therefore, within the SALUD project an online application was developed - simSALUD - that allows users to conduct spatial simulation step by step, to validate the results immediately after the simulation execution has finished, and to visualise the results in form of maps. The latter is an especially valuable tool to help communicate results to policy makers.In Austria, spatial microsimulation has not been applied in the area of health research, and this project could make a major national and international contribution in this research area. Further, the SALUD project can help to promote the advantages of health and geoinformation (health-GIS) research to public health institutions.