The Delphi method – determining the importance of individual areas
To determine the importance of each area, an expert evaluation was carried out using the Delphi method. A total of 10 independent experts from different disciplines related to population health anonymously determined the weight of each of the nine areas. The experts’ estimates were refined in three rounds, during which feedback and opinions of other experts were anonymously provided to all of them. After each round of assessment, the range of weights was determined and averaged for each of the nine areas separately. After that, the questionnaire was sent out again at an interval of approximately two months to refine the weights (the questionnaire contained an overview of the areas and their indicators). The Covid-19 pandemic has led to a partial reassessment of the importance of some areas.
- Economic conditions and social protection (weight 15)
- Education (weight 14)
- Demographic changes (weight 6)
- Environmental conditions (weight 11)
- Individual living conditions (weight 7)
- Road safety and crime (weight 3)
- Health and social care resources (weighting 8)
- Health (weight 16)
- Quality of life (weight 20)
Individual clusters (clusters, groups) are created using cluster analysis. Cluster analysis is a multivariate statistical method used to classify objects. It is used to classify units, in this case districts, into groups (clusters) so that units belonging to the same group are more similar than objects from other groups. The clusters are characterized by the same set of features, i.e. 103 indicators of all nine areas.
The software allows the creation of indices (composite indicators) of individual areas 1 to 8, as well as a summary Health Index. The processing is set up using a multi-criteria variance scoring method: the WSA method.
The software is developed using MySQL and ESRI AGOL technologies. User data is uploaded to the database via a web page. The uploaded data is checked for completeness and relevance. In case of an error, the deficiencies are reported to the user with the possibility of correction. After checking, the software visualizes the entered data into maps for the overall Health Index and individual areas. The numerical database data are assigned a location at the district level and the data are transformed into geographic polygon layers in the S-JTSK coordinate system. This preprocessed data is uploaded to the ESRI AGOL cloud environment. Here they are used in a web map and then in the resulting web application, which includes analytical widgets in addition to the visualization. Using the data embedded in the generated software, the user is presented with specific graphical and map documents. It is also possible to insert data for historical predefined periods and the software visualizes the districts graphically and cartographically, which is useful for comparing changes in given regions of different periods. In this way, data from 1980 to 2000 can be displayed at five-year intervals. After 2000 in three-year intervals. Data can be visualized up to 2040.