Method Comparison: ATHIS_Jemals, ATHIS_Behandlung, ATC-ICD, ExpertInnen
Select an age category and click through the methods. Click on the provinces and see what happens.
Note that in Vorarlberg in the age group over 85, many individuals with diabetes were identified by the methods ATC-ICD and ExpertInnen, but none by ATHIS_Jemals and ATHIS_Behandlung. This could be due to the small sample size in the respective age and gender group in the Health Interview Survey (ATHIS) for this province. Under such close examination large differences can occur due to large variances in small sample sizes. When analysed, for instance, only by age group as in representation only by age group (see comparison 3) the sample size is larger, resulting in a smaller variance.
Observe the different values per province. It is striking that the values obtained by ATHIS_Jemals for younger age groups in Salzburg are always 0, while ATC-ICD and ExpertInnen show a low prevalence rate unequal to 0. This outcome could be attributed to the small sample size in the respective age group of females in Salzburg in the Health Interview Survey (ATHIS). In ATHIS, when analysing at a high level of detail, large differences in outcome can result when the variance is too large for extrapolation. When evaluating at a lower level of detail, as for instance solely by gender (see Comparison 2), the sample size is larger and the variance is smaller.
Also of interest are the prevalence figures in the age group 65–74 years in Lower Austria in ATC-ICD and ExpertInnen. They are higher than in Carinthia. In ATHIS_Behandlung it is exactly the opposite.