AdDIT UK | AdDIT Canada | AdDIT Australia | |
---|---|---|---|
# of sites | 20 sites | 5 sites | 7 sites |
Regions/provinces/states | 8 | 1 | 4 |
Regional representation pattern | Not representative | Partially representative | Fully representative |
Over-recruitment of high SES and under-recruitment of low SES | Under-recruitment of low SES | Recruitment consistent with regional population | |
Recruitment approach | Larger number of diabetes centres from wider geographic area with fewer participants recruited per clinic | Diabetes centres from a single urban area | Diabetes centres from 4 large urban areas |
Interpretation | Potential for selection bias in participants with greater interest, time or perceived suitability (Higher SES) | Selection bias in background urban population with higher proportions of high and low SES with a “U shaped” distribution consistent with urban income polarization | Skew in regional population consistent with census data showing that most advantaged population quintiles are clustered around Australian cities and selected coastal areas |
National representation | Not representative | Not representative | Not representative |