Table 1 Data extraction strategy
Data Extraction Strategy | Options | Notes |
|---|---|---|
Data Source Type | Electronic health records, linked health data study, cross-sectional survey, longitudinal cohort, randomised controlled trials | Â |
How was ethnicity asked and reported? If self-reported, how many categories? | Self-reported, prescribed, based on medical records, other | We separated ethnicity reported in medical records from self-reported ethnicity given the known inconsistencies of how ethnicity is asked in medical settings [21] |
How were missing, mixed, and other ethnic groups described? | Whether other, missing and mixed are reported or included in analysis | Â |
Was ethnicity aggregated in description or analysis? | Yes, No | Â |
If so, were justifications provided? | Yes (describe), No | Justification is defined very broadly A simple one sentence explanation on why ethnicity is aggregated is sufficient to be classed as having provided a justification. We do not evaluate the quality of the justifications |
How was ethnicity used in analysis? | Descriptor, predictor, other | Â |
Was ethnicity conceptualised or theorised to link with health outcomes? | Yes, No | Only applicable to papers using ethnicity as a predictor Conceptualisation is defined very broadly. A simple one sentence description on how ethnicity is theorised to associate with health outcome is sufficient to be classed as having provided a theory |