Record linked retrospective cohort study of 4.6 million people exploring ethnic variations in disease: myocardial infarction in South Asians
1 Information Services Division, NHS National Services Scotland, Gyle Square, 1 South Gyle Crescent, Edinburgh EH12 9EB, UK
2 General Register Office for Scotland, Ladywell House, Ladywell Road, Edinburgh EH12 7TF, UK
3 Public Health Sciences, University of Edinburgh, Teviot Place, Edinburgh EH8 9AG, UK
BMC Public Health 2007, 7:142 doi:10.1186/1471-2458-7-142Published: 5 July 2007
Law and policy in several countries require health services to demonstrate that they are promoting racial/ethnic equality. However, suitable and accurate data are usually not available. We demonstrated, using acute myocardial infarction, that linkage techniques can be ethical and potentially useful for this purpose.
The linkage was based on probability matching. Encryption of a unique national health identifier (the Community Health Index (CHI)) ensured that information about health status and census-based ethnicity could not be ascribed to an identified individual. We linked information on individual ethnic group from the 2001 Census to Scottish hospital discharge and mortality data.
Overall, 94% of the 4.9 million census records were matched to a CHI record with an estimated false positive rate of less than 0.1 %, with 84.9 – 87.6% of South Asians being successfully linked. Between April 2001 and December 2003 there were 126 first episodes of acute myocardial infarction (AMI) among South Asians and 30,978 among non-South Asians. The incidence rate ratio was 1.45 (95% CI 1.17, 1.78) for South Asian compared to non-South Asian men and 1.80 (95% CI 1.31, 2.48) for South Asian women. After adjustment for age, sex and any previous admission for diabetes the hazard ratio for death following AMI was 0.59 (95% CI 0.43, 0.81), reflecting better survival among South Asians.
The technique met ethical, professional and legal concerns about the linkage of census and health data and is transferable internationally wherever the census (or population register) contains ethnic group or race data. The outcome is a retrospective cohort study. Our results point to increased incidence rather than increased case fatality in explaining high CHD mortality rate. The findings open up new methods for researchers and health planners.