Coronary heart disease is the major cause of morbidity and mortality in the South Asian population in the United Kingdom, and its incidence is higher than in the white population.1 This excess risk seems to be determined by a combination of metabolic factors leading to the insulin resistance syndrome, psychosocial factors, and established risk factors.2 Ten out of 15 coronary risk factors measured were reported to be higher in South Asian patients than in their European counterparts, and several of these factors are believed to reflect relative deprivation. South Asian people are also at risk owing to high triglyceride concentrations and low concentrations of high density lipoprotein cholesterol. Although substantial evidence shows the value of lowering cholesterol in people at risk, studies have shown that many patients are not receiving appropriate treatment.3 We investigated the relation between ethnicity and prescribing of lipid lowering drugs.
Methods and results
We approached all general practices in one health authority to obtain consent to use their prescribing analyses and cost data for 1996-7. Sixty two (63.9%) of 97 practices gave consent. We obtained the following information for each practice from the health authority: proportion of South Asian patients in the nested age bands 35-69, 40-69, 45-69, 50-69, and 55-69, identified by using name based analysis software (Nam Pehchan)4; whether single handed or group practice; proportion of general practitioners of South Asian origin; fundholding status (particularly relevant at the time); Jarman index (surrogate measure for practice workload) for the practice's council ward; and Townsend score (measure of deprivation) for the ward. Comparative analyses of these demographic factors for each practice showed that consenting and non-consenting practices did not differ significantly (table).
We determined the number of defined daily doses of all lipid lowering drugs prescribed per 1000 South Asian patients in each nested age band for each consenting practice. We used multiple regression analysis (backward and forward selection techniques) to explore the relation between the number of defined daily doses prescribed per 1000 patients (aged 35 to 69) and the practice characteristics. Because of non-linearity and heteroscedasticity of the residual errors, we reanalysed the data after logarithmic transformation of the response variable. We identified two practices as extreme cases (as defined by SPSS) and excluded them from the analysis.
The median number of defined daily doses per 1000 patients was 4775 (interquartile range 2592 to 7336). Owing to strong correlation, we analysed Townsend score and Jarman index separately. The table shows the factors ranked in order of importance for predicting volume of prescribing, with Townsend score included. The parsimonious model includes only the percentage of South Asian patients and deprivation of the practice ward. The negative regression coefficients indicate reduction of prescribing levels with increasing numbers of South Asian patients and levels of deprivation. The results were not significantly changed by use of the various nested age bands or by replacement of Townsend score with Jarman index.
Comment
Patients in practices with a greater South Asian population are less likely to be prescribed lipid lowering drugs. This may be surprising, given the higher cardiovascular morbidity and mortality among South Asian people in the United Kingdom1 and a possible need for lipid lowering treatment that is equal to, if not greater than, that for the white population.2 Although this type of analysis does not show a causal link between ethnicity, deprivation, and the prescribing of lipid lowering drugs, the identified trend may demand explanation. Further analysis is needed to ascertain the effects of subsequent prescribing guidelines and recent government strategies promoting the use of lipid lowering drugs.5 Given the limitations of an ecological study, a standardised assessment is needed to determine the extent of unmet need and risk profiles at the level of the individual patient.
Table.
General practice characteristics and results of regression analysis
| Rank order | Factor | Consenting (n=62) |
Non-consenting (n=35) | P value | R2 (n=60) | Regressing log (DDD/1000) against all five factors (forward selection)*
|
|
|---|---|---|---|---|---|---|---|
| Coefficient (95% CI) | Standardised | ||||||
| 1 | % South Asian patients (median (interquartile range)) | 4.14 (0.99 to 35.70) | 5.39 (0.88 to 55.31) | 0.738† | 0.496 | −0.00490 (−0.00686 to −0.00294) |
−0.567 |
| 2 | Townsend score (median (interquartile range)) | 1.14 (−0.97 to 4.51) | 1.31 (−2.89 to 4.92) | 0.910† | 0.537 | −0.0183 (−0.03398 to −0.00262) |
−0.246 |
| 3 | Fundholding practices (No (%)) | 28 (45) | 12 (34) | 0.406‡ | 0.556 | No further factor identified as significant | |
| 4 | Single handed general practitioners (No (%)) | 16 (26) | 15 (43) | 0.133‡ | 0.559 | — | — |
| 5 | % South Asian general practitioners (mean (SD)) | 31.41 (43.71) | — | — | 0.560 | — | — |
DDD=defined daily dose.
Constant=3.778 (95% CI 3.717 to 3.839).
Mann-Whitney U test.
χ2 analysis (adjusted using Yates's correction).
Acknowledgments
We thank D Naylor, R J Naylor, A Hobbiss, and E Kernohan for providing invaluable support and guidance throughout and Bradford Health Authority and all participating general practices for permitting data collection.
Footnotes
Funding: None.
Competing interests: None declared.
References
- 1.Wild S, McKeigue P. Cross sectional analysis of mortality by country of birth in England and Wales, 1970-92. BMJ. 1997;314:705–710. doi: 10.1136/bmj.314.7082.705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bhopal R, Unwin N, White M, Yallop J, Walker L, Alberti KGMM, et al. Heterogeneity of coronary heart disease risk factors in Indian, Pakistani, Bangladeshi, and European origin populations: cross sectional study. BMJ. 1999;319:215–220. doi: 10.1136/bmj.319.7204.215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Primatesta P, Poulter NR. Lipid concentrations and the use of lipid lowering drugs: evidence from a national cross sectional survey. BMJ. 2000;321:1322–1325. doi: 10.1136/bmj.321.7272.1322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Cummins C, Winter H, Cheng KK, Maric R, Silcocks P, Varghese C. An assessment of the Nam Pehchan computer program for the identification of names of South Asian origin. J Public Health Med. 1999;21:401–406. doi: 10.1093/pubmed/21.4.401. [DOI] [PubMed] [Google Scholar]
- 5.Department of Health. National service framework for coronary heart disease: modern standards and service models. London: Stationery Office; 2000. [Google Scholar]