Skip Navigation


Journal of Public Health Advance Access originally published online on October 27, 2006
Journal of Public Health 2007 29(1):40-47; doi:10.1093/pubmed/fdl068
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/1/40    most recent
fdl068v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ashworth, M.
Right arrow Articles by Rowlands, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ashworth, M.
Right arrow Articles by Rowlands, G.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006, Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved.

Social deprivation and statin prescribing: a cross-sectional analysis using data from the new UK general practitioner ‘Quality and Outcomes Framework’



M. Ashworth, Honorary Senior Lecturer1

D. Lloyd, Applied Statistician2

R. S. Smith, GIS Analyst3

A. Wagner, Research Fellow4

G. Rowlands, Professor of Primary Care and Public Health5
1 Department of General Practice and Primary Care, King’s College London School of Medicine at Guy’s, King’s College and St Thomas’ Hospitals, 5 Lambeth Walk, London SE11 6SP, UK
2 Prescribing Support Unit, The Information Centre for Health and Social Care, 1 Trevelyan Square, Boar Lane, Leeds LS1 6AE, UK
3 Informatics Collaboratory of the Social Sciences (ICOSS), University of Sheffield, 219 Portobello, Sheffield S1 4DP, UK
4 National Primary Care Research and Development Centre, Williamson Building, University of Manchester, Oxford Road, Manchester M13 9PL, UK
5 Faculty of Health, London South Bank University, 103 Borough Road, London SE1 0AA, UK

Address correspondence to Mark Ashworth, E-mail: mark.ashworth{at}kcl.ac.uk

We aimed to study the relationship between the prescribing of lipid-lowering medication, social deprivation and other general practice characteristics. We conducted a cross-sectional survey of all general practices in England, 2004–05. For each practice, the following variables were obtained: standardized cost and volume data for lipid-lowering medication, descriptors of general practices, Index of Multiple Deprivation, 2004, ethnicity data from the 2001 Census and Quality and Outcomes Framework data. A regression model was constructed which explained 34.5% of the variation in statin prescribing by general practitioners. The most powerful predictors were higher social deprivation, higher prevalence of coronary heart disease and achievement of cholesterol targets for diabetics. Negative regression coefficients were demonstrated for the proportion of elderly patients in the practice and, to a lesser extent, for the proportion of south Asian and Afro-Caribbean patients. In conclusion, contrary to previous local studies, we found that statin prescribing was higher in more deprived communities, even after adjustment for increased disease prevalence and practice variables associated with deprivation. Statin prescribing was also independently associated with success at achieving cholesterol targets in established disease (secondary prevention). However, our findings suggest under-prescribing of statins to the elderly and possibly also to ethnic minorities.

Keywords: primary care quality indicators, social deprivation, statin prescribing


   Introduction
TOP
 Introduction
Methods
 Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
Many aspects of the management of cardiovascular disease appear to follow the pattern of the ‘inverse care law’ in which there is a mismatch between need and the provision of services. The situation is complex because cardiovascular morbidity is greater in poorer and more socially deprived areas of the county,1 whereas investment in the services required to provide cardiovascular health care may bear no geographical relationship to areas of high cardiovascular disease prevalence.2 Access to revascularization services may, paradoxically, be increased in deprived areas, although this finding was probably the result of confounding by proximity to specialist cardiac centres.3

Inequalities in coronary artery bypass graft and coronary angioplasty rates have been demonstrated with low rates for these procedures in deprived communities3 and areas of high morbidity.4 In one example, council tenants with coronary heart disease (CHD) were found to be under half as likely to have had a coronary revascularization procedure.5 Not all the shortfall in delivery of services is likely to be attributable to a lack of provision. Population characteristics such as differences in health expectations or in health-seeking responses to cardiovascular illness may account for some of the apparent under-provision.6–8

Similar inequalities appear to exist in statin prescribing for patients with cardiovascular disease. Since 1994, good evidence has existed for the benefit of statins in reducing cardiovascular morbidity and mortality, both as primary and as secondary prevention.9 Nevertheless, surveys in selected areas of England have shown that the volume of statin prescribing by general practitioners (GPs) was lower in more deprived areas.10,11 In explanation, it was postulated either that fewer patients at high risk of cardiovascular disease presented to the GP in deprived areas or that patients in these areas made less demands for statin treatment.10

The new contract for GPs has provided a wealth of new descriptive and epidemiological data about the performance of general practice [Quality and Outcomes Framework (QOF) data].12 Data are collected from each participating general practice in England and collated at the end of the financial year—the first such data collection was at the end of March 2005. The resultant data set covers 146 quality indicators in the following domains: chronic disease management (76 indicators covering 11 chronic diseases), ‘practice organization’ (56 indicators), ‘patient experience’ (4 indicators) and ‘additional services’ (10 indicators). Each indicator is weighted, contributing to an overall maximum quality score for each practice of 1050 points. Up to one-third of GP income may be derived from success at achieving these indicators,13 so there is considerable financial motivation, and no doubt professional motivation too, to achieve highly.

Twelve of the QOF indicators are targets for the management of CHD, 10 for cerebrovascular disease (CVD) and 18 for diabetes. Some indicators describe processes (such as maintaining a CHD register), whereas others relate to outcomes (such as achieving target cholesterol and blood pressure levels).14

We aimed to study the relationship between social deprivation, the prescription of lipid-lowering medication and success in achieving the cholesterol targets specified in the new GP contract. We also wanted to describe the extent to which various practice variables influenced prescribing and may have acted as confounding variables, confusing any apparent relationship between deprivation and prescribing.


   Methods
TOP
 Introduction
 Methods
Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
QOF data
We obtained 2004–05 QOF data for all general practices in England from the The Information Centre for Health and Social Care, Leeds. Data for each clinical indicator covering CHD, CVD and diabetes mellitus (DM) were analysed separately. On the basis of the literature and our understanding of primary care, some organizational indicators were explored in more detail: whether the practice offered 10-min appointments,15 conducting medication reviews on patients receiving repeat prescriptions (a large proportion of repeat prescriptions are for the management of chronic cardiovascular disease or diabetes), whether the practice conducted significant event reviews (because these might suggest the presence of structured practice-based learning) and two indicators suggesting the implementation of cardiovascular risk factor screening: blood pressure monitoring in at least 75% of all registered patients aged ≥45 years and recording the smoking status of at least 75% of patients aged 15–75 years. Finally, raw prevalence data for the three chronic diseases listed above (calculated as disease register number divided by practice list size) were obtained from the QOF data set.

Practice characteristics
A detailed summary of practice characteristics was obtained from the Primary Care Research and Development Centre, University of Manchester. Variables included practice list size, age/sex breakdown of registered population, number of full-time equivalent GPs, training practice status, Personal Medical Services or General Medical Services status.16

Census-based variables
Data from the 2001 national Census were obtained and linked to practice data using the super output area (SOA) for each practice.17 SOAs are geographical, ‘socially homogeneous’ areas containing an average population of around 1500, which are arguably a better link to social measures than political units such as local authority wards. SOAs form the basis for calculating the Index of Multiple Deprivation (IMD), 2004.18 This relatively new deprivation index contains more wide-ranging domains than previously used indices. Deprivation is described according to seven domains: income, employment, health and disability, education skills and training, barriers to housing and services, crime and living environment. Although mostly derived from national Census data, and thus fixed to the year 2001, some variables such as education have been updated more recently. We used IMD data based on the SOA of each participating practice because place-of-residence data were not available for national data sets of registered patients. Deprivation data were therefore obtained at practice rather than at patient level.

Deprivation indices in common use do not include information about ethnic minorities, even though such data were elicited in the 2001 national Census. Because cardiovascular risk is unevenly distributed among ethnic groups, we obtained Census-derived figures, again based on SOAs linked to general practice postcodes, estimating the proportion of the local population from each ethnic group. We were particularly interested in the proportion from south Asian and Afro-Caribbean ethnic groups who are known to have an increased risk of cardiovascular disease.19 These data were provided by the Informatics Collaboratory of the Social Sciences (ICOSS), University of Sheffield, derived from variables using the Neighbourhood Statistics website (http://neighbourhood.statistics.gov.uk/).

Prescribing data
Prescribing data were obtained for two types of lipid-lowering drugs: statins and ezetimibe. Other lipid-lowering medication was not included because we were interested primarily in drugs that acted to reduce serum cholesterol levels rather than triglyceride levels. Data were collected over a 12-month period from April 2004 to March 2005 from the national Prescribing Analyses and CosT (PACT) data.20 Both cost and volume data were obtained and standardized according to the age/sex breakdown of the registered population in each practice, using specific therapeutic group age–sex weightings-related prescribing units (STAR-PUs).21 The STAR-PU used here is based on the use of lipid-regulating medication. The volume of prescriptions issued was measured by using the average daily quantity (ADQ)22—a standardized quantity that is preferable to using the number of items as it avoids the apparently high prescribing volumes of practices that issue multiple prescriptions for short periods. The cost was measured using the net ingredient cost (NIC). These data were provided by the Prescription Pricing Authority in Newcastle.

Statistical methods
We constructed a data set containing data from all 8576 practices in England, their QOF data, practice and Census-based variables and prescribing data. Sixty-one practices were removed from the analysis because they were no longer independent at the end of the study year or had a list size of under 750 patients or under 500 per full-time GP on the grounds that these were likely to be newly formed or about to be closed. The final QOF data set contained information on 8515 practices. Postcode and SOA code anomalies meant that IMD data could only be matched to 8480 practices. Disease prevalence data were available for 8430 of these practices.

Firstly, we explored the univariate associations between the cost and the volume of the two lipid-lowering drugs and other practice and Census-based variables, using linear regression. All univariate associations were then included in a multivariate analysis using multiple regression and a forward stepwise method.


   Results
TOP
 Introduction
 Methods
 Results
Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
Table 1 summarizes the distribution of all the variables included in our study. Statins are more widely prescribed than ezetimibe by a factor of ~100. For this reason, detailed results of our analyses are shown only for statins.


View this table:
[in this window]
[in a new window]

  		Table 1 Summary of the study variables

 

The standardized volume of statin prescribing varied widely between practices: the mean volume was 4220 ADQs per 1000 STAR-PUs, standard deviation 1595. Ten per cent of practices prescribed under 2629 ADQs per STAR-PU, whereas 10% prescribed above 6207.

Social deprivation scores (IMD score) correlated significantly (P < 0.001) with the volume of statin prescribing (Pearson’s r = 0.29), CHD prevalence (r = 0.10), CVD prevalence (r = –0.08) and DM prevalence (r = 0.28). The IMD score was also correlated with the proportion of the registered population aged ≥75 years (r = –0.25).

The prevalence of the three chronic diseases included in this study was closely correlated. The highest correlation coefficient was between CHD and CVD (r = 0.69). There were lesser correlation coefficients between DM and CHD (r = 0.37) and between DM and CVD (r = 0.17).

The achievement of cholesterol control in the three chronic diseases also correlated closely: for CHD and CVD (r = 0.72), for CHD and DM (r = 0.74) and for DM and CVD (r = 0.63).

Univariate associations based on standardized values for the volume of statin prescribing are summarized in Table 2. On the basis of this approach to analysis, the main factors associated with the volume of statin prescribing are social deprivation, the prevalence of the three chronic diseases included in our study (CHD, CVD and DM) and the achievement of target cholesterol levels for these diseases.


View this table:
[in this window]
[in a new window]

  		Table 2 Univariate associations between statin-prescribing volume and predictor variables

 

A parsimonious regression model was constructed. Nine variables explained 34.5% of the variation in statin prescribing volume (Table 3). The adjusted regression coefficients summarized in Table 3 demonstrate the effect of each variable when adjusted for confounding by the other variables. Thus, the adjusted regression coefficient for deprivation is 21.2 (Table 3), whereas for deprivation alone, unadjusted for other confounders such as the higher prevalence of cardiovascular disease in deprived areas, the regression coefficient is higher at 27.4 (Table 2). To compare the relative importance of each of the nine variables in the final regression model, we required standardized regression coefficients, because each variable is measured in different units. On the basis of these standardized values (Table 3), the most powerful predictors remain social deprivation, the prevalence of disease, achievement of cholesterol targets and the proportion of patients aged ≥75 years. Variables such as longer GP consultations, training practice status, practice-based learning (as evidenced by significant event analyses) and active screening for cardiovascular risk factors had no significant effect on the model.


View this table:
[in this window]
[in a new window]

  		Table 3 Multivariate associations between statin-prescribing volume and nine predictor variables

 

A similar analysis was conducted for predictors of the cost of statin prescribing. The standardized cost was closely associated with the standardized volume (Pearson’s r = 0.83, P < 0.001). Nine variables explained 33.3% of the variation in the cost of statin prescribing (regression coefficients available from the authors). Seven of these variables were identical to those predicting the volume of statin prescribing (Table 3), but two were excluded (the proportion of patients from a south Asian ethnic group and cholesterol ≤5.0 in patients with CVD) and there were two additions (unadjusted prevalence of CVD and single-handed practices).

Finally, a similar analysis was conducted for predictors of the volume of ezetimibe prescribing. A model containing all the variables listed in Table 2 only explained 4.1% of the variation in ezetimibe prescribing. The IMD score was not significant as an independent predictor. The highest standardized adjusted regression coefficient was for the variable ‘cholesterol ≤5.0 in all DM patients’, standardized Beta 0.13, P < 0.001.


   Discussion
TOP
 Introduction
 Methods
 Results
 Discussion
Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
Main findings of this study
Higher volumes of statins were prescribed in general practices serving more deprived populations. This effect remained even after adjustment for other factors such as the increased prevalence of cardiovascular disease and diabetes in deprived areas.

The volume of statin prescribing also appeared to be based on the morbidity needs of the practice population, being higher in practices reporting higher prevalences of cardiovascular disease or diabetes.

Some practice factors influenced statin prescribing: single-handers and small practices prescribed more statins although the effect was not large on the regression model. However, some organizational factors, which might have been thought to influence statin prescribing, were found to have little independent effect: longer GP consultations, training practice status, practice-based learning (as evidenced by significant event analyses) and active screening for other cardiovascular risk factors.

The independent association of statin prescribing with reported cardiovascular and diabetes morbidity suggests that a substantial proportion of statin prescribing is linked to secondary prevention.

Practices with a higher proportion of patients aged >75 years prescribed proportionally less statins. This association remained relatively strong even when other factors, such as social deprivation and the higher reported prevalence of cardiovascular disease and diabetes, were controlled and may suggest that patients aged >75 years are relatively disadvantaged in terms of statin treatment.

Practices based in areas with higher proportions of patients of Afro-Caribbean or south Asian ethnicity had lower volumes of statin prescribing. This association was independent of other variables including social deprivation and, although not strong, might imply that ethnic minorities are relatively disadvantaged in terms of statin treatment. However, the results must be interpreted with caution because the proportion of ethnic minorities is highly negatively skewed, and the relatively few areas with high proportions of ethnic minorities would have had a disproportionate effect on the regression model.

The cost of statin prescribing has been a concern,23 but our findings show that both higher volumes and higher costs did translate into better control of cholesterol in all three chronic diseases included in the QOF and that the higher the reported prevalence of each disease, the higher the cost of statins. Again, statin prescribing was more costly in deprived areas.

What is already known on this topic?
Other studies report either lower statin prescribing in more deprived communities10 or the lack of an association with deprivation or social class.24–26 However, previous studies10,11 only included relatively small numbers of practices (118 and 132, respectively) and were unable to derive practice-based disease prevalence rates or measures of cholesterol control (both variables only becoming widely available in the QOF data set).

Our findings are congruent with those of several studies demonstrating relatively low levels of statin prescribing in the elderly. On the basis of an analysis of data from the British Regional Heart Survey, Ramsay et al.27 found that patients with established CHD (74–85 years) were 60% less likely to be taking a statin than younger patients. Data from the Health Survey for England demonstrated that the likelihood of receiving statin therapy for CHD was greatly reduced in patients aged >75 years (odds ratio 0.11).25 Our findings are of interest in the light of increasing evidence of benefit from statin therapy in older age groups.28

The relationship between ethnic group and statin prescribing has been explored in other studies. Findings comparable with our own were reported from a study of primary care prescribing in Bradford.29 The prescribing of lipid-lowering drugs was found to be lower in practices with a high south Asian population, although their study identified south Asians on the basis of name-based analysis software rather than on self-report.

What this study adds
On the basis of a national study, we have presented the first report of higher levels of statin prescribing in general practices serving more deprived communities.

Limitations of this study
Several limitations may affect the interpretation of our findings. Firstly, prevalence data available in the QOF database are not age/sex standardized. Moreover, GPs may not have organized their disease registers sufficiently during the first year of QOF to have included all known patients with the three chronic diseases included in our study. The 2003 Health Survey for England reported a national crude prevalence for CHD in adults of 7.4% in men and 4.5% in women.30 Data obtained from a database of >2 million case records in primary care included all age groups and found a crude CHD prevalence of 4.0%.31 When compared with the QOF-derived CHD prevalence figure of 3.54%, both these surveys imply incomplete case finding in QOF. Findings about the volume of statin prescribing therefore need to be interpreted in the light of disease prevalence figures that only report recognized disease rather than the true level of need in the community as determined by community surveys incorporating case finding.

The final regression model omitted the reported prevalence of CVD as a predictor variable. Although high reported prevalence of CVD did predict high statin prescribing, the reported prevalence of CVD and CHD was so closely correlated that its inclusion in the regression model added little to the predictive power of the model. A similar argument applies to the omission of cholesterol control in CHD as a predictor variable of statin prescribing.

The lack of data about individual patient locations led to general practice postcodes being used as a proxy. The interpretation of findings at an individual level, when data have been obtained at an area level, may give rise to the ecological fallacy.32 In small-scale studies, there are likely to be significant discrepancies between social variables, such as ethnicity or income, obtained from an SOA, and the same variables attributed to the registered list of a general practice based in that particular SOA. However, these discrepancies may remain even in large studies, meaning that findings have to be interpreted with caution. A recent study of 38 general practices in Rotherham provided some justification for using practice-level deprivation data, concluding that practice postcode-linked IMD scores did provide a valid proxy for patient-level deprivation measures.33 However, Strong et al.33 also found that practice-level deprivation data tended to underestimate the strength of the association between deprivation and all-cause mortality.

Exception reporting has not been included in our study. Exception reporting is the process whereby GPs exclude certain patients from the targets set by QOF on the grounds of unsuitability, for example, if the patient refuses treatment with statins, has unacceptable side effects or refuses to attend review appointments.34 The median rate or exception reporting in 2004–05 was 6%.35 True achievement of cholesterol targets may therefore be somewhat lower for the whole population of patients with cardiovascular disease than the declared results. However, unless there are differential exception reporting rates in deprived areas, this process would not affect our finding of the relationship between statin prescribing and deprivation.

Finally, our survey was not able to obtain consultation data. Without consultation data, we do not know whether the excess of statins in deprived communities was prescribed for primary or secondary cardiovascular prevention. Nor do we know whether the threshold for diagnosis was similar in practices in deprived and non-deprived areas. If, for example, GPs working in more deprived communities only recognized more severe cases of CHD, then the finding of increased statin prescribing in deprived areas could still be consistent with poorer treatment for deprived patients with similar disease severity. Similarly, we have no information about whether patients were purchasing over-the-counter (OTC) statin medication, although the volume of OTC statin sales is tiny compared with the volume of GP prescriptions.36 It is likely that more OTC statins were obtained in more affluent areas, and such a pattern might have contributed to our finding of higher GP statin-prescribing volumes in more deprived areas.


   Conclusions
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
General practices located in more deprived communities with a higher reported prevalence of cardiovascular disease and diabetes and reporting more success at achieving cholesterol targets were found to prescribe higher volumes of statins. Inequities in statin prescribing appear to apply to the elderly population and possibly also to south Asians and Afro-Caribbeans. Although we could not identify the indication for prescribing, the finding of substantially higher statin prescribing in deprived communities requires further research.


   Contributor
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
 Contributor
Conflict of interest
 Ethical approval
 Acknowledgements
 References
 
M.A. is the guarantor.


   Conflict of interest
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
Ethical approval
 Acknowledgements
 References
 
None declared.


   Ethical approval
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
Acknowledgements
 References
 
Guy’s Research Ethics Committee (Chairman’s action, 8 February 2006).


   Acknowledgements
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
References
 
We acknowledge the roles of Stevo Durbaba (Department of General Practice and Primary Care, King’s College London) in preparing the database.

Funding

M.A. has been part funded by the South Thames Research Network (STaRNet), London.


   References
TOP
 Introduction
 Methods
 Results
 Discussion
 Conclusions
 Contributor
 Conflict of interest
 Ethical approval
 Acknowledgements
 References
 

  1. Harding S, Bethune A, Maxwell R et al. Mortality trends using the longitudinal study. In: Drever F, Whitehead M (eds). Health, Inequalities Series DS No 15. Government Statistical Service. London: ONS, 1997, p. 143.
  2. Hippisley-Cox J, Pringle M. Inequalities in access to coronary angiography and revascularisation: the association of deprivation and location of primary care services. Br J Gen Pract 2000;50:449–54.[Web of Science][Medline]
  3. Black N, Langham S, Petticrew M. Coronary revascularisation: why do rates vary geographically in the UK? J Epidemiol Community Health 1995;49:408–12.[Abstract/Free Full Text]
  4. Langham S, Basnett I, McCartney P et al. Addressing the inverse care law in cardiac services. J Public Health Med 2003;25:202–7.[Abstract/Free Full Text]
  5. Morris RW, Whincup PH, Papacosta O et al. Inequalities in coronary revascularisation during the 1990s: evidence from the British regional heart study. Heart 2005;91:635–40.[Abstract/Free Full Text]
  6. Richards HM, Reid ME, Watt GC. Socioeconomic variations in responses to chest pain: qualitative study. BMJ 2002;324:1308.[Abstract/Free Full Text]
  7. Gardner K, Chapple A. Barriers to referral in patients with angina: qualitative study. BMJ 1999;319:418–21.[Abstract/Free Full Text]
  8. Richards H, McConnachie A, Morrison C et al. Social and gender variation in the prevalence, presentation and general practitioner provisional diagnosis of chest pain. J Epidemiol Community Health 2000;54:714–8.[Abstract/Free Full Text]
  9. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease. The Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383–9.[CrossRef][Web of Science][Medline]
  10. Packham C, Robinson J, Morris J et al. Statin prescribing in Nottingham general practices: a cross-sectional study. J Public Health Med 1999;21:60–4.[Abstract/Free Full Text]
  11. Ward PR, Noyce PR, St Leger AS. Are GP practice prescribing rates for coronary heart disease drugs equitable? A cross sectional analysis in four primary care trusts in England. J Epidemiol Community Health 2004;58:89–96.[Abstract/Free Full Text]
  12. U.K. Department of Health. Investing in General Practice: The New GMS Contract. London: DoH, 2003.
  13. http://www.primarycarecontracting.nhs.uk/qanda.php?article_request=64 (1 June 2006, date last accessed).
  14. http://www.dh.gov.uk/assetRoot/04/08/86/93/04088693.pdf (1 June 2006, date last accessed).
  15. Howie JG, Porter AM, Heaney DJ et al. Long to short consultation ratio: a proxy measure of quality of care for general practice. Br J Gen Pract 1991;41:48–54.[Web of Science][Medline]
  16. A Guide to Personal Medical Services Pilots Under the NHS (Primary Care Act) 1997. London: Department of Health, 1997. http://www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidanceArticle/fs/en?CONTENT_ID=4009667&chk=ARdoqC (1 June 2006, date last accessed).
  17. Beginners’ guide to UK geography. Super Output Areas (SOAs). http://www.statistics.gov.uk/geography/soa.asp (1 June 2006, date last accessed).
  18. Indices of deprivation 2004 – summary (revised). Office of the Deputy Prime Minister. http://www.odpm.gov.uk/index.asp?id=1128442 (1 June 2006, date last accessed).
  19. Cappuccio F. Ethnicity and cardiovascular risk: variations in people of African ancestry and South Asian origin. J Hum Hypertens 1997;11:571–6.[CrossRef][Web of Science][Medline]
  20. Majeed A, Evans N, Head P. What can PACT tell us about prescribing in general practice? BMJ 1997;315:1515–9.[Free Full Text]
  21. Lloyd DC, Harris CM, Roberts DJ. Specific therapeutic group age–sex weightings related prescribing units (STAR-PUs): weightings for analysing general practices’ prescribing in England. BMJ 1995;311:991–9.[Abstract/Free Full Text]
  22. Walley T, Roberts DJ. Average daily quantities – a tool for measuring prescribing volume in England. Pharmacoepidemiol Drug Saf 2000;9:55–8.
  23. Cook J, Yin D, Alemao E et al. Development and validation of a model to project the long-term benefit and cost of alternative lipid-lowering strategies in patients with hypercholesterolaemia. Pharmacoeconomics 2004;22(S3):37–48.[CrossRef]
  24. Strong M, Maheswaran R, Radford J. Socioeconomic deprivation, coronary heart disease prevalence and quality of care: a practice-level analysis in Rotherham using data from the new UK general practitioner Quality and Outcomes Framework. J Public Health 2006;28:39–42.
  25. Reid FD, Cook DG, Whincup PH. Use of statins in the secondary prevention of coronary heart disease: is treatment equitable? Heart 2002;88:15–9.[Abstract/Free Full Text]
  26. Britton A, Shipley M, Marmot M et al. Does access to cardiac investigation and treatment contribute to social and ethnic differences in coronary heart disease? Whitehall II prospective cohort study. BMJ 2004;329:318–23.[Abstract/Free Full Text]
  27. Ramsay SE, Morris RW, Papacosta O et al. Secondary prevention of coronary heart disease in older British men: extent of inequalities before and after implementation of the National Service Framework. J Public Health 2005;27:338–43.
  28. Ebrahim S, Davey Smith G, McCabe C et al. What role for statins? A review and economic model. Health Technol Assess 1999;3:12.[Medline]
  29. Patel MG, Wright DJ, Gill PS et al. Prescribing of lipid lowering drugs to South Asian patients: ecological study. BMJ 2002;325:25–6.[Free Full Text]
  30. Sproston K, Primatesta P (eds). Health Survey for England 2003, Vol. 1. Cardiovascular Disease. London: The Stationery Office, 2003.
  31. Horsfield P, Teasdale S. Generating information from electronic patient records in general practice: a description of clinical care and gender inequalities in coronary heart disease using data from over two million patient records. Inform Prim Care 2003;11:137–44.[Medline]
  32. Longley PA, Goodchild G, Maguire DJ et al. Geographic Information Systems and Science. Chichester: John Wiley and Sons Ltd, 2001.
  33. Strong M, Maheswaran R, Pearson T. A comparison of methods for calculating general practice level socioeconomic deprivation. BMC Int J Health Geographics 2006;5:29–34.[CrossRef]
  34. Roland M. Linking physicians’ pay to the quality of care – a major experiment in the United Kingdom. N Engl J Med 2004;351: 1448–54.[Free Full Text]
  35. Doran T, Fullwood C, Gravelle H et al. Pay-for-performance programs in family practices in the United Kingdom. N Engl J Med 2006;355:375–84.[Abstract/Free Full Text]
  36. OTC statins: a bad decision for public health. Lancet 2004;363:1659.[CrossRef][Web of Science][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/1/40    most recent
fdl068v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ashworth, M.
Right arrow Articles by Rowlands, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ashworth, M.
Right arrow Articles by Rowlands, G.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?