Article Text

Download PDFPDF

Recent UK trends in the unequal burden of coronary heart disease
  1. Jonathan Pearson-Stuttard1,
  2. Madhavi Bajekal2,
  3. Shaun Scholes2,
  4. Martin O'Flaherty3,
  5. Nathaniel Mark Hawkins4,
  6. Rosalind Raine2,
  7. Simon Capewell3
  1. 1Christ Church, University of Oxford, Oxford, UK
  2. 2Department of Applied Health Research, University College London, London, UK
  3. 3Division of Public Health & Policy, University of Liverpool, Liverpool, UK
  4. 4Institute of Cardiology Medicine & Science, Liverpool Heart and Chest Hospital, Liverpool, UK
  1. Correspondence to Jonathan Pearson-Stuttard, Medical Student, Christ Church, St Aldate's, Oxford, OX1 1DP, UK; jonathan.pearson-stuttard{at}chch.ox.ac.uk

Abstract

Introduction The burden of coronary heart disease (CHD) in the UK is substantial. However, recent trends and associated socioeconomic inequalities are not well studied. We aim to identify and analyse these trends stratified by age, gender and socioeconomic quintiles.

Methods We quantified the CHD burden and analysed trends from 1999 to 2007 in all adults aged over 25 years resident in England. Data sources included deaths (from ONS), health surveys, and hospital admissions (from Hospital Episode Statistics), all using ICD9 and ICD10 coding. Socioeconomic inequalities were calculated in both absolute and relative terms.

Results In 2007, the CHD burden comprised approximately 205 000 hospital admissions (acute and elective), including approximately 110 000 admissions with acute coronary syndrome. There were approximately 1.5 million CHD patients with chronic disease living in the community. Approximately 67 500 of these were admitted during 2007 for revascularisation. There were approximately 173 000 CHD patients living with heart failure, of whom some 14% required hospital admission during 2007. Between 1999 and 2007, age-specific hospital admission rates generally decreased by 20%–35%. Community prevalence decreased by 10%–20%. Strong socioeconomic gradients were apparent in all patient groups, persisting or worsening between 1999 and 2007.

Conclusions The burden of CHD is immense, costly and unequal. Hospital admissions attract more attention than the far more numerous patients living with chronic disease in the community. Population-based rates for hospital admissions and CHD prevalence have been declining by 3%–4% per annum. However, marked socioeconomic gradients have persisted or worsened—there is no room for complacency.

  • Socioeconomic inequalities
  • coronary
  • heart
  • disease burden
  • trends
  • quality of care and outcomes
  • epidemiology
  • heart failure
  • cardiac resynchronisation therapy
  • systolic heart failure
  • mortality statistics
View Full Text

Statistics from Altmetric.com

Introduction

The burden of cardiovascular disease in the UK is very large. Coronary heart disease (CHD) and stroke, together cause over 150 000 cardiovascular disease deaths annually, with millions of patients affected by chronic disease.1 CHD mortality rates have halved over the past 2 decades.2 However, in spite of improved treatments, the actual number of patients affected by the disease remains large, especially affecting older age groups.3 Furthermore, CHD costs the UK economy over £30 billion per annum, with some £9 million falling on the NHS.4

UK health inequalities in socioeconomic circumstances (SEC) remain large.5 Moreover, as the leading cause of death in the UK, CHD is a major contributor to inequalities, especially in premature mortality. CHD mortality rates halved between 1980 and 2000, some 60% of this fall was attributed to risk factor reduction, while approximately 40% was attributed to improved treatments.2 However, treatment uptake appears surprisingly equitable across SECs.6 Much of the inequality in CHD deaths must, therefore, be attributable to differences in major cardiovascular risk factors.7

The large reductions in CHD mortality and the equitable administration of CHD treatment are both laudable. However, recent trends in both inequalities and total disease burden appear less rosy. When taking office in 1997, New Labour pledged to reduce the inequality gap in life expectancy between deprived areas and the general population by 2010. However, this has not been achieved.8 Helping understand why will require a deep analysis. We aim to both quantify the current burden of CHD, and identify recent trends and inequalities in this burden, stratified by age, gender and socioeconomic status.8 The results could thus be invaluable. First, to address shortcomings of previous inequality reduction policies; second, to assess implications for healthcare commissioning services; and third, to inform future prevention strategies.

Methods

Data sources

We synthesised data on the disease burden for the English population aged 25 years and older in 1999 and 2007(17 002 155 men and 18 278 670 women). We then quantified trends in disease burden between 1999 and 2007 stratified by age, sex and deprivation, using a range of databases as detailed below.

All potentially relevant data sources specific to the population in England were critically appraised.7 ,8 When several sources were available, we chose the most up-to-date, representative and least biased dataset.

Population estimates and CHD death counts (1999: ICD9 410–414; 2007: ICD10 I20–I25) by sex, 5-year age-bands to age 85+ years, and deprivation quintile were obtained from the Office for National Statistics.

Emergency admissions for acute myocardial infarction were extracted from Hospital Episode Statistics. This was supplemented with data from the Myocardial Ischaemia National Audit Project to disaggregate patient groups; we assumed that approximately 60% were ST elevation acute myocardial infarction, and 40% non-ST elevation acute coronary syndrome.9 For admissions for heart failure due to CHD, the NHS Heart Failure Survey 2005 was used to estimate in-hospital treatment uptake.10 The General Practice Research Database11 and the Health Survey for England provided data on disease prevalence and treatment uptake in the community.12

Stratifying data according to socioeconomic circumstances

All data sources recorded the individual's postcode of residence. We therefore used the Index of Multiple Deprivation (IMD) as a proxy indicator of the socioeconomic circumstances. This is a widely used measure of relative area deprivation covering 32 482 small areas (average population of 1500).13

The IMD is a composite index of relative deprivation at small-area level based on seven domains: income, employment, health deprivation and disability, education (skills and training), barriers to housing and services, crime and disorder, and living environment.13

We used the IMD 2007 to rank all lower super-output areas in England in ascending order of increasing deprivation, and then grouped them into equal quintiles. Although scoring low on the deprivation index does not necessarily equate to affluence, to assist readability we refer to quintile 1 as ‘most affluent’, and to quintile 5 as ‘most deprived’.

Based on the postcode of residence, the data providers matched CHD deaths and treated patients to their corresponding deprivation quintile before releasing the data to us.

This dataset therefore allowed us to:

  1. analyse trends in the burden of disease from 1999 to 2007,

  2. analyse trends within socioeconomic quintiles—SECs—over the 7-year period and

  3. compare the burden of disease from community-based and hospital patients.

Statistical analyses

In order to assess trends, we compare the incidence of hospital admission cases and the prevalence of community-based disease groups from 1999 with those of 2007.

We assessed the strength of SEC gradients by calculating age-standardised rate ratios for the start year (1999) and end year (2007), comparing SEC1 with SEC5 for all disease groups. We used 95% CIs for this analysis.

All rates were directly standardised to the European standard populations using 10-year age-bands to age 85+ years. Rate ratios and their 95% CIs were estimated using the Poisson approximation.14

Average reduction within age group

Between 1999 and 2007, the average age-group reduction was calculated as the mean reduction within each age group in men and women for each disease class.

Results

Overall CHD disease burden in 2007

Overall, of the 35 million adult population aged 25 and over living in England in 2007, approximately 1.7 million patients had CHD.

The CHD disease burden in 2007 comprised two major groups, hospital admissions and community-based patients living with chronic disease.

Approximately 1 536 425 patients (88% of the total) were treated solely in the community (mostly having a previous myocardial infarction, chronic angina or heart failure) (table 1).

Table 1

Trends in the burden of coronary heart disease patients

Approximately 205 000 CHD patients (12%) were admitted to hospital at least once during 2007 for either acute coronary syndromes (ST elevation myocardial infarction, Non-ST elevation myocardial infarction, unstable angina: 4%) revascularisation (4%) or heart failure due to CHD, (1.5%) (table 1).

Trends in hospital admissions 1999–2007

Hospital admissions for acute coronary syndromes

Hospital admissions included approximately 141 000 patients with acute coronary syndromes in 1999 falling to some 112 000 admissions in 2007 (table 1).

During the period 1999–2007, admission rates for acute coronary syndromes fell in both men and women across all age groups. The age-standardised incidence rate for acute myocardial infarction fell by approximately 21% (RR 0.79 (0.78 to 0.80)) during this period (figure 1, and online appendices).

Figure 1

Trends in acute myocardial infarction admissions. Hospital admission rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data source: Hospital Episode Statistics.

The age-standardised incidence rate for unstable angina fell even more during this period, by approximately 33% (RR 0.67 (0.66 to 0.68)).

Hospital admissions for CHD patients with heart failure

Approximately 24 600 CHD patients required hospital admission for heart failure during 2007 (table 1). Between 1999 and 2007, hospital admission rates for heart failure reduced substantially across all age groups in men and women. The age-standardised incidence rate fell by approximately 38% (RR 0.62 (0.61 to 0.64)) (figure 3).

Trends in community-based patients with chronic disease: 1999–2007

There were approximately 1.74 million patients with CHD living in the community in 1999, decreasing to approximately 1.53 million patients in 2007.

In the 8-year period, 1999–2007, age-specific rates for these disease groups, likewise, fell significantly in both men and women; the most marked reductions being seen in middle-aged patients (figures 4–6).

Patients with a previous myocardial infarction

In 1999, there were approximately 505 000 patients with a previous myocardial infarction in the community (32% women) decreasing to some 400 000 (36% women) in 2007.

The age-standardised prevalence rate reduced by approximately 37% (RR 0.63 (0.63 to 0.63)) during this period (figure 4).

Community angina

There were approximately 1 015 000 patients with chronic angina in 1999 falling to 985 000 in 2007.

Between 1999 and 2007, angina prevalence rates in the community fell in young and middle-aged groups. The age-standardised prevalence rate fell significantly by 10% (RR 0.90 (0.90 to 0.90)) during this period.

However, rates apparently rose in those over 75 years of age (figure 5).

In 2007, of approximately 67 500 patients (7% of the total community patients with chronic angina), some 18 500 patients were admitted for coronary bypass surgery, and approximately 49 000 patients for angioplasty procedures.

CHD patients with heart failure

In 1999, there were approximately 223 000 CHD patients with heart failure in the community (57% women) decreasing to some 172 800 (51% women) in 2007 ( table 1 ).

Between 1999 and 2007, heart failure prevalence rates in the community fell substantially in men and women in all age groups. The age-standardised prevalence rate decreased by 20% (RR 0.80 (0.798 to 0.813)) (figure 6).

Trends in disease inequalities in CHD patients

Large and persistent socioeconomic gradients were apparent in all disease groups, generally being steepest in youngest age groups, and most marked in hospital admissions of unstable angina or heart failure (figures 1–6).

Figure 2

Trends in unstable angina admissions. Hospital admission rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data source: Hospital Episode Statistics.

Figure 3

Trends in heart failure admissions. Hospital admission rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data source: Hospital Episode Statistics.

Figure 4

Trends in the prevalence of survivors of myocardial infarction in the community. Prevalence rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data source: General Practice Registry Database.

Figure 5

Trends in prevalence of angina in the community. Community prevalence rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data source: General Practive Registry Database.

Figure 6

Prevalence of heart failure in the community. Prevalence rates by age, sex and socioeconomic circumstance quintile (SEC) in 1999 and 2007. (A) Men, (B) Women. Data Source: General Practice Registry Database.

Inequalities were consistently greatest in the youngest groups aged <55 years, and decreasing through middle age (55–75 years).

Thus, comparing SEC5 (the most deprived quintile) with SEC1 (the most affluent) in patients aged under 55 years of age, there was a twofold ratio for myocardial infarction admission rates, and a threefold gradient in patients admitted with unstable angina.

Socioeconomic gradients in patients over 75years of age were generally more modest or had even disappeared (Figures 1–6).

Inequalities in hospital admissions

Acute coronary syndromes

There was a clear SEC gradient in hospital admissions with acute myocardial infarction. This was greatest in young and middle-aged men and women (figure 1). Age-standardised rate ratios, encompassing all gender and age groups, confirmed a clear and significant gradient in 1999 (RR 1.06 (1.03 to 1.10)) which persisted to 2007 (RR 1.10 (1.07 to 1.14)).

Unstable angina admissions demonstrated a substantial and persistent socioeconomic gradient across all age groups, again being steepest in the younger age groups. Overall, age-standardised rate ratios showed a marked gradient in 1999 (RR 1.34 (1.31 to 1.38)) which worsened by 2007 (RR 1.43 (1.38 to 1.47) (figure 1).

Inequalities in CHD-related heart failure admissions

Clear socioeconomic gradients were apparent across all age groups (figure 3) Age-standardised rate ratios showed a significant gradient in 1999 (RR 1.09 (1.04 to 1.13)) which persisted to 2007 (RR 1.13 (1.07 to 1.19).

Trends in disease inequalities in community-based patients

Significant and persistent SEC gradients were apparent, particularly in young and middle-aged patients.

In patients with a previous myocardial infarction, SEC gradients were present in 1999, persisting in 2007 in young and middle-aged patients. Age-standardised rate ratios showed a significant gradient in 1999 (RR 1.07 (1.06 to 1.08)) which persisted to 2007 (RR 1.07 (1.06 to 1.07)) (figure 4).

There were also significant and persistent gradients for community angina prevalence. These were most marked in middle-aged men and women. However, despite these persistent gradients in many of the age groups, age-standardised rate ratios showed an inverse gradient in 1999 (RR 0.68 (0.67 to 0.68)). This inverse gradient reduced to 2007 (RR 0.76 (0.75 to 0.77)) (figure 5).

Finally, there were significant socioeconomic gradients in heart failure prevalence in the community across most age groups in 1999. These became more marked in 2007. However, similar to age-standardised rate ratios for community angina, community heart failure showed an inverse age-standardised rate ratio in 1999 (RR 0.81 (0.79 to 0.82)), which reduced to 2007 (RR 0.96 (0.95 to 0.98)) (figure 6).

More detailed results are summarised in online appendices 1–5.

Discussion

The results demonstrate a reduction in both hospital admission rates and community rates from 1999 to 2007. However, a similar reduction in socioeconomic inequalities was not observed.

The ‘burden’ of CHD remains immense. The very visible and costly 200 000+ hospital admissions annually attract much interest, potentially diverting attention from the eightfold larger mass of patients with chronic disease in the community (approximately 1.6 million). However, these community patients also have a reduced life expectancy, impaired quality of life, a high risk of further cardiac events and associated costs.4

The burden of disease is best measured by patient numbers rather than rates.

Happily, this burden clearly declined, even within the brief 8-year period of this study. The true falls are likely to be even larger, because case ascertainment is likely to have improved between 1999 and 2007, reflecting financial incentives to maximise case finding in the community (notably Quality and Outcomes Framework (QoF)),1 ,15 and also incentives remunerating hospital admissions (Payment by Results).16

Population-based rates for CHD, likewise, generally declined between 1999 and 2007. Encouragingly, almost all disease groups shrank during this period, particular improvements being seen in middle-aged men and women.

CHD mortality rates also fell by approximately 35% between 1999 and 2007, with even steeper reductions in the most affluent groups. This reflects an improved environment (risk factor declines) as well as therapeutic advances, which account for approximately half this mortality reduction.7

The 35% decline in CHD-related heart failure prevalence rates, and in-hospital admission rates over the 8-year time period was particularly encouraging. This likely reflects a reduction in major risk factors (and hence, incidence) as well as an increasing use of evidence-based medication.17 However, heart failure due to CHD continues to disable over 170 000 patients in the community, quality of life is often poor, and some 14% require hospital admission each year.

Approximately 1 million patients live in the community with chronic angina. Every year, approximately 7% are admitted for angioplasty or CABG surgery, a surprisingly small proportion, given the high profile and costs of revascularisation.1

As in other disease groups, prevalence rates fell in young and middle-aged men and women. However, in those aged over 75 years, crude numbers and rates both apparently increased between 1999 and 2007. This was unexpected, and may simply reflect improvements in case ascertainment. Financial incentives, through the ‘QoF’ schemes may have powerfully encouraged more active case finding.15 Reassuringly, the exemplary British Regional Heart Study also found no increase in community CHD prevalence between 1975 and 2004.18

The apparent increased prevalence of chronic angina in elderly community patients might also partly reflect an increased life expectancy due to improved secondary prevention. For example, the uptake of statins and ACE-inhibitors doubled during this brief period.17

Furthermore, 14% of HF patients and 7% of chronic angina patients still required hospital admission in a single year. This represents huge personal, family and economic costs, and strengthens demands for better treatment and prevention.

Treatment uptake rates have been gratifyingly equitable across socioeconomic quintiles.6 This further emphasises the importance of addressing the inequalities in major cardiovascular risk factors in order to reduce the inequalities in disease.

The consistent falls in patient numbers, population-based rates and mortality rates are very encouraging. Lessons should be drawn from this success, especially the likely powerful role of reductions in major cardiovascular risk factors. Most notably, systolic blood pressure fell substantially during this period explaining some 30% of the mortality decline, with a further 6% coming from falls in total cholesterol. Both declines were attributed more to improved diet than to treatment with antihypertensives or statins, respectively.7 ,19

The marked falls in disease burden are encouraging. However, large socioeconomic gradients have persisted or even increased in some groups. This is very worrying and further highlights the inadequacy of previous policies to effectively reduce inequalities. Further research is clearly indicated. Moreover, diabetes and obesity worsened between 1999 and 2007, especially in deprived groups.7 In contrast, more pronounced reductions were seen in the lower SECs for systolic blood pressure and smoking. These may well reflect relatively effective population-wide policies used for tobacco control and salt reduction, respectively.20 Further lessons might therefore be learnt from prevention successes elsewhere. Effective and powerful policies in other countries have resulted in successful tobacco control and substantial dietary reductions in salt, saturated fats, trans-fats and sugars concealed in processed food, takeaways and sweetened drinks. These, in turn, have resulted in large and surprisingly rapid reductions in cardiovascular burden and mortality.20–22 Analysing the respective compliance rates, both with primary and secondary prevention methods across the socioeconomic groups may help to explain the relative success and failure of previous policies resulting in the unequal risk factor reductions described above.

This analysis has several strengths with large datasets covering the entire adult population of 35 million. Data quality is generally very good.7 ,12 ,13 ,23 ,24

This is the first study to quantify socioeconomic trends in the burden of coronary disease during recent years. The datasets used are also fairly representative of the socioeconomic distribution of the English population, while the large population (35 million) allows fairly precise estimates of these changes, and permits conclusions to be drawn.

This study also has limitations. The period of analysis from 1999 to 2007 was relatively short. However, the trends are plausible and consistent with other longer studies.18 This study also used an area-level categorisation of socioeconomic circumstances. This may, therefore, be suboptimal when analysing trends within individuals, or overall national trends in hospital admission and community rates. However, area deprivation measures generally correlate well with measures of individual socioeconomic position.7

Further, the hospital admission numbers reported may underestimate the true burden due to failures to capture all readmissions during the same year. Conversely, we may have double-counted a number of individuals. However, the community results probably offer reasonably representative proxy measures of disease prevalence and, hence, disease burden.

Although, detailed sex and age-group analyses provide more in-depth results, they may also increase complexity. The age-standardised rates are, therefore, useful to analyse overall trends. A future study might usefully analyse regional inequalities and recent trends.

In conclusion, this study stratifies the burden of CHD by age, gender and socioeconomic status both in 1999 and 2007. The burden of CHD in the English population is immense and unequal. The very visible hospital admissions potentially divert attention from the many more patients with chronic disease concealed in the community who also have a reduced life expectancy and high risks of further cardiac events.

Further research, stratifying these data by socioeconomic inequalities, and then specifically by geographical area might be valuable in identifying more specific targets for policy.

Population-based rates for hospital admissions and CHD prevalence have significantly declined since 1999. This is encouraging and probably reflects improved risk factor levels as well as therapeutic benefits. However, marked socioeconomic gradients have persisted, and even worsened. These persisting inequalities, therefore, leave no room for complacency—the case for effective cardiovascular prevention policies has never been stronger.

What is already known on this subject

  • The burden of cardiovascular disease in the UK is substantial, and dominated by coronary heart disease.

  • However, recent trends in the associated socioeconomic inequalities are not well studied.

What this study adds

  • Between 1999 and 2007, age-specific hospital admission rates for coronary heart disease decreased by 20%–35%. Community prevalence also decreased by 10%–20%.

  • In 2007, there were approximately 205 000 hospital admissions for myocardial infarction, unstable angina, revascularisation or heart failure, plus some 1.5 million patients with chronic angina, previous myocardial infarction or heart failure living in the community.

  • Strong socioeconomic gradients were apparent in all patient groups. These inequalities persisted or worsened between 1999 and 2007. There is, thus, no room for complacency.

Acknowledgments

We thank Iain Buchan, David Blane, Peter Goldblatt, Adrian Gallop, Klim McPherson, Martin McKee, Richard Morris and Colin Sanderson of the IMPACTsec advisory group for comments on study design and emerging findings. Particular thanks go to Dr Kate Walters for help with clinical coding lists to extract prescribing data from the GPRD, and to Dr Gianluca Baio for statistical advice.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Funding MB and SS are honorary research staff at UCL, funded by Legal & General Assurance Society Limited (L&G) as part of its wider research collaboration with UCL on understanding the drivers of longevity. JPS, RR and SC are supported by the Higher Education Funding Council, and RR is partly funded by the National Institute for Health Research University College London Hospital/University College London Comprehensive Biomedical Research Centre. MOF is partly funded by the UK Medical Research Council and from European Community's Seventh Framework Programme (FP7/2007-–2013) under grant agreement no223075—the MedCHAMPS project and NH by the NHS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The IMPACTsec team had access to all data sources and has the responsibility for the contents of the report and the decision to submit for publication.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Supplementary appendix (attached) available with open access from the corresponding author at jonathan.pearson-stuttard@chch.ox.ac.uk.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.