Secondary prevention of coronary heart disease in older British men: extent of inequalities before and after implementation of the National Service Framework

Abstract

Background

Deficiencies in implementation of secondary prevention of coronary heart disease (CHD) have been identified. We explored the extent of medication use for secondary prevention of CHD since the introduction of the National Service Framework (NSF) for CHD and the influence of patient age, social class, region and time since diagnosis in older British men.

Methods

Prospective study in 24 British towns using patient information on medication use in 1998-2000 and 2003. Subjects were men with medically recorded diagnosis of myocardial infarction or angina, aged 62-85 years in 2003. Prevalence of medication use (aspirin, statins, ACE inhibitors and beta-blockers) in 1998-2000 and 2003 was ascertained.

Results

Prevalence of use of all drugs increased in 2003 and was markedly higher in patients with a history of myocardial infarction than angina. Older age was related to lower prevalence of drug use, particularly statins. In 2000, older subjects (74-85 years) were 60% (95% CI 41%, 72%) less likely to receive statins compared with younger subjects (62-73 years); this pattern changed very little between 2000 and 2003. Although social class appeared to have little relation to drug use, the prevalence of use of all medications decreased with increasing time since diagnosis.

Conclusions

Although the uptake of medications for secondary prevention in CHD patients increased since the NSF in 2000, marked age inequalities in statin use were present both in 1998-2000 and 2003. Further action is needed to reduce these inequalities, since older patients are at particularly high risk of recurrent and fatal CHD.

Introduction

Coronary heart disease (CHD) remains the most common cause of death in the UK and a major cause of morbidity.¹ Among patients with established CHD, there is strong evidence that medications including aspirin, statins, ACE inhibitors and beta-blockers reduce the risk of recurrent CHD.² However, despite this clinical evidence and guidelines for the use of drugs in secondary prevention, several surveys conducted up to 2000 have shown that these medications were greatly underused in CHD patients.^3-8 Moreover, inequalities in the distribution of secondary prevention were also identified, with older patients being less likely to receive treatment.^9-14

In 2000, the National Service Framework for Coronary Heart Disease (NSF-CHD) set out a blueprint for the management of CHD, with secondary prevention established as a high priority.¹⁵ However, little is known about the extent to which inequalities in the provision of secondary prevention have changed since the introduction of the NSF. We have therefore examined the extent of overall changes in use of medications for secondary prevention between 2000 and 2003, and the influence of age, social class, region and time since diagnosis on the use of the medications at the two time points in a population based study of older British men aged 63-85 years with CHD.

Methods

The British Regional Heart Study (BRHS) is a prospective study of cardiovascular disease comprising a socially and geographically representative sample of men from one general practice in each of 24 towns representing all major British regions. Participants were enrolled in 1978-80 and have been followed up through the National Health Service Central Register (Southport and Edinburgh), general practice records and questionnaires to study participants. Details of the BRHS are reported elsewhere.¹⁶

For the present report, myocardial infarction and angina are based on diagnoses documented by the General Practitioner (GP), identified from regular two yearly reviews of GP records.¹⁶ A diagnosis of myocardial infarction was validated in accordance with World Health Organization criteria. The categories of myocardial infarction and angina have been considered separately; patients with both diagnoses have been included as myocardial infarction cases. Reports of current medication are based on questionnaires completed by study participants in 1998-2000 (at the time of a 20 year follow-up examination) and in 2003 (postal questionnaire). Details of all current medications including non-prescription medications were sought on each occasion, with a specific additional question on aspirin use.

Prevalence of use of different medications (aspirin, statins, ACE inhibitors and beta-blockers) were calculated at the two time points according to age, social class and time since last diagnosis of myocardial infarction or angina. Subjects were grouped into older and younger age groups, divided at the mid year of birth of the study population (1929). In 2003 younger subjects were aged 62-73 years and older subjects were 74-85 years. Social class was derived from the longest held occupation recorded at the time of baseline questionnaire using the Registrar General’s classification of occupations, with categories grouped as non-manual (I, II and III non-manual) and manual (III manual, IV and V). Men in the armed forces at baseline were excluded from this analysis. Regions were divided into north and south depending on whether the town of residence was north or south of a line joining Bristol and the Wash. Time since diagnosis was ascertained using the date of last (most recent) diagnosis of myocardial infarction or angina before 2003 and therefore is calculated only at this point in time. The ‘time since diagnosis’ variable was categorised into three groups of less than 5 years, 5-10 years and greater than 10 years.

For each questionnaire time point, prevalence ratios were calculated to ascertain the effect of age, social class, region and time since diagnosis on the use of different medications. Since the outcome of our analyses, use of medication, is a common outcome, prevalence ratios were considered to be a more appropriate measure of effect than odds ratios. Prevalence ratios and their confidence intervals were calculated using the ClinStat software. However, to mutually adjust for the effect of variables such as age, social class, region and time since diagnosis logistic regression was carried out, using SAS version 8. To investigate whether age and social class effects differed between 2000 and 2003, generalised estimating equations were applied using STATA as the same men contributed data at both time points.¹⁷

Results

The response rate for questionnaires to all men was 77% in 2000 and 80% in 2003. Analyses were based on 332 myocardial infarction cases and 485 angina cases studied in 2000, and on 336 myocardial infarction cases and 521 angina cases studied in 2003. Details of medication prevalence by age and social class are presented in table 1, with the corresponding prevalence ratios. The prevalence of medication use was generally highest for aspirin followed by statins, beta-blockers and ACE inhibitors. Prevalence rates were higher for myocardial infarction cases than for angina and tended to increase between 2000 and 2003, except for the use of aspirin in older myocardial infarction subjects, already over 80% in 2000. Statins and to a lesser extent beta-blocker medications were more likely to be received by younger rather than older subjects. The relative differences in statin use were particularly marked for patients with myocardial infarction (the proportion receiving statin was reduced by 60% in older subjects compared with younger subjects both in 2000 and by 39% in 2003). For angina patients, the proportion of older subjects receiving statins was reduced by 38% compared with younger subjects in 2000 and by 31% in 2003. Prevalence of aspirin and ACE inhibitor use, in contrast, showed no strong relation with age. Prevalence of medication use tended to be lower among manual social class groups, though these differences were not generally statistically significant. There was no evidence of any difference in medication prevalence between regions (data not presented).

Table 1.

Prevalence ratios with 95% confidence intervals (CI) for use of secondary prevention medication for CHD by age and social class in 2000 and 2003

MYOCARDIAL INFARCTION CASES*							ANGINA ONLY CASES
YEAR	2000 (N = 332)			2003 (N = 336)			2000 (N = 485)			2003 (N = 521)
Use of medication by age
	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)
Aspirin  Younger  Older	173 159	142 (82) 130 (82)	1.00 0.99 (0.90, 1.10)	185 151	161 (87) 122 (81)	1.00 0.93 (0.84, 1.02)	238 247	161 (68) 157 (64)	1.00 0.94 (0.82, 1.07)	288 233	201 (70) 170 (73)	1.00 1.04 (0.94, 1.16)
Statins  Younger  Older	173 159	75 (43) 28 (18)	1.00 0.40 (0.28, 0.59)	185 151	146 (79) 73 (48)	1.00 0.61 (0.51, 0.73)	238 247	61 (26) 39 (16)	1.00 0.62 (0.43, 0.88)	288 233	165 (57) 92 (39)	1.00 0.69 (0.57, 0.83)
ACE inhibitors  Younger  Older	173 159	43 (25) 49 (31)	1.00 1.24 (0.87, 1.75)	185 151	85 (46) 62 (41)	1.00 0.89 (0.69, 1.14)	238 247	40 (17) 41 (17)	1.00 0.98 (0.66, 1.47)	288 233	88 (31) 67 (29)	1.00 0.94 (0.72, 1.23)
Beta blockers  Younger  Older	173 159	70 (40) 48 (30)	1.00 0.75 (0.55, 1.00)	185 151	93 (50) 54 (36)	1.00 0.71 (0.55, 0.92)	238 247	80 (34) 63 (26)	1.00 0.76 (0.57, 1.00)	288 233	117 (41) 73 (31)	1.00 0.77 (0.61, 0.97)
Use of medication by social class
	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)
Aspirin  Non-manual  Manual	138 181	117 (85) 143 (79)	1.00 0.93 (0.84, 1.03)	143 180	122 (85) 149 (83)	1.00 0.97 (0.88, 1.06)	202 272	139 (69) 171 (63)	1.00 0.91 (0.80, 1.04)	241 264	178 (74) 182 (69)	1.00 0.93 (0.84, 1.04)
Statins  Non-manual  Manual	138 181	40 (29) 57 (31)	1.00 1.08 (0.77, 1.52)	143 180	86 (60) 124 (69)	1.00 1.14 (0.97, 1.35)	202 272	52 (26) 45 (17)	1.00 0.64 (0.45, 0.91)	241 264	123 (51) 123 (47)	1.00 0.91 (0.76, 1.09)
ACE inhibitors  Non-manual  Manual	138 181	35 (25) 54 (30)	1.00 1.17 (0.81, 1.69)	143 180	64 (45) 79 (44)	1.00 0.98 (0.76, 1.25)	202 272	40 (20) 39 (14)	1.00 0.72 (0.48, 1.08)	241 264	81 (34) 70 (27)	1.00 0.78 (0.60, 1.03)
Beta blockers  Non-manual  Manual	138 181	56 (41) 54 (30)	1.00 0.73 (0.54, 0.99)	143 180	67 (47) 76 (42)	1.00 0.90 (0.71, 1.15)	202 272	59 (29) 79 (29)	1.00 0.99 (0.75, 1.32)	241 264	83 (34) 98 (37)	1.00 1.07 (0.85, 1.36)

^*Myocardial infarction cases include those with or without angina;

N – total in each category; n – prevalence of medication use for each category

Table 2 shows the relationship between date of diagnosis and the use of medications. Date of diagnosis was strongly related to prevalence of drug use. For all medications, subjects diagnosed within the last five years had a higher prevalence of receiving secondary prevention compared with earlier diagnoses. The decline in prevalence of medication use became marked with a diagnosis 10 years ago or more but the decline appeared more recent, at 5-10 years for ACE inhibitors and beta-blockers in myocardial infarction patients.

Table 2.

Prevalence ratios with 95% CI for use of secondary prevention medications for CHD according to time since last diagnosis before 2003

	Myocardial infarction cases*			Angina only cases
	N=336			N=521
Medication type	N	n (%)	Prevalence ratio (95% CI)	N	n (%)	Prevalence ratio (95% CI)
Aspirin  Time since diagnosis   < 5 years   5-10 years   10 or more	81 101 154	72 (89) 90 (89) 121 (79)	1.00 1.00 (0.90, 1.11) 0.88 (0.79, 0.99)	246 156 119	180 (73) 118 (76) 73 (61)	1.00 1.03 (0.92, 1.16) 0.84 (0.71, 0.99)
Statins  Time since diagnosis   < 5 years   5-10 years   10 or more	81 101 154	62 (77) 64 (63) 93 (60)	1.00 0.86 (0.68, 1.00) 0.79 (0.66, 0.94)	246 156 119	130 (53) 78 (50) 49 (41)	1.00 0.95 (0.77, 1.15) 0.78 (0.61, 0.99)
ACE inhibitors  Time since diagnosis   < 5 years   5-10 years   10 or more	81 101 154	47 (58) 43 (43) 57 (37)	1.00 0.73 (0.55, 0.98) 0.64 (0.48, 0.84)	246 156 119	81 (33) 46 (29) 28 (24)	1.00 0.89 (0.66, 1.21) 0.71 (0.49, 1.03)
Beta blockers  Time since diagnosis   < 5 years   5-10 years   10 or more	81 101 154	56 (69) 39 (39) 52 (34)	1.00 0.56 (0.42, 0.74) 0.48 (0.37, 0.64)	246 156 119	99 (40) 59 (38) 32 (27)	1.00 0.94 (0.73, 1.21) 0.67 (0.48, 0.93)

^*Myocardial infarction cases include those with or without angina;

N – total in each category; n – prevalence of medication use in each category

The effects of age, social class and time since diagnosis were unchanged by mutual adjustment (results not presented). The relation between age and prevalence of drug use was further explored by examining whether increasing comorbidity (particularly with cancer) might account for the lower rates of statin use in older subjects. However, the prevalence of cancer was extremely low (in 2003 none of the myocardial infarction patients had cancer and only 6% of angina patients reported having cancer). Excluding these subjects from analysis had no effect on the age differences in prevalence of myocardial infarction and angina.

Discussion

Main findings of this study

Completeness of uptake of secondary prevention has increased from 2000 to 2003, though even in 2003 fewer than three quarters of patients with CHD were receiving a statin and fewer than half were receiving a beta-blocker or ACE inhibitor. The prevalence of medication use has remained lower among angina patients compared with those with myocardial infarction. Prevalence of statin use in particular has been lower among older patients; the age gradient has remained unchanged between 2000 and 2003. Medication use also decreased with an increase in the time since diagnosis of a CHD event.

What is already known about this topic

Medication use, amongst other forms of secondary prevention, reduces the risk of recurrence of CHD. In earlier studies,^18-24 patients at high risk of not receiving medications of secondary prevention were those with milder disease (angina rather than myocardial infarction), older patients and those whose diagnosis was not recent. Similar differences have also been described in earlier reports from the BRHS. ^19;25 However, little is known about the extent of these inequalities since the NSF for CHD in 2000. It is now clearly established that the relative benefits of statin treatment are maintained among men above the age of 75 years and that the absolute benefits of secondary prevention are high, especially given the high case-fatality of major CHD events among older people.^26-29 Underuse of beta-blockers in secondary prevention among older patients has also been shown to relate to reduced survival.³⁰

What this study adds

Our study is based on a highly representative sample of older British men.¹⁶ The focus on men who had a confirmed doctor diagnosis of either myocardial infarction or angina should provide conservative estimates of undertreatment, since by definition all the CHD cases studied were known to their doctors. The population-based (rather than hospital based) nature of the study and the lack of intervention in the participating General Practices should mean that the results of our study are generalisable to older British men and to primary care settings.

The use of secondary prevention medications remained incomplete older British men even in 2003, particularly for ACE inhibitors and beta-blockers and to a lesser extent for statins; only aspirin was used in more than 70% of patients with myocardial infarction and angina by 2003. Although statin use had increased markedly from 2000, by 2003 it was still being received by only 65% of all patients with myocardial infarction and half of patients with angina. Age-related inequalities in uptake (particularly of statins and to a lesser extent beta-blockers) continued to be evident in 2003.

The reason for continuing underuse of secondary prevention medications, particularly at older ages - despite strong evidence of benefit - remains uncertain. This phenomenon, described as the ‘treatment-risk paradox’³¹ may reflect a time lag between publication of research findings (including relevant clinical trials) and their dissemination into clinical practice.^32;33 Similar discrepancies in adherence to guidelines and recommendations have also been reported in Europe, Canada and the USA.^34-37 The limited use of statins in older subjects could reflect the view of practitioners that life is not worth prolonging at older ages, or it may reflect rationing on cost grounds, with discrimination against older patients because of their shorter life expectancy.³⁸ The National Service Framework’s focus on targets specifically aimed at subjects under 75 years may have also contributed to the lower treatment rates among older subjects seen in our study.

Uptake of secondary prevention was somewhat lower among patients in manual occupations in this study, though these differences were not for the most part statistically significant. It is therefore difficult to exclude a moderate social class difference, given the limited statistical power of analysis in the number of CHD cases available for analysis. However, the absence of marked social class gradients in statin use in other studies suggests that the pattern of statin distribution does not contribute strongly to social class differences of CHD.^39-41,42

Prevalence of secondary prevention medication use also varied by time since diagnosis of last CHD event, for all medications studied. In our study subjects with older diagnoses were receiving less medication, an effect that was independent of the age of subjects. However, the time course differed, with reasonably consistent patterns of aspirin use among men whose last event had occurred over the previous ten years but with high use of other medications being maintained for a relatively short period (less than five years) since diagnosis. This reflects to some extent the earlier establishment of agreement about the use of aspirin, compared with statins and ACE inhibitors. The decreasing likelihood of receiving medication with an increase in time since diagnosis has been observed previously in studies on aspirin and statin use.^19;43 This suggests that there has been a stronger focus on establishing secondary prevention among subjects with recent acute CHD, rather than optimising prevention in patients diagnosed sometime previously. Given the long-term prognostic consequences of CHD diagnosis, particularly in the case of myocardial infarction,⁴⁴ this imbalance needs to be addressed.

Limitations of this study

Since this study ascertains current medication use, recall bias is unlikely. But, it is possible that under-reporting is more marked at older ages. However, studies have suggested that compliance with cardiovascular medications is high^45;46, and that this applies in older patients (≥65 years).⁴⁷ This suggests that patients would be likely to report medication usage accurately. Moreover, the difference between age groups was observed predominantly for statin use; it is unlikely that any reporting differences between age groups would apply to statins alone. The alternative method of using prescription rates to ascertain drug use may not always reflect the true use of medication by patients.

Implications of this study

Further efforts to improve the completeness of secondary prevention medication among older patients with CHD are warranted. The reasons underlying the lower rates of use (particularly of statins) among older patients – whether lack of knowledge about benefits in older people, a perception of the lack of value of extending life at older ages or age-related discrimination and rationing – needs to be further explored.⁴⁸ If the age inequality apparent in our study is a result of implicit rationing of resources then the basis of priority setting should be re-examined. Rationing of health resources needs to be assessed on the basis of the individual’s ability to benefit and not on the basis of age.⁴⁹ It may be useful to explore the attitudes of general practitioners and physicians to investigate whether they consider medication use in secondary prevention worthwhile in the elderly. Investigating the views of general practitioners may also help in identifying other potential barriers to implementation of guidelines such as structural or process issues (i.e. lack of resources for prescribing or staff shortage).⁵⁰ In addition, targets for secondary prevention should be amended where necessary to ensure that they do not facilitate age discrimination, as was the case in the NSF for CHD. Since majority of CHD patients receive most of their clinical care in Primary Care, improvements in completeness of secondary prevention in the CHD population as a whole are likely to have their origins in Primary Care. Implementation of the National Service Framework in tertiary care has already brought about dramatic increases in revascularisation rates. There is now a need to refine CHD prevention strategies with a special emphasis on older patients to ensure that all the potential benefits of secondary prevention are realised.