North of England evidence based guideline development project: guideline on the use of aspirin as secondary prophylaxis for vascular disease in primary careBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7140.1303 (Published 25 April 1998) Cite this as: BMJ 1998;316:1303
- Martin Eccles (), professor of clinical effectivenessa,
- Nick Freemantle, senior research fellowb,
- James Mason, research fellowb,
- the North of England Aspirin Guideline Development Group.
- a Centre for Health Services Research, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4AA
- b Centre for Health Economics, University of York, York YO1 5DD
- Correspondence to: Professor Eccles
- Accepted 16 December 1997
Patients who have had cardiovascular disease and stroke are treated with aspirin to reduce their subsequent risk of vascular events or death and thereby to increase the length and quality of their life. This guideline aims to provide general practitioners with evidence linked recommendations on the use of aspirin as secondary prophylaxis for cardiovascular disease and stroke in patients at high risk of these disorders. It is assumed that doctors will use their knowledge and clinical judgment in managing individual patients in the light of available resources. Recommendations may not be appropriate for use in all circumstances. This is a summary of the full version of the guideline.1
The use of aspirin in the secondary prophylaxis of vascular disease is cost effective
Aspirin should be used in patients with acute myocardial infarction, prior myocardial infarction, stable and unstable angina, and prior stroke or transient ischaemic attack
In acute myocardial infarction a dose of 150 mg daily should be used
In the other indications a dose of 75 mg daily should be used
In general practice, patients with a raised risk of vascular disease present with several disorders—acute or previous myocardial infarction, unstable or stable angina, transient ischaemic attacks, and peripheral vascular disease. The incidence and prevalence of these conditions and the workload associated with them in general practice can be estimated from the recent national morbidity survey in general practice for England and Wales, and is shown in the table.2
Throughout this guideline the strength of statements on evidence and of recommendations is categorised according to the scheme discussed in the first paper in the series.3 The box below shows these categories in descending order of importance.
Categories of strength used in statements
Strength of evidence
Ia—Evidence from meta-analysis of randomised controlled trials
Ib—Evidence from at least one randomised controlled trial
IIa—Evidence from at least one controlled study without randomisation
IIb—Evidence from at least one other type of quasi-experimental study
III—Evidence from descriptive studies, such as comparative studies, correlation studies, and case-control studies
IV—Evidence from expert committee reports or opinions or clinical experience of respected authorities, or both
Strength of recommendations
A—Directly based on category I evidence
B—Directly based on category II evidence or extrapolated recommendation from category I evidence
C—Directly based on category III evidence or extrapolated recommendation from category I or II evidence
D—Directly based on category IV evidence or extrapolated recommendation from category I, II, or III evidence
Use of aspirin
Aspirin as an antiplatelet agent
The potential importance of aspirin treatment in patients with a raised risk of vascular diseases has been well described in a recent meta-analysis, the latest update of work reported by the Antiplatelet Trialists' Collaborative Group.4 The treatment recommendations in this guideline draw on that work and, where necessary, develop it further by including subsequent trials. We have calculated a pooled risk ratio in relation to clinical subgroups and overall for the impact of antiplatelet treatment on subsequent myocardial infarction, stroke, and death from vascular causes.
This meta-analysis of 83 randomised trials on antiplatelet treatment is summarised in figure 1, which shows that the overall pooled risk ratio is 0.79 (95% confidence interval 0.76% to 0.82%). Tests for heterogeneity provide good evidence of homogeneity between the trials included in this analysis (Q=74.81, df=82, P=0.70). Thus, strong evidence exists that antiplatelet treatment has a protective effect in patients at raised vascular risk. Trials used several different antiplatelet drugs, including aspirin. Indirect comparison of trials in which aspirin or an alternative antiplatelet drug was used provides no evidence of systematic differences in effect (fig 2).
Substantially different risk reductions in mortality and morbidity are found for different disorders. This may be confounded by differences in the design of studies (particularly the length of follow up), but nevertheless reflects true variation in the potential benefits of treatment in diverse groups of patients. Pooled risk differences for each disorder are described in figure 3.
While 5% of patients in the unstable angina trials benefit from aspirin, the mean benefit in patients with diabetes is much smaller, and the confidence intervals around it are wide. The overall estimate of effect is 3.4% (2.9% to 4.0%) in the fixed effects model. The substantial heterogeneity of effect between studies is unlikely to have occurred by chance (Q=107.38, df=82, P=0.032). The pooled estimate of overall effect in relation to the random effects model is 2.7% (1.7% to 3.7%).
Duration of treatment
The major trials of aspirin have used different study durations ranging from 1 to 48 months. Those recommendations that are made for treatment over a time period covered by the evidence from clinical trials are designated “A”; those that result from extrapolation beyond the period covered by a trial are designated “D.” However, in the case of unstable angina, a “D” recommendation was upgraded. This is because the evidence from trials in these patients extends for 18 months, and the guideline development group felt that after this time patients could be regarded as having stable angina and be treated under the recommendations for that condition.
Dosage of aspirin
Trials have used different doses of aspirin; more recent trials have tended to use lower doses. There is no evidence that aspirin in doses greater than 75 mg provides greater benefit, and three recent major trials have used this regimen.5-8 Comparison of the effects of treatment in studies examined here, or in the broader range of comparisons reported by the antiplatelet trialists' collaboration, shows no evidence of differences (fig 4).4 We have therefore recommended that the dosage for antiplatelet treatment should be 75 mg of aspirin daily, except in the case of acute myocardial infarction, where most of the evidence is provided by a single trial using twice that dose.9
Aspirin and vascular disease
Acute myocardial infarction
Statement: giving aspirin within 24 hours of acute myocardial infarction lowers the risk of a vascular event over the subsequent month (Ia)
Nine trials have examined the role of antiplatelet treatment after acute myocardial infarction.9-19 These trials provide a pooled risk difference of 3.8% (2.8% to 4.7%). The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one month results in a 3.3% reduction in the risk of myocardial infarction, stroke, or death from vascular causes—or a number needed to treat value of 30. Limited follow up from the second international study of infarct survival9 indicates that the benefit of one month's treatment with aspirin may be maintained for four years.20
Recommendations: acute myocardial infarction
Patients with a suspected acute myocardial infarction should be treated with 150 mg of aspirin daily (A)
Patients with proved acute myocardial infarction should be given 150 mg of aspirin daily for one month (A)
After one month, patients should be treated according to the section on previous myocardial infarction (A)
Previous myocardial infarction
Statement: aspirin given to patients who have previously had a myocardial infarction lowers their risk of a subsequent vascular event (Ia)
Within this area of investigation there are 11 trials, and average treatment periods vary from 12 to 41 months.21-40 Overall, these trials give a risk difference of 3.2% (2.2% to 4.2%) for myocardial infarction, stroke, or vascular death. The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one year results in a 1.5% reduction in the risk of myocardial infarction, stroke, or death from vascular causes—or a number needed to treat of 65.
Recommendations: previous myocardial infarction
Patients who have previously had a myocardial infarction should be treated with 75 mg of aspirin daily for three years (A)
After three years, aspirin should be continued long term at a dose of 75 mg daily (D)
Statement: aspirin given to patients with stable angina lowers their risk of having a subsequent vascular event (Ia)
In the six studies examining the effectiveness of antiplatelet treatment for patients with stable angina, the overall risk difference was 4.5% (1.9% to 7.1%). 5 40-46 The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one year results in a 0.7% reduction in the risk of myocardial infarction, stroke, or death from vascular causes—or a number needed to treat of 150.
Recommendations: stable angina
Patients who have stable angina should be treated with aspirin 75 mg daily for four years (A)
After four years, aspirin should be continued long term at a dose of 75 mg daily (D)
Aspirin given to patients with unstable angina lowers their risk of having a subsequent vascular event (Ia)
Seven trials provide evidence of the effectiveness of antiplatelet treatment in unstable angina. 7 47-54 Overall, a risk difference of 5.5% (3.4% to 7.5%) in the incidence of myocardial infarction, stroke, or death from vascular causes was achieved. There is no evidence of heterogeneity between estimates of treatment effect (Q=8.72, df=6, P=0.19). The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one year results in a 6.6% reduction in the risk of myocardial infarction, stroke, or death from vascular causes—or a number needed to treat of 15.
Recommendations: unstable angina
Patients with suspected unstable angina should be treated with 75 mg of aspirin daily for 18 months (A)
After 18 months, patients with a history of unstable angina should be treated in accordance with the recommendation for stable angina (A)
Previous stroke or transient ischaemic attack
Statement: aspirin given to patients with a history of transient ischaemic attack or mild to moderate stroke lowers their risk of a subsequent vascular event (Ia)
Within this area of investigation there are 19 trials, showing an overall risk reduction of 4.3% (2.8% to 5.8%). 6 55-83 The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one year brings a 1.4% reduction in the risk of myocardial infarction, stroke, or death from vascular causes—or a number needed to treat of 69.
Recommendations: previous stroke or transient ischaemic attack
Patients with a history of a stroke or transient ischaemic attack should be treated with 75 mg of aspirin daily for four years (A)
Computed tomography is unnecessary before starting treatment in these patients (D)
After four years, aspirin should be continued long term at a dose of 75 mg daily (D)
Statement: aspirin given to patients with intermittent claudication seems to have a small and statistically uncertain effect on the risk of a vascular event (Ia)
Substantial evidence from 23 randomised trials shows that antiplatelet treatment for intermittent claudication is unlikely to have a beneficial effect on the subsequent incidence of non-fatal myocardial infarction, non-fatal stroke, and death from vascular causes.84-110 Overall, the risk difference in favour of treatment is 1.3% (−0.1% to 2.7%), a difference of borderline significance. The use of antiplatelet treatment in patients with intermittent claudication for reasons other than secondary prophylaxis of vascular events is discussed elsewhere. 4 111
Recommendations: intermittent claudication
Patients with intermittent claudication who have additional indications of raised vascular risk should be treated in line with the recommendations for that indication (D)
There is insufficient evidence to support use prophylactic aspirin in patients with intermittent claudication but no additional vascular risk factors (A)
Statement: aspirin given to patients with diabetes seems to have a small and statistically uncertain effect on the risk of a vascular event (Ia)
Within this area of investigation there are eight trials.112-122 These show an overall estimate of risk difference of 1.2% (−0.9% to 3.3%), which is of uncertain significance. There is no evidence of heterogeneity of treatment effect between these trials (Q=7.66, df=7, P=0.36). The pooled incidence rate difference, shown by random effects model, estimates that antiplatelet treatment for one year brings a non-significant reduction in the risk of myocardial infarction, stroke, or death from vascular causes of 0.3%—or a number needed to treat of 360.
Diabetic patients with additional indications of a raised risk of vascular events should be treated in line with the recommendations for those indications (D)
There is insufficient evidence to support the use of prophylactic aspirin in patients with diabetes but no additional risk factors (A)
Safety and cost effectiveness of aspirin
Side effects and costs of aspirin
Statement: the benefits of using aspirin in the secondary prophylaxis of vascular disease considerably outweigh the attributable risks of gastrointestinal or cerebrovascular bleeding (Ib) Statement: the use of aspirin is likely to be cost saving or cost neutral (IV)
The dosages, cautions, contraindications, and side effects of aspirin as an antiplatelet drug are described in the British National Formulary, section 2.9.123 All guideline recommendations for treatment apply only in the absence of recognised cautions, contraindications, side effects, or interactions as documented in the latest version of the formulary.
The net value of aspirin to individual patients must balance the increased risk of haemorrhage against the reduced likelihood of a cardiovascular event. A recent review examined all trials listed in the antiplatelet trialists' collaboration for information on toxicity.124 When patients receiving aspirin and placebo were compared, the pooled odds ratio for all forms of gastrointestinal bleeding was 2.0 (1.5 to 2.8) and for bleeding leading to hospital admission was 1.9 (1.1 to 3.1). Similarly, when these groups were compared for either peptic ulcer or gastrointestinal symptoms leading to withdrawal of treatment, the pooled odds ratios were 1.3 (1.1 to 1.6) and 1.5 (1.1 to 1.9). The review found a consistent tendency of lower rates of adverse events in lower dose trials.
Two major trials in which a dosage of 75 mg aspirin daily was used have been reported. 5 6 It is possible to project the major benefits and risks attributable to this treatment regimen, but estimates should be treated with caution, since the trials did not have enough power to measure adverse effects at conventional levels of statistical significance. Assuming 1000 person years of treatment, the following effects are attributable to 75 mg of aspirin daily:
Ten patients with stable angina will avoid vascular events (non-fatal or fatal myocardial infarction or sudden death). However, one patient will have a fatal bleed, one will have a major non-fatal bleed, and one patient will experience a minor bleed (where “bleed” includes stroke and gastrointestinal haemorrhage).
Twenty nine patients who have previously had a transient ischaemic attack or minor stroke will avoid a vascular event (non-fatal or fatal stroke or other vascular death). However, six patients will have a serious bleed (possibly fatal) and eight patients will suffer less serious bleeds.
We found no adequate cost-benefit analyses of aspirin as an antiplatelet drug. The cost of aspirin itself is negligible. Generic aspirin, in 75 mg dispersible tablets, costs approximately £1 per year to prescribe, although proprietary brands may cost 10-20 times more. From a health service perspective, the net cost includes the cost of aspirin, treatment for attributable adverse events, and savings from fewer vascular events. Patients with vascular disease tend to consult their general practitioner regularly, and any increase in consultation because of treatment with aspirin would probably be small. Since the reduction in vascular events exceeds considerably the attributable adverse events, and given the nature of the medical interventions for both, aspirin treatment probably results in a net cost saving to the health service. The balance of costs could shift adversely if it were necessary to provide expensive H2 agonists to ameliorate gastrointestinal symptoms in a number of patients. The trials reported above, however, do not indicate that this would be the case. Aspirin is probably cost saving or cost neutral, although formal cost calculation has not proved possible because of inadequate hospital cost data.
Recommendation: secondary prophylaxis of vascular disease
Use of aspirin in the secondary prophylaxis of vascular disease is cost effective (D)
The following research needs were identified by the guideline development group during the preparation of this document:
Many of the trials of antiplatelet treatment were conducted before the introduction of other important treatments for some groups of patients with ischaemic heart disease (for example, angiotensin converting enzyme inhibitors for heart failure): all potential interactions between the actions of relevant drugs have not been explored.
Further research on the appropriate duration of treatment is required.
A formal evaluation of the cost effectiveness of aspirin and other antiplatelet agents is required.
• Further trials are needed to examine the effect of 75 mg of aspirin daily in patients with intermittent claudication or diabetes as the only risk factor for vascular disease.
The guideline development group comprises the following members, in addition to the authors: Mr Mark Campbell, prescribing manager, Wolfson Unit of Clinical Pharmacology, University of Newcastle upon Tyne; Dr David Graham, general practitioner, Hexham; Dr Keith MacDermott, general practitioner, York; Dr Tony McKenna, general practitioner, Stockton-on-Tees; Dr Maureen Norrie, general practitioner, Eston, Middlesborough; Dr Colin Pollock, medical director, Wakefield Health Authority; Dr Helen Rogers, senior lecturer and consultant in stroke medicine, Centre for Health Services Research, University of Newcastle; Dr Jeff Rudman, general practitioner, Distington.
The project steering group comprises: Professor Michael Drummond, Centre for Health Economics, University of York; Professor Andrew Haines, Department of Primary Care and Population Sciences, University College, London Medical School and Royal Free Hospital School of Medicine; Professor Ian Russell, Department of Health Sciences and Clinical Evaluation, University of York; Professor Tom Walley, Department of Pharmacology and Therapeutics, University of Liverpool.
We thank the following for reviewing the full version of the draft guideline: Dr Phil Ayres, Dr Richard Baker, Professor Stuart Cobbe, Dr Chris Griffiths, and Dr Andrew Herxheimer. Janette Boynton, Julie Glanville, Susan Mottram, and Anne Burton are thanked for their contribution to the functioning of the guideline development group and the development of the practice guideline.
Funding: The work was funded by the Prescribing Research Initiative of the UK Department of Health.
Conflict of interest: None.