Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Aspirin in high risk patients

Particular high risk groups
Particular antiplatelets
Aspirin or clopidogrel?
Most of us are content that aspirin is effective in reducing vascular events in people at high risk, and while not every candidate patient receives aspirin, every year more of them do. Much of the impetus has come from impressive meta-analyses showing solid evidence of efficacy. Reinforcement in the form of an updated meta-analysis with many more patients in randomised trials is a welcome addition [1]. No précis can do justice to the detail of a paper that takes 16 pages of the BMJ plus more available electronically, but the main results carry the message.


All relevant trials available up to September 1997 were sought through an intensive search strategy of electronic databases, trial registers, manual searching, reference lists and inquiry to researchers and pharmaceutical companies. Trials had to compare an antiplatelet regimen with control or with another antiplatelet regimen in patients at high annual risk (over 3% a year). Trials had to be properly randomised.

The outcome was one of preventing 'all bad things happening', in this case defined as a serious vascular event (non-fatal myocardial infarction, non-fatal stroke, or death). Individual trialists were asked about details of randomisation, blinding, duration of treatment and follow up, and those with at least 200 randomised patients were asked for information on individual patients (age, sex, blood pressure, medical history) and outcomes in individual patients.

Deaths were divided into those with a vascular cause (cardiac, cerebrovascular etc) and those that were definitely non-vascular. Strokes were subdivided into intracranial haemorrhage and those that were ischaemic or of unknown cause. Major extracranial bleeds were also considered an endpoint, and the definition of serious was applied for those needing hospital admission or blood transfusion.


There were 195 trials with information on vascular events (135,000 patients) that compared antiplatelet with control, and 89 (77,000 patients) that compared different antiplatelet regimens. There were 7,705 serious vascular events (10.7%) in 71,912 patients with antiplatelet therapy and 9,502 (13.2%) in 72,139 with control.

The analysis provides results in various combinations, for instance combining all doses of all antiplatelets used for any duration above one day for particular conditions. When analysing particular antiplatelets against control, only aspirin doses are given separately. Numbers needed to treat were not calculated, nor odds ratios. Rather results are given in terms of benefit or harm for 1,000 patients treated together with the percentage reduction in odds of an event.

Particular high risk groups

The results for the five main categories of high risk patients are shown in Table 1 for the outcome of any vascular event. There was a consistent reduction in odds of any vascular event of about 20-30%, except for treatment of acute stroke with a mean of only three weeks of treatment. In all other high risk categories the benefit per 1,000 patients was 20-40, over one month for treatment of acute myocardial infarction, and over 22-29 months for other conditions.

Table 1: Results in particular high risk groups, antiplatelet versus control

Number (%) vascular events


Number of trials



Percent odds reduction

Benefit/1000 patients

Mean months of treatment

Previous myocardial infarction 12 1345/9984
25 36 27
Acute myocardial infarction 15 1007/9658
30 38 1.0
Previous stroke, TIA 21 2045/11493
22 36 29
Other high risk 140 1638/20359
26 22 22
All high risk 188 6035/51494
Acute stroke 7 1670/20418
11 9 0.7
All trials 195 7705/71912

Table 2 breaks down these into more detail, showing the particular vascular events avoided, together with all-cause death avoided. Benefit has to be balanced against risk of harm, the major risk being major extracranial bleeds. Treatment of acute myocardial infarction with antiplatelet for one month entailed considerable benefit in terms of vascular death or non-fatal reinfarction over one month. Treatment of previous stroke for three years avoided additional non-fatal strokes, but at the cost of some extracranial bleeds.

Table 2: Benefit or harm per 1000 patients in particular high risk groups, antiplatelet versus control

Benefit or harm/1000 patients


Mean treatment months

Non-fatal reinfarction

Non-fatal stroke

Vascular death

Any death

Fatal and non-fatal major extracranial bleeds

Previous myocardial infarction 24 18 5 14 12 0
Acute myocardial infarction 1 13 2 23 24 0
Previous stroke, TIA 36 6 25 7 15 5
Acute stroke 0.7 no data 4 5 5 4
Other high risk no data no data no data no data no data 9

Particular antiplatelets

Dose and target patients complicate indirect comparison. The main results for the main antiplatelets are shown in Table 3. Again, there is a reasonably consistent pattern of event rates and degree of odds reduction. Aspirin 75-150 mg a day appears to be as effective as any other dose, and no other antiplatelet had spectacularly better effect.

Table 3: Antiplatelet versus control, particular drugs and doses



Percent vascular events


Drug and dose (mg)

Number of trials



Percent odds reduction

Aspirin 500-1500 34 14.5 17.2 19
Aspirin 160-325 19 11.5 14.8 26
Aspirin 75-150 12 10.9 15.2 32
Aspirin <75 3 17.3 19.4 13
Any aspirin 65 12.9 16.0 23
Dipyridamole 15 14.5 16.8 16
Dipyridamole + aspirin 46 10.7 14.3 30
Ticlopidine 42 8.1 11.1 32

Antiplatelet trials have often used aspirin as a comparison, though none of the comparisons has demonstrated spectacularly better results than aspirin alone (Table 4).

Table 4: Antiplatelet regiments versus aspirin alone



Percent vascular events


Antiplatelet or combination

Number of trials

Antiplatelet regimen


Percent odds reduction

Aspirin + dipyridamole 25 11.8 12.4 6
Dipyridamole 3 16.7 16.5 -2
Ticlopidine 4 21.1 23.2 12
Clopidogrel 1 10.1 11.1 10


This paper is a 'must-read' for those delivering antiplatelet therapy. And not just the abstract. Although it is a meta-analysis, and people are often mightily turned off with the prospect of wading though one of those, this paper is easily understandable by a reader without special knowledge.

There's something Shakespearean about it, with the main theme that antiplatelets work, but with a myriad subplots, from loading doses of aspirin after a heart attack, to the reasoning of a useful effect of antiplatelets in diabetics, even though the evidence for an effect in diabetes is weak. So whether your interest is atrial fibrillation in your granny, or cardiac valve surgery in a neighbour, you could answer the question about antiplatelets.

But one word of caution. The pooling of different doses is a potential problem. For instance, out of 65 comparisons of aspirin against control, only 12 (with 10% of the patients) were of the 75-150 mg doses commonly used in the UK. The problem is only averted by the consistency of results over different doses of different antiplatelets in many different conditions. That won't be the case in other meta-analyses, and in those, and here for very specific questions, other analyses will need to be done.

Aspirin or clopidogrel?

This question has also been the subject of review [2]. The bottom line was that clopidogrel and ticlopidine prevent a few more vascular events than aspirin (NNT 92), at a cost of somewhat more adverse events. The implications for costs are significant. Treating 1000 patients with clopidogrel rather than aspirin could prevent as many as 19 vascular events at a cost of £57,000 ($85,000) per event prevented, or as few as two events at a cost of £533,000 ($800,000) per event prevented. The evidence is abstracted in more detail here.


  1. Antithrombotic Trialists' Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002 324: 71-86.
  2. GJ Hankey et al. Thienopyridines or aspirin to prevent stroke and other serious vascular events in patients at high risk of vascular disease? A systematic review of the evidence from randomized trials. Stroke 2000 31: 1779-1784.
previous or next story in this issue