Aspirin is one of the most widely used drugs worldwide, including the UK, where it is estimated that 50% of people aged 75 and use it.1 It is an old drug, with well-established indications, well-known side effects, and is cheap and widely available to the public.

This familiarity and cost-effectiveness may be the reason why, beyond the requirements of the Quality and Outcomes Framework (QOF) stipulating its use in secondary prevention of cardiovascular disease, it does not frequently feature in practice-based medicines audit. For this reason, we have decided to focus on potentially unnecessary prescribing for primary prevention of cardiovascular disease.

As recently as a decade-and-a-half ago, there was a significant body of professional opinion advocating aspirin’s use in primary prevention for all high risk people, including notionally healthy adults aged 50 or over,2 the sell being ‘it’s good for the heart’ or ‘it thins the blood’. More recently, however, that body of opinion has swung significantly against any use in primary prevention, where risks outweigh benefits.3

On the matter of safety, one key study by Peter Rothwell and colleagues on behalf of the Oxford Vascular study, analysing patients appropriately given aspirin for secondary prevention, estimated by extrapolation that aspirin causes 20,000 significant bleeds and 3,000 deaths annually in the UK.4

Another landmark study was ASPREE (Aspirin Reducing Events in the Elderly), in which 19,114 patients over 70 years old and without cardiovascular disease were randomised to aspirin or placebo. At 4.7 years follow-up CVD event rates were the same in both groups; however, the aspirin group had 8.6 major haemorrhage events per 1000 person-years compared with 6.2 events per 1000 person-years in the placebo group.5  

Chorleywood Health Centre is a 4-doctor undergraduate teaching and postgraduate training practice with a list size of 6,900, located in a suburban / semi-rural commuter zone to the North-West of London. The practice has a low turnover of around 10% of the list per annum. Though the area is affluent, there is significant health need due to a sizeable elderly population and the chronic disease workload resulting from this. Having recognised, anecdotally, that several patients continue to take aspirin wedded to advice given in the distant past which is not supported by current evidence, we decided to investigate the scale of unnecessary prescribing of aspirin.


A computer search (EMIS) was undertaken to identify all patients receiving a prescription for aspirin 75mg within a 3-month period of November 2019 to January 2020 inclusive. Patients who had diagnoses of coronary heart disease (CHD), stroke or TIA (CVD) and peripheral vascular disease (PVD), and who were therefore receiving aspirin for secondary prevention were excluded.

This left a cohort of patients with no immediately apparent coded indication. In turn, these patients had their medical records analysed to ascertain if they were indeed taking aspirin, and if so, whether there was evidence for an indication, which in most cases, unless immediately obvious, involved tracing the initiation of aspirin, and relevant information from both primary and secondary care. The patients thus analysed were reclassified into five groups:

  • Those who had stopped aspirin
  • Those whose record gave a clear indication for prescription
  • Equivocal cases, deemed to be those without definite pathological indications, but considering the overall clinical management the prescription seemed reasonable
  • Those with no definite indication who clearly expressed a wish to continue
  • No indication – the group taking the drug unnecessarily

A particular subgroup studied was those patients prescribed aspirin when initial symptoms suggested a high likelihood of a reasonable indication, but who continued taking it even when given the all-clear or an alternative diagnosis after investigation.


The initial computer search generated 346 patients receiving aspirin 75mg on repeat prescription, and flagged up 90 deemed not to have an indication and whose records were analysed in detail. Of these 90, the breakdown into the 5 aforementioned categories is as follows:

Stopped aspirin

Nine patients had stopped, including two obstetric cases who used it due to recurrent miscarriage - a reasonable indication.

Appropriate indication

Five patients had proven CHD, two had proven CVD, nine proven PVD, two had a central retinal vein thrombosis, one had a history of deep venous thrombosis (DVT) attributable to Factor V Leiden heterozygous status, and one had polycythaemia rubra vera; these formed a cohort of 20 patients with appropriate indications for prescription. In each case the medical records were updated to reflect these diagnoses, which were not corrected coded contemporaneously.

Equivocal cases

Two patients were referred to the TIA clinic with an equivocal opinion on cause, but were advised to adopt secondary prevention. One patient had atypical Parkinson’s disease and the neurologist postulated a partly vascular cause. One patient had ischaemic heart disease due to vasospasm rather than significant atheroma, one had mitral regurgitation with no angiographic evidence of significant coronary stenosis but was advised to reduce overall cardiac risk, and one kidney transplant patient was advised to do so by the renal clinic; this subgroup numbered six.   

Patients who expressed a wish to continue

Two patients fell into this category, expressing a wish to continue despite there being no indication to do so, and being warned of the risks.

This cohort collectively comprising those no longer taking aspirin, those found to have a definite or reasonable indication to have it prescribed, and those who exercised their personal choice in the matter collectively numbered 37. This was deducted from the original 90, to leave a cohort of 53 patients of the 346 identified (15.3%) almost certainly taking aspirin unnecessarily. Conversely, it meant that 84.7% had a reasonable indication or were unwilling to stop.

Our investigation into the subgroup who were prescribed aspirin initially on the supposition of cardiovascular disease, and who continued taking it despite subsequent negative investigations discounted comprised a total of seven. These were four with chest pain that was non-cardiac, two who had PVD excluded (one being diagnosed with spinal stenosis with symptoms indistinguishable from intermittent claudication), and one referred to the TIA clinic deemed to have a negligible risk of the event having been a TIA.


At the outset we must acknowledge the limitations of our study, conducted as it was within a small practice, where small absolute numbers may significantly skew percentages. Furthermore, given that aspirin is freely and cheaply available over-the-counter, a study based on searches of prescribed aspirin necessarily underestimates the number of patients actually using this drug.

Nevertheless, this has proven a useful exercise, facilitating the retrospective updating of significant pathology in many patients in whom this was not done at the time of diagnosis, and being reminded of less frequent conditions such as central retinal vein thrombosis and Factor V Leiden Heterozygous which are, nonetheless, sound indications for aspirin. Indeed, almost all the patients analysed were older (average age 74 years), and had cardiovascular risk factors, with many being started on aspirin when it was much more in vogue to do so for both primary and secondary prevention.

Indeed, though the latter is a fairly exact science within the often inexact science of clinical medicine, the reality of the discourse in the consulting room and outpatient clinic is often nuanced, and factors such as personal preference and intuition can be expected to play a part in decisions to prescribe or recommend aspirin, especially in equivocal cases.

Furthermore, evidence that discontinuing aspirin, even when given to low-risk patients in whom it is not definitely indicated increases the risk of thrombosis by an even greater proportion that that by which initiating aspirin reduces it, probably due to a rebound increase in platelet aggregability,6,7,8 forms a sound argument against stopping it in such patients. Though not explicitly articulated in the discharge summaries, this is probably the reason why none of the 7 patients investigated for vascular disease in secondary care and given the all-clear were advised to stop aspirin. Our study is thus more likely to influence future prescribing habits than change the status quo for the current patient cohort.   

Edin Lakasing and Alison E. James are GPs at Chorleywood Health Centre, Hertfordshire

Conflict of interest: None


  1. Elwood P Morgan G White J et al. Aspirin taking in a south Wales County. Br J Cardiol 2011; 18: 238-240
  2. Ellwood P, Morgan G, Brown G, Pickering J. Aspirin for everyone older than 50? BMJ 2005; 330:
  3. O’Sullivan J. Aspirin for the primary prevention of cardiovascular disease in the elderly. BMJ Evidence-Based-Medicine 2019; 24(4):
  4. Linxin L, Geraghty OC, Mehta Z, Rothwell PM. Age-specific risks, severity, time course, and outcome of bleeding an long-term antiplatelet treatment after vascular events: a population-based cohort study. The Lancet 2017; 390: 490-499.
  5. McNeil JJ, Wolfe R, Woods RL, et al. Effect of aspirin on cardiovascular events and bleeding in the healthy elderly. N Engl J Med 2018; 379: 1509–18.
  6. Sever P S. Rapid Response: Discontinuation of Aspirin for Primary Prevention? BMJ 2011; 343:
  7. Biondi-Zoccai GG, Lotrionte M, Agostoni P, Abbate A, Fusaro M,
    Burzotta F, et al. A systematic review and mega-analysis on the hazards
    of discontinuing or not adhering to aspirin among 50,279 patients at risk
    for coronary artery disease. Eur Heart J 2006; 27: 2667-2674.
  8. Lotrionte M, Biondi-Zoccai GG. The hazards of discontinuing
    acetylsalicylic acid therapy in those at risk of coronary artery disease.
    Curr Opin Cardiol 2008; 23: 487-493.