Logo of brjclinpharmLink to Publisher's site
Br J Clin Pharmacol. 2004 Feb; 57(2): 188–198.
PMCID: PMC1884435

Cardiovascular prophylaxis with aspirin: costs of supply and management of upper gastrointestinal and renal toxicity



To determine the cost to the NHS of prescribed low-dose aspirin.


This was a population based observational cohort study. Patients from Tayside Scotland (17 244 new users of dispensed aspirin each with 10 matched comparators) were included. A pragmatic analysis totalled costs from the start to end of the study and compared these with a matched cohort of aspirin nonusers to estimate excess costs. Fastidious analyses were done of subjects with no prior history of upper gastrointestinal (UGI) or renal disease where the cost that occurred during aspirin exposure, the 30 days following aspirin exposure and subsequent nonexposure was calculated adjusting for risk factors in each period.


Subjects took aspirin for only 1.18 of the 2.53 years follow-up (47% compliance). Aspirin use cost an additional £49.86 per year (pragmatic analysis) made up of £1.96 for aspirin tablets (4%), £5.49 for dispensing costs (11%), £24.60 for UGI complications (49%) and £17.81 for renal complications (36%). The costs for managing complications were substantially lower in the fastidious analysis (£2.66 for UGI complications and £2.92 for renal complications). Assuming that the antiplatelet trial meta-analysis is an accurate assessment of the benefits of aspirin, the costs of preventing one vascular event lay between £62 500 (primary prevention, pragmatic analysis) and £867 (secondary prevention, fastidious analysis). These costs may be underestimates due to the low compliance observed.


Compliance with aspirin was poor. Serious adverse events were uncommon but despite this aspirin cost the NHS between 6 and 25 times the cost of aspirin tablets due to dispensing costs and the cost of managing adverse effects.

Keywords: adverse effects, aspirin, compliance, cost


Low-dose aspirin is commonly prescribed for prophylaxis of coronary and cerebrovascular disease in subjects with [1, 2] and without [35] established vascular disease. Aspirin is an example of a treatment that has inexpensive prescription ingredient costs, but may incur significant additional costs due to adverse effects such as upper gastrointestinal haemorrhage (UGI) [6, 7] or renal impairment [8]. No studies of the true cost of aspirin prescribed within an NHS setting have been performed. We have done a study to examine the range of possible aspirin associated costs. We did both a pragmatic analysis to estimate the upper limit and a fastidious analysis to estimate the lower limit of events and hence costs.


The aim of this study was to determine the costs to the NHS of community prescribed low dose aspirin (up to 325 mg day−1) including the costs of the associated gastrointestinal and renal adverse events.

The study used the Medicines Monitoring Unit (MEMO) record Linkage database which has been described elsewhere [9]. In brief, MEMO receives all dispensed prescriptions for Tayside from the Practitioner Services Division of the Information and Statistics Division of the Common Services Agency of the National Health Service in Scotland. These prescriptions are entered onto a database along with a unique patient identifier which is used on all medical records in Tayside. MEMO also receives copies of all hospital discharge diagnostic and procedure data called Scottish Morbidity Records, copies of laboratory data, copies of the General Registrar Office death certification data and a variety of other healthcare data. These data also contain the unique patient identifier and so dispensed prescribing records can be record-linked with other healthcare data. MEMO studies use data that have been acceptably anonymized. Studies are approved by the Tayside committee on research medical ethics and the operating procedures of MEMO have been approved by the Tayside Caldicott Guardians.

Study population

The study population consisted of a fixed population of residents of Tayside who were registered with a general practitioner (GP) between both on 1st January 1989 and 31st December 1995, or who died in the intervening period. (Subjects who died were censored at the date of death).

Study subjects

Subjects included were those dispensed one or more prescriptions for low-dose aspirin between 1st January 1990 and 31st December 1995. Patients who were dispensed doses of aspirin higher than 325 mg day−1 at any time during the study period were excluded.

Screening period

In order to identify new users of aspirin, subjects who were dispensed aspirin prescribed during 1989 were excluded from the study.

Exposure classification

The date, number of tablets, preparation and treatment regimen recorded on each aspirin prescription were used to determine the cost and the period of exposure associated with each prescription. New prescriptions dated within 7 days of the end of these exposure periods were regarded as providing uninterrupted aspirin exposure. Those dated more than 7 days after the end of the previous exposure period defined the start of a new period of exposure.

Study cohorts

For each subject in the aspirin cohort each period of aspirin exposure was matched by date to 10 patients of the same age (within a decade) and sex among the remaining 354 933 patients in the study population. The first 30 days of nonexposure and subsequent periods of nonexposure were also matched by date in the same 10 comparator patients. This matching process was done to provide comparator cost data in an unexposed cohort taking into account the possible temporal changes in health care provision.

Study outcomes

The following were defined as outcome events:

  1. Hospitalization with a primary diagnosis of renal failure. We assumed that 10% of patients with such admissions required dialysis. A sensitivity analyses was done assuming that only 1% of patients got dialysis.
  2. Creatinine ≥150 mmol ml−1. It was assumed that 10% would go on to require dialysis. A sensitivity analysis was also done assuming only 10% of creatinine ≥500 mmol ml−1 required dialysis.
  3. Hospitalization with a primary diagnosis of an upper gastrointestinal event.
  4. An endoscopy.
  5. H2-receptor antagonist prescription.
  6. Proton pump inhibitor prescription.
  7. Misoprostol prescription.

Risk factors

The patient-level risk factors considered were age, sex and the Carstair's social deprivation score [10].

The following time-dependent risk factors were determined on the first day of each period of aspirin exposure or nonexposure for patients in the aspirin cohort, and on the same dates in the comparator patients: a prior history of each of the outcomes; the use of the following noncardiovascular medications: NSAIDs, H2-receptor antagonists, proton pump inhibitors, misoprostol; the use of the following cardiovascular drugs, loop diuretics, nitrates and ACE inhibitors.

Statistical methods

Pragmatic analysis

A pragmatic analysis that totalled all the costs of aspirin and anti-ulcer drugs (including dispensing charges), the costs of hospitalizations for GI and renal events and the costs of dialysis was performed for each aspirin user, starting from the first use of aspirin until the end of the study period. The same costs were calculated for the comparator patients associated with each aspirin user over the same periods of time. This type of analysis summed the cost of recurrent events and was thought to represent the type of analysis that would produce an upper limit of aspirin attributable costs.

In addition to matching on sex, age and date, patients were further matched on risk factors that, in preliminary analyses, were significant for the outcome being analysed. This was done because no suitable statistical model could be found that would allow parametric adjustments to be made. Within the 10 comparator patients for each aspirin user, the additional matching criteria shown in Table 1 were applied.

Table 1
Additional matching criteria for exposed subjects and comparators in the pragmatic analyses.

Fastidious analysis

A fastidious analysis was also done. This analysis examined only low risk subjects free from disease at entry and it recalculated risk factors prior to each period of aspirin exposure. This type of analysis was thought to produce a lower estimate to the aspirin attributable events and hence costs. In this analysis, the numbers of days of aspirin exposure and the numbers of outcome events in those periods of exposure were aggregated for every combination of risk factors. Days and outcome events in same periods of time in the comparator patients were also aggregated for every combination of risk factors. The numbers of outcome events were modelled using Poisson regression with the logarithm of the exposure time as an offset variable. Similar analyses were carried out, separately, for periods of recent exposure and for subsequent periods of nonexposure.

Analysis of renal outcomes was restricted to patients with no history of renal events, and analyses of UGI outcomes, including endoscopies and the use of anti-ulcer drugs, were restricted to patients with no history of UGI events. All of the other potential risk factors were included in the statistical models. Those that were not statistically significant (P > 0.05) in any of the three analyses (for current, recent and nonexposure) for an outcome were eliminated from the models for that outcome. For each outcome relative risks and their confidence intervals were estimated for aspirin use and for each risk factor that was statistically significant in at least one of the three analyses.

The numbers of outcome events attributable to aspirin were estimated from these models, the costs of the attributable-events were calculated.

Cost data

Hospitalization costs were taken from the Scottish Health Service Costs for 1996/97 [11]. The total cost per day for each speciality in each hospital was used. The costs of endoscopy were obtained from Tayside Universities Hospital Trust.

The yearly costs of renal dialysis used were an upper value of €35 000 [12] (approximately £26 000 by 1996/7 exchange rates) to a lower value of between £11 000 [13] and £17 000 [14]. The variation in price was driven by methodological issues and the scope of what was included in the cost exercise. In addition, it is recognized that survival is reduced in these patients [15]. After discussion with renal physicians we used the upper limit as a reasonable estimate of total cost. However, we also provide analyses using the lower cost of £11 000/year.

Drug costs for each preparation dispensed were taken from the British National Formulary March 1996.

The dispensing cost used was £1.58, a composite cost of professional fees and cost of processing prescriptions as published by the Information and Statistics Division of the Common Services Agency in Scotland [16].


The study used acceptably anonymized data using methodology approved by the Tayside Caldicott Guardians.


The aspirin cohort contained 17 244 subjects, 77% of who were aged 60 years or greater.

Pragmatic analysis

The results of the pragmatic analysis are summarized in Table 2. There was an average of 2.53 years of observation per patient during which there was 1.18 years of exposure to aspirin, 0.20 years of recent exposure and 1.15 years of nonexposure. Thus new users of aspirin actually took prescribed aspirin for only 46.6% of the study period. The actual cost of aspirin prescriptions for the 1.18 years of use was an average of £4.95 per subject or £4.20 per subject per year. When divided by the number of patient years of observation, the cost was £1.96 per year.

Table 2
Results of the pragmatic analyses. Costs (£) incurred by aspirin users and comparators between first dose of aspirin and the end of the study period (an average of 2.53 years/patient).

The total cost incurred by aspirin users, in terms of aspirin treatment and the management of renal and upper gastrointestinal problems, during the study was £126.15 per patient greater than the costs incurred for nonusers or £49.86 per patient per year. Of this yearly excess cost, aspirin ingredients cost £1.96 (4%), dispensing costs were £5.49 (11%), GI hospitalizations, endoscopies and prescriptions for anti-ulcer drugs accounted for £24.60 (49%) and renal events and dialysis accounted for £17.81 (36%).

Fastidious analysis

Hospitalizations with upper gastrointestinal diagnoses

Thirteen hospitalizations with upper gastrointestinal diagnoses (UGI events) occurred during the 13 thousand patient years exposure to aspirin in this study, a rate of 0.96 events per thousand patient years (Table 3). In the comparator periods corresponding to this aspirin exposure (154 thousand patient years) 70 events occurred, an event rate of 0.45 per thousand patient years. The difference in event rate between the aspirin and comparator cohorts suggests that an excess of approximately seven events occurred during aspirin exposure, out of the total of 13 events in this cohort. However, risk factors differed between the aspirin and comparator cohorts. When these were taken into account, the risk of a UGI event during aspirin exposure was estimated to be 2.02 times greater than in the comparator cohort, suggesting that approximately six of the UGI events (46% of the total number) were attributable to aspirin.

Table 3
Results of the fastidious analyses. Hospitalizations with upper gastrointestinal (UGI) diagnoses, number of endoscopies, prescriptions for H2-receptor antagonists, proton pump inhibitors and misoprostol, hospitalizations for renal failure and creatinine ...

Data from similar analyses done for endoscopy, prescriptions for H2-receptor antagonists, proton pump inhibitors, misoprostol, hospitalizations with renal failure, and creatinine concentrations higher than150 µmol ml−1 are also given in Table 3.

Costs attributable to aspirin

The estimated cost attributable to the use of aspirin in this study obtained from the fastidious analysis is given in Table 4. This was £32.97 per patient during the study or £13.03 per patient per year. Of this yearly cost, aspirin tablets cost £1.96 (15%), dispensing charges cost £5.49 (42%), UGI hospitalizations, endoscopies and ulcer healing drugs cost £2.66 (21%) and renal complications and dialysis cost £2.92 (22%) per patient per year.

Table 4
Fastidious analysis: Costs attributable to aspirin exposure

Sensitivity analyses are shown in Table 4 for renal costs assuming that 10% or 1% of renal admissions resulted in dialysis and assuming that 100% or 10% of subjects with creatinine >500 µmol ml−1 resulted in dialysis. The higher estimates were used in the total cost calculations but these lower cost assumptions could reduce renal costs attributable to aspirin by about £5 per subject in total.

Costs attributable to aspirin during aspirin exposure

The fastidious analysis calculated the excess cost of aspirin during aspirin exposure to be £24.39. This excess cost of aspirin was incurred wholly during aspirin exposure and since there was only 1.18 years of exposure during the study, the cost per year of aspirin exposure was £20.67 per subject per year of exposure.


Aspirin exposure

Aspirin is the most commonly prescribed drug in Scotland with about 2 million prescriptions or 544 438 person year's exposure per year in Scotland in 2002 [16]. The number of prescriptions of antiplatelet therapy has been increasing and between 1998 and 2000 the number of prescriptions dispensed increased by about 20% per year [16].

In 1998 a maximum of 153 448 000 aspirin 75 mg tablets were sold over the counter without prescription (OTC) in the UK [17]. This equates to 420 400 patient years of exposure in the UK (if all subjects took one tablet per day: some will have taken two). Assuming that Scotland consumed 10% of the UK total then 42 040 patient years of exposure will have occurred in Scotland. Tayside has about 7.3% of the Scottish population so about 3070 patient years exposure might have occurred from OTC 75 mg aspirin in 1998 among the Tayside population. On the other hand, there was an estimated 43 555 patient years of exposure to prescribed aspirin in Tayside in 2002 [16].

Our study cohort of 17 244 new aspirin users were exposed to 20 347 patient years of prescribed aspirin during the study period. It seems unlikely that these subjects who were prescribed aspirin, most of whom were over 60 years old and exempt from prescription charges, would also buy OTC aspirin.

The population of Tayside (about 400 000), from which the cohort and 10 matched controls was drawn will have consumed an absolute maximum of 3363 patient years of aspirin OTC (aspirin use has been increasing: prior to 1998 less OTC aspirin 75 mg was sold). It seems very unlikely that the 10 control subjects will all have been taking OTC aspirin. However, even if all of the OTC aspirin sold in Tayside were taken in the control group of our study there would be a maximum of an 8% misclassification of exposure. Therefore we doubt if OTC 75 mg aspirin use would have altered the conclusions of our study.

As for aspirin as an analgesic, OTC drugs are sold for intermittent and infrequent use. In Tayside 1124 person years of exposure to analgesic doses of aspirin were sold OTC in 1998 to a population of 400 000 [17]. We doubt if this is a major source of bias. Even if this were a bias, then such a bias would be likely to drive the difference between exposed and unexposed subjects towards the null and thus underestimate the costs of prescribed aspirin.


We have shown that in an unselected population of 17 244 newly treated patients, the excess costs/year associated with use of prescribed aspirin lay between £13.03 (fastidious analysis) and £49.86 (pragmatic analysis) per patient per year of treatment, irrespective of whether or not aspirin was continued after the first prescription. Indeed, the average duration of aspirin use during the study was 1.18 years per person whilst the average duration of follow-up of subjects in the study was 2.53 years per person. Thus, for only 47% of the study time were subjects actually exposed to aspirin suggesting poor compliance with aspirin.

Aspirin-attributable serious adverse events were uncommon in the fastidious analysis of this study. However, this analysis was confined to low risk subjects with no history of events. The pragmatic analysis included those subjects with a history of events who have much increased event rates. Thus for the 17 244 newly treated subjects included in the analyses the excess costs per year of aspirin treatment plus those of UGI and renal complications in Tayside lay between £224 689 (fastidious analysis) and £859 786 (pragmatic analysis). Multiplied pro-rata to Scotland, these figures would be between £3 million and £11 million per year for newly treated patients. Thus the cost of aspirin therapy in Scotland is considerable.

In the pragmatic analysis patients were matched for age, sex and other risk factors for the outcome being analysed at the time of first dose of aspirin. During the average 2.53 years study period for each patient their risk factors may have changed substantially, and they may have changed differentially for the aspirin users and nonusers. If risks increased more rapidly in aspirin users than in nonusers, for reasons other than their use of aspirin, the pragmatic analysis may overestimate the costs attributable to aspirin.

The fastidious analysis adjusted event rates for risk factors present at the start of each period of exposure or nonexposure. Thus, risks that had changed due to previous exposure to aspirin were adjusted for, and this analysis therefore estimated the marginal effects of aspirin exposure in each period. This provided a rigorous test of the hypothesis that aspirin increases the risk of renal and UGI events during and after each course of treatment, but because it estimated the marginal effects in each exposure period it will have substantially underestimated the total attributable costs. In addition this analysis examined only those subjects at low risk for adverse events and costs differences were calculated despite poor compliance with aspirin. The marginal excess cost of aspirin during aspirin exposure was also calculated at £20.67 per subject per year. This may represent a more meaningful lower estimate of excess cost but this is a matter for debate.

The pragmatic and fastidious analyses provide reasonable upper and lower bounds for the costs attributable to aspirin. The true cost is likely to be closer to the upper bound, provided by the pragmatic analysis, than to the lower bound, which estimated only the marginal costs attributable to each course of treatment.

Sensitivity analyses

Sensitivity analyses of differing estimates of the costs of dialyses were done as there was some uncertainty as to the proportion that would undergo this procedure and wide variation in cost estimates. As shown in Table 4, varying these indices widely made only a few pounds difference to the total cost and thus the outcome of this study was not sensitive to these changes.

The pragmatic analysis estimated the attributable-costs at £126 per patient for the study period, and the fastidious analysis estimated cost at £33 per patient. Approximately £19 of cost, in both cases, was the cost of prescribing aspirin, so the additional costs were approximately £107 and £14 by the two analyses. Of these additional costs, £30 and £6, respectively, were incurred during aspirin exposed periods and £77 and £8 during subsequent nonexposed periods. Even the fastidious analysis revealed a marginal carry-over effect of aspirin following each period of exposure with associated costs similar to those incurred during exposure. The pragmatic analysis indicated that the cumulative costs attributable to aspirin after exposure ceased were more than twice those incurred during exposure.

A recent small study from Manchester has examined the costs of aspirin plus the cost of antidyspeptic medication [18]. These authors estimated the cost of aspirin to be 4.7p per day or £17.16 per year. Clearly this study did not take into account many of the costs considered in the present study. This lends weight to the argument that the true costs of aspirin therapy are nearer the pragmatic analysis results than the fastidious estimate.

Confounding by indication

Subjects who were prescribed aspirin got this drug for secondary or primary prevention of cardiovascular disease. The presence of cardiovascular disease increases the rate of NSAID-associated upper gastrointestinal complications [19, 20]. Subjects with cardiovascular disease are also more likely to have hypertensive renal disease or atherosclerotic renovascular disease. Thus subjects with cardiovascular disease who are prescribed aspirin are likely to be at increased risk for both renal and upper gastrointestinal complications that are not directly attributable to aspirin and are confounded by the underlying conditions. We adjusted for known risk factors in order to try and control for such confounding but it is likely that residual confounding remained. The fastidious analysis may be less confounded because risk factors were adjusted for at each exposure and nonexposure period. However, at least some of the aspirin-attributable costs and carry-over effects are likely to be due to residual confounding. For example, the risk ratios for endoscopies, H2-receptor antagonists, proton pump inhibitors, and creatinine >150 mmol l−1 (Table 3) are all statistically significantly elevated in the aspirin cohort when off treatment compared with the comparator cohort and this strongly suggests the presence of confounding. In addition, a recent case-control study from Sweden has found evidence that aspirin may at least be a contributing factor to renal dysfunction [21].

However, the data on upper GI bleeding from major unconfounded trials of antiplatelet therapy vs placebo (http://bmj.com/cgi/content/full/324/7329/71/DC1/3) show that there were about twice as many bleeds with antiplatelet therapy as with placebo in subjects with a variety of cardiovascular conditions (415 events vs 211 over a wide range of conditions). We found an event rate of 0.96 events per thousand patient years in the aspirin exposed cohort vs 0.46 events per thousand patient years in the comparator group which is almost exactly the same proportion as in unconfounded trials. For this reason we think that it is unlikely that the present data are so significantly confounded. In view of this, it seems likely that our findings may be reasonable estimates of aspirin-induced renal costs.

Other costs

We have studied costs due to gastrointestinal and renal complications. As with other NSAIDs, aspirin may increase hospitalizations for other conditions such as colitis [22], congestive heart failure [23], hypertension [24, 25], cerebral haemorrhage [1] and epistaxis [26]. We have not included these costs in this study, however, but they could be substantial. For example, Page & Henry have suggested that 19% of all hospitalizations for congestive heart failure might be attributable to NSAIDs and low dose aspirin may play a part in this toxicity [23].

Prescribing costs

The average cost of aspirin tablets was modest at £4.20 per year or £4.95 for the average exposure of 1.18 years. However, the cost of dispensing aspirin was £1.58 per prescription [16] or £11.77 per year or £13.88 for the average duration of exposure in the study. Since 77% of subjects were over the age of 60 years and since some of the remaining 27% will have been exempt from prescription charges, the majority of these costs were borne by the NHS.

Benefits of aspirin

The benefits of aspirin have been estimated by the Antiplatelet Trialists' Collaboration [1] and the Antithrombotic Triallists Collaboration [2]. In the most recent analysis [2], antiplatelet therapy given for the secondary prevention of MI, prevented 36 vascular events (myocardial infarction, stroke or vascular death) per thousand subjects treated for 27 months, or approximately 16 events per year. For subjects treated for secondary prevention of stroke or transient ischaemic attacks, antiplatelet treatment prevented 36 vascular events per thousand over 29 months or about 15 events per year of treatment. For low risk or primary prevention [1], a net of four vascular events (five nonfatal myocardial infarctions, two nonfatal strokes and one vascular death) per thousand subjects were prevented over 5 years of treatment or 0.8 events per year. However another recent analysis has suggested that the benefits of aspirin in low-risk subjects may be more marginal and that there is no effect of aspirin on mortality [27].

The above benefits of aspirin have been derived from the results of randomized controlled trials. It is likely that the compliance with aspirin (or antiplatelet) treatment in these studies was good or at least better than the compliance experienced in ‘real world’ prescribing. Compliance with aspirin does appear to predict the cardiovascular benefit [28]. The costs of aspirin in the present study were calculated over a period of time during which subjects were taking aspirin less than one day in two. It is thus likely that the benefits of aspirin therapy will have been similarly reduced and this should be borne in mind when assessing the cost-effectiveness of aspirin therapy. The fastidious analysis provided costs for aspirin therapy during aspirin exposure. It is notable that this cost per year of exposure is about £21 which is considerably more than the average cost per year in the study of £13 (Table 4).

The precise aspirin-induced gastrointestinal and renal risk profile of each of the above treatment groups will vary, with secondary prevention probably having higher risks than primary prevention (for example, gastrointestinal haemorrhage is more common in subjects with established cardiovascular disease). However, as an approximation, the excess cost of treating subjects with aspirin and paying for the gastrointestinal and renal events associated with aspirin in 1000 subjects over 1 year will range from £13 000 (fastidious analysis) to £50 000 (pragmatic analysis). Thus for secondary prevention of myocardial infarction or stroke (and assuming that the benefits will be the same as those seen in randomized trials) the excess cost per vascular event prevented lay between £867 and £3330. For primary prevention the cost lay between £16 250 to £62 500 per event prevented. However, if cardiovascular events prevented are linearly related to aspirin compliance, then the cost per event prevented could be much higher. For example, if the excess costs during aspirin exposure are used (rather than the average excess costs in the study), then the cost-effectiveness calculations for aspirin in the fastidious analysis are £1723 per vascular event prevented for secondary prevention and £25 838 for primary prevention.

A study that examined the cost effectiveness of aspirin and clopidogrel for secondary prevention in the USA healthcare setting has recently been published [29]. They found that extending the use of aspirin from current levels to all eligible subjects for 25 years would cost $11 000 per quality adjusted life year (QALY) gained whereas using clopidogrel alone would cost $130 000 per QALY. These data show that the cost-effectiveness of prophylaxis is sensitive to the cost of the drug used.

For primary prevention in the general practice setting, the best estimate of risk indicates that aspirin will significantly reduce total cardiovascular events (cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, angina, transient ischaemic attack, peripheral artery disease or revascularization procedures) by 20 events per thousand subjects treated for 4 years or 5 events per 1000 per year at a cost of between £2600 and £10 000 per event prevented [5].

There have recently been calls to give aspirin to subjects at even lower cardiovascular risk than has been recent practice, perhaps as low as 3% 5 year coronary risk [27]. However there have also been cogent arguments against this [30]. Clearly the cost per event prevented in very low risk subjects will be much higher. It has also been argued that the benefits of aspirin may be overestimated or even absent [31, 32]. If this is the case then the cost-effectiveness figures given above may be optimistic. Thus there are economic arguments for restricting aspirin to those who have higher cardiovascular risk.

One interesting finding from the present study is the high cost of dispensing aspirin. In view of the low base cost of the aspirin tablets, it would appear to be more cost-effective to dispense much longer duration prescriptions at least for subjects established on long-term prophylactic aspirin. If prescriptions were given for 1-year duration, this would reduce the overall cost of dispensing costs from £11.77 to £1.58. The cost of the aspirin itself would remain unchanged. Alternatively, aspirin could be supplied by a pharmacist by a mechanism that did not involve a prescription or the OTC purchase of aspirin could be encouraged. A recent study suggested that such OTC use could be increased further [33].

There have recently been calls to introduce a ‘Polypill’ that would have aspirin as one of its components [34]. Such a combination therapy would reduce the cost of supply of aspirin but it would still incur the same adverse effects. Thus the use of aspirin in a combination product would not be without cost.

The strengths of the present study are that it was carried out in a population-based setting and included all subjects treated. In reality, some subjects are likely to have existed within this population who should have received aspirin prophylaxis but did not [35]. The matching process employed in our study may have enriched the comparator population with such subjects and this will have reduced the confounding by indication associated with aspirin use. On the other hand, subjects who were aspirin-intolerant or who had suffered a prior gastrointestinal event that was not recorded on the database may have been included in the comparator population. Such subjects are more likely to experience a further upper gastrointestinal event and this may have minimized the differences in costs between the cohorts.

The weaknesses of this study lie in its observational (nonrandomized) nature. Thus subjects who received aspirin may have upper gastrointestinal and renal event rates that are different from those who did not receive aspirin for reasons other than the aspirin itself. However, when assessing such rare adverse events, observational studies are likely to remain the major and most practical data sources for the foreseeable future [36]. We may also have underestimated the true costs of aspirin's toxicity by confining our study to gastrointestinal and renal outcomes. Finally, subjects taking regular aspirin purchased over-the-counter were not identified although we doubt that this will have affected our findings.


Generic low-dose aspirin tablets are cheap. Serious adverse events were uncommon. However, the cost to the NHS of prescribed aspirin prophylaxis was significant adding between £13 and £50 per patient per year or about 6–25 times the cost of aspirin tablets to the cost of management of new users. These increased costs were due to dispensing costs and the cost of managing adverse effects and were incurred despite poor compliance with treatment. The cost per event prevented for secondary prevention and for high-risk primary prevention suggest that aspirin use is good value. However, its use for primary prevention in subjects at low cardiovascular risk is less clear.

A full report that incorporates detail tabulations and analyses is available from ku.ca.eednud@dlanodcam.m.t.


Dr Steve V. Morant did the statistical analyses and assisted in drafting the report. Professor John G. F. Cleland assisted in drafting the protocol and contributed to the report. Professor Peter G. Davey assisted in drafting the protocol and contributed to the pharmacoeconomic analyses and the report. Dr Alex D. McMahon assisted in drafting the protocol and the report. Professor Thomas M. MacDonald assisted in drafting the protocol, assisted with analyses, drafted the report and is guarantor of the study. ISD Scotland provided SMR1 data and CHI data. This study was funded by the Chief Scientists Office of the NHS in Scotland grant number K/OPR/2/2/D352.


1. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet therapy. 1: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Br Med J. 1994;308:81–106. [PMC free article] [PubMed]
2. Antithrombotic Trialists' Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. Br Med J. 2002;324:71–86. [PMC free article] [PubMed]
3. The Medical Research Council's General Practice Research Framework. Thrombosis prevention trial. randomised trial of low-intensity oral anticoagulation with warfarin and low-dose aspirin in the primary prevention of ischaemic heart disease in men and women. Lancet. 1998;351:233–41. [PubMed]
4. Hansson L, Zanchetti A, Carruthers SG, et al. Effects of intensive blood-pressure lowering and aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial. Lancet. 1998;351:1755–62. [PubMed]
5. Collaborative Group of the Primary Prevention Project (PPP) Low-dose aspirin and vitamin E in people at cardiovascular risk: a randomised trial in general practice. Lancet. 2001;357:89–95. [PubMed]
6. Weil J, Colin-Jones D, Langman M, Lawson D, Logan R, Murphy M, Rawlins M, Vessey M, Wainwright P. Prophylactic aspirin and risk of peptic ulcer bleeding. Br Med J. 1995;310:827–30. [PMC free article] [PubMed]
7. Derry S, Loke YK. Risk of gastrointestinal haemorrhage with long term use of aspirin: meta-analysis. Br Med J. 2000. pp. 1183–7. [PMC free article] [PubMed]
8. Evans JM, McGregor E, McMahon AD, McGilchrist MM, White G, Jones M, McDevitt DG, MacDonald TM. Non-steroidal anti-inflammatory drugs and hospitalization for acute renal failure. Q J Med. 1995;88:551–7. [PubMed]
9. Evans JMM, MacDonald TM. The Tayside Medicines Monitoring Unit (MEMO) In: Strom BL, editor. Pharmacoepidemiology. 3. Chichester: John Wiley; 2000. pp. 361–74. Chapter 22.
10. McLoone P. Carstairs Codes for Scottish Postcode Sectors from the 1991 Census. Glasgow: University of Glasgow, Public Health Research Unit; 1991.
11. Information and Statistics Division National Health Service in Scotland. Scottish Health Service Costs. Edinburgh: ISD Publications; 1997.
12. Renal purchasing guidelines. London: NHS Executive; 1996.
13. Beech R, Mandalia S, Melia J, Mays N, Swan A. Purchasing services for end stage renal failure: the potential and limitations of existing information sources. Health Trends. 1993;25:60–4. [PubMed]
14. Chandna SM, Schulz J, Lawrence C, Greenwood RN, Farrington K. Is there a rationale for rationing chronic dialysis? A hospital based cohort study of factors affecting survival and morbidity. Br Med J. 1999;318:217–23. [PMC free article] [PubMed]
15. MacLeoad A, Grant A, Donaldson C, Khan I, Campbell M, Daly C, Lawrence S, Wallace S, Vale L, Cody J, Fitzhugh K, Montague G, Ritchie C. Effectiveness and efficiency of methods of dialysis therapy for end-stage renal disease: systematic reviews. Health Technol Assessment. 1998;2:1–140. [PubMed]
17. Sheen CL, Dillon JF, Bateman DN, Simpson K, MacDonald TM. Paracetamol pack size restriction. the impact on paracetamol poisoning and over-the-counter supply of paracetamol, aspirin and ibuprofen. Pharmacoepidemiol Drug Safety. 2002;11:329–31. [PubMed]
18. Drummond A, Kwok S, Morgan J, Durrington PN. Cost of aspirin and statins in general practice. Q J Med. 2002;95:23–6. [PubMed]
19. Simon LS, Hatoum HT, Bittman RM, Archambault WT, Polisson RP. Risk factors for serious nonsteroidal-induced gastrointestinal complications. regression analysis of the MUCOSA trial. Fam Med. 1996;8:204–10. [PubMed]
20. Weil J, Langman MJ, Wainwright P, Lawson DH, Rawlins M, Logan RF, Brown TP, Vessey MP, Murphy M, Colin-Jones DG. Peptic ulcer bleeding: accessory risk factors and interactions with non-steroidal anti-inflammatory drugs. Gut. 2000;46:27–31. [PMC free article] [PubMed]
21. Fored CM, Ejerblad E, Lindblad P, Fryzek JP, Dickman PW, Signorello LB, Lipworth L, Elinder C-G, Blot WJ, McLaughlin JK, Zack MM, Nyren O. Acetaminophen, aspirin, and chronic renal failure. N Engl J Med. 2001;345:1801–8. [PubMed]
22. Evans JMM, McMahon AD, Murray FE, McDevitt DG, MacDonald TM. Non-steroidal anti-inflammatory drugs are associated with emergency hospitalisation for colitis due to inflammatory bowel disease. Gut. 1997;40:619–22. [PMC free article] [PubMed]
23. Page J, Henry D. Consumption of NSAIDs and the development of congestive heart failure in elderly patients. An under-recognized public health problem. Arch Intern Med. 2000;160:777–84. [PubMed]
24. Pope JE, Anderson JJ, Felson DT. A meta-analysis of the effects of nonsteroidal anti-inflammatory drugs on blood pressure. Arch Intern Med. 1993;153:477–84. [PubMed]
25. Johnston AG, Ngyuyen TV, Day RO. Do non-steroidal anti-inflammatory drugs affect blood pressure? A meta-analysis. Ann Intern Med. 1994;121:289–300. [PubMed]
26. Tay HL, Evans JMM, McMahon AD, MacDonald TM. Aspirin, non-steroidal anti-inflammatory drugs and epistaxis: a regional record linkage case control study. Ann Otolaryngol Rhinol. 1998;107:671–4. [PubMed]
27. Hayden M, Pignone M, Phillips C, Mulrow C. Aspirin for the primary prevention of cardiovascular events: a summary of the evidence for the U.S. preventive services task force. Ann Intern Med. 2002;136:161–72. [PubMed]
28. Glynn RJ, Buring JE, Manson JE, LaMotte F, Hennekens CH. Sdherence to aspirin in the prevention of myocardial infarction. The physicians' health study. Arch Intern Med. 1994;154:2649–57. [PubMed]
29. Gaspoz J-M, Coxson PG, Goldman PA, et al. Cost effectiveness of aspirin, clopidogrel or both for secondary prevention of coronary heart disease. N Engl J Med. 2002;346:1800–6. [PubMed]
30. Lauer MS. Aspirin for primary prevention of coronary events. N Engl J Med. 2002;346:1468–74. [PubMed]
31. Cleland JGF For debate: Preventing atherosclerotic events with aspirin. Br Med J. 2002;324:103–5. [PMC free article] [PubMed]
32. Cleland JGF. Is aspirin ‘The weakest link’ in cardiovascular prophylaxis. Prog Cardiovasc Dis. 2002;44:275–92. [PubMed]
33. Bedson J, Whitehurst T, Lewis M, Croft P. Factors affecting over-the-counter use of aspirin in the secondary prophylaxis of cardiovascular disease. Br J Gen Pract. 2001;51:1001–3. [PMC free article] [PubMed]
34. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80% Br Med J. 2003;326:1419–24. [PMC free article] [PubMed]
35. EUROASPIRE I and II Group. Clinical reality of coronary prevention guidelines: a comparison of EUROASPIRE I and II in nine countries. Lancet. 2001;357:995–1001. [PubMed]
36. McMahon AD, MacDonald TM. Design issues for drug epidemiology. Br J Clin Pharmacol. 2000;50:1–7. [PMC free article] [PubMed]

Articles from British Journal of Clinical Pharmacology are provided here courtesy of British Pharmacological Society
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    NCBI Bookshelf books that cite the current articles.
  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...