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

The price of aspirin

Preventing gastrointestinal events
Low-dose aspirin costs
And aspirin resistance?

Were you to overhear conversations in any coal mining community, at some time you would come across a discussion about the price of coal. What you were listening to was no discussion about economics, but rather an oblique reference to the hidden costs of injuries and deaths resulting from coal extraction. In medicine the hidden costs are the costs of the adverse events that almost inevitably accompany attempts to generate benefit.

The balance of benefit and risk with aspirin prophylaxis was examined in Bandolier 86, where the main risk examined was upper gastrointestinal bleeding. There are other risks, though, mainly renal failure and congestive heart failure.

Preventing gastrointestinal events

Upper gastrointestinal bleeding can be reduced by concomitant use of proton pump inhibitors [1]. One hundred and twenty-three patients who had an ulcer complication after using low dose aspirin for more than one month, and who had Helicobacter pylori infection had their ulcers healed with triple therapy using PPI and antibiotics for one week followed by 20 mg famotidine twice a day for a further five weeks. Those patients whose Helicobacter infection had been cured were randomised to 100 mg aspirin and 30 mg lansoprazole or 100 mg aspirin and matching placebo, for 12 months, with follow up every two months. Patients were told to avoid taking NSAIDs. The primary end point was the recurrence of ulcer complications of bleeding, perforation or obstruction.


Patients were 71 years old on average, and 70% were men. During the 12 months of the study, 14 upper gastrointestinal tract events were adjudicated by an end-points committee, and 10 were confirmed. One occurred in the 62 patients taking lansoprazole (2%) and nine in the 61 patients taking placebo (15%) (Figure 1). The relative risk was 0.1 (95% CI 0.01 to 0.8) and the number needed to treat with 30 mg lansoprazole for one year to prevent one upper gastrointestinal bleed was 8 (95% CI 4 to 27).

Figure 1: Upper GI events with lansoprazole and placebo

Low-dose aspirin costs

The costs of low dose aspirin have been explored in a massive study from Tayside [2]. In this population of about 400,000 people there is a well-established record-linking scheme, so that primary care prescriptions can be linked to medical history, other prescriptions, and hospital admissions and diagnostic interventions.

People who were dispensed one or more prescriptions for low dose aspirin (defined as not higher than 325 mg per day) at any time over the six years of 1990-1995 were included. For each subject there were 10 age and sex matched controls. Only new users were included.

The study outcomes were:


Of the 17,244 subjects taking aspirin, 77% were aged 60 or older. There was an average of 2.53 years of observation per patient, during which aspirin was taken for 1.18 years. The implication was that compliance (concordance) was 47%.

Aspirin users had more renal and upper gastrointestinal adverse events (Table 1). The average annual cost was £18 per person for renal adverse events and £25 for gastrointestinal adverse events, compared with an average £2 for the aspirin and £5.50 for dispensing. These annual average costs were obtained by division of the total cost by the number of years of follow up, not by years of aspirin exposure.

Table 1: Drug costs and costs of renal and upper gastrointestinal adverse events with low dose aspirin

Over an average follow up of 2.5 years per patient
Aspirin users
Excess cost
Excess cost/year
Drug costs
Dispensing cost
Total drug cost
Renal costs
Renal admission
Total renal cost
Gastrointestinal costs
GI admission
Anti-ulcer drugs
Total GI cost
Total cost
Acutal exposure was 1.18 years of the average 2.53 years between the first prescription and end of the study

A second analysis was limited only to low risk subjects free of disease at study entry (30% of total), and adjusted for different risk factors between aspirin users and non-users. This produced a lower cost estimate of average annual cost of £2.90 per person for renal adverse events and £2.66 for gastrointestinal adverse events, and a total of £13 per patient per year.


This cost paper [2] is a brilliant read. It produces cogent arguments why the full analysis (called a pragmatic analysis in the paper) is likely to be the most sensible. It makes the point that it did not include costs involved with congestive heart failure, nor did it include any costs with any bleeding events not in the upper bowel. It also considers sources of bias, like use of non-prescribed aspirin or NSAIDs, and possible confounding by indication.

The projected annual costs, multiplied pro-rata for Scotland would be £3 to £11 million, and therefore for the UK would be ten times greater. And, of course, this is for newly treated patients, and with costs at 1996/7 prices.

It also examined costs per event prevented, using information from the largest meta-analysis (Bandolier 108). Using the benefit of 16 events prevented per 1,000 people per year after a heart attack, the cost of one vascular event prevented would be £3,330. This used the pragmatic analysis and average costs over 2.5 years, not the 1.2 years of actual exposure, which would double the figure. At lower levels of cardiovascular risk the cost per event prevented rises substantially.

This paper is a must-read for those who provide guidance on prescribing in primary and secondary care. There is masses of good discussion that has to be read and understood, so that prescribing makes sense for individuals and for payers. It might also help to provide advice for the many older people who have several conditions with competing quality of life issues, to help decide what is best for them.

And aspirin resistance?

Just to confuse things further, there is now an acceptance that in some people aspirin may not prevent platelet aggregation. In those where this failure occurs, risks of vascular events are higher. It is too early to know just where this is leading, but for those who need to know more, a review [3] is a useful starting place.


  1. KC Lai et al. Lansoprazole for the prevention of recurrences of ulcer complications from long-term low-dose aspirin use. New England Journal of Medicine 2002 346: 2033-2038.
  2. SV Morant et al. Cardiovascular prophylaxis with aspirin: costs of supply and management of upper gastrointestinal and renal toxicity. British Journal of Clinical Pharmacology 2003 57: 188-198.
  3. R Altman et al. The antithrombotic profile of aspirin. Aspirin resistance, or simply failure? Thrombosis Journal 2004 2:1 (

previous or next story