Intended for healthcare professionals


Non-steroidal anti-inflammatory drugs and elderly patients

BMJ 1995; 310 doi: (Published 01 April 1995) Cite this as: BMJ 1995;310:817
  1. D N Bateman,
  2. J G Kennedy
  1. Reader in therapeutics Lecturer in primary care therapeutics Northern and Yorkshire Regional Drug and Therapeutics Centre, Wolfson Unit of Clinical Pharmacology, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 4LP

    The medicine may be worse than the disease

    Aches and pains are a feature of growing old, but managing these symptoms in elderly people can pose problems. Patients have often used paracetamol and aspirin, which are available over the counter without prescription, before consultation. Opiates cause constipation and non-steroidal anti-inflammatory drugs cause ulcers. Although these associations are well known, 20 million prescriptions for non-steroidal anti-inflammatory drugs (many for elderly people) were dispensed in 1993 in Britain, at a cost of over pounds sterling180m.1

    Several factors affect the relative risk of adverse gastrointestinal effects during treatment with non-steroidal anti-inflammatory drugs. These include the age of the patient2 3 4 5 6 7 8; their medical history3 4 6 7; the drug used and its dosage4 9; its route of administration10 and the duration of treatment3 4 8; the concomitant use of more than one such drug4; and other independent risk factors, particularly alcohol use, anti-coagulant treatment, corticosteroid treatment, and smoking.4

    The risk of upper gastrointestinal bleeding and ulceration with non-steroidal anti-inflammatory drugs clearly increases with age: Laporte et al calculated an estimated annual incidence of upper gastrointestinal bleeding of 210 per million people over 60 compared with 35 per million for people under 60.7 This effect of age is particularly important as the background incidence of upper gastrointestinal bleeding in the general population also rises steeply with age, and even a moderate rise in relative risk in elderly users of non-steroidal anti-inflammatory drugs therefore represents an important increased hazard for them. Elderly patients are also at increased risk of the renal and cardiac adverse effects of non-steroidal anti-inflammatory drugs owing to these drugs' higher background incidence of impaired renal and cardiac function.

    Some non-steroidal anti-inflammatory drugs carry a higher relative risk of gastrointestinal ulceration than others,4 6 7 8 and the risk of peptic ulceration is 10 times greater with azapropazone than with ibuprofen. The daily dose of non-steroidal anti-inflammatory drugs has a direct linear relation with the risk of gastrointestinal complications for subjects of any age.4 Elderly patients commonly receive non-steroidal anti-inflammatory drugs on repeat prescriptions,12 and, although the relative risk of gastrointestinal problems is highest during the initial period of treatment,4 8 11 long term users also develop gastrointestinal adverse effects, which indicates that the risk associated with the drugs does not disappear with prolonged exposure. Garcia Rodriguez and Jick estimated that patients who had recently changed from one non-steroidal anti-inflammatory drug to another or who had received more than one non-steroidal anti-inflammatory drug simultaneously had more than twice the risk of patients exposed to only one such drug.4 No elderly patient should receive two non-steroidal anti-inflammatory drugs at the same time, and if a change in treatment is clinically indicated careful supervision is required.

    Changing the route of administration is not necessarily the answer. Henry et al showed that suppositories carried a higher relative risk than tablets.10 Increased risk was attributed to the prescribing of rectal forms of non-steroidal anti-inflammatory drugs for patients at high risk in the mistaken belief that rectal forms were safer than oral forms. This finding supports the hypothesis that non-steroidal anti-inflammatory drugs exert at least part of their adverse effects on the gastrointestinal tract after systemic absorption. No controlled evaluations of injectable non-steroidal anti-inflammatory drugs have been reported, but this form is unlikely to be less toxic than oral forms. Topical non-steroidal anti-inflammatory drugs are believed to be safer owing to their reduced systemic absorption but are clinically less effective and have limited applications.

    Ten million people aged 60 and older live in England and Wales. Langman et al estimated that in this population 10000 episodes of ulcer bleeding and 2000 ulcer perforations occur each year.8 This yields an overall annual risk of bleeding of an ulcer of 1 in 1000 and of perforation of 1 in 5000; about a quarter of these episodes are associated with the use of non-steroidal anti-inflammatory drugs.8 If the drugs in current use were replaced by those with the lowest relative risk then the number of episodes of upper gastrointestinal adverse events associated with the drugs would be halved. The incidence of adverse effects could be cut further if the lowest effective doses of non-steroidal anti-inflammatory drug were prescribed.8

    The strategy of minimising the use of non-steroidal anti-inflammatory drugs through rational prescribing contrasts with the implications of coprescribing antiulcer drugs. Two hundred deaths attributable to ulcers caused by non-steroidal anti-inflammatory drugs occur in Britain each year.11 If the 20 million prescriptions for these drugs issued each year in Britain are each assumed to be for, on average, one month's treatment then the cost of coprescribing misoprostol or ranitidine would range from pounds sterling200m (misoprostol 400 μg daily) to pounds sterling600m (ranitidine 150 mg twice daily). Even if these drugs were totally effective in preventing deaths from non-steroidal anti-inflammatory drugs (which they are not) the cost per life saved would be pounds sterling1m to pounds sterling3m.

    What are the implications for managing elderly patients with chronic pain? Although pain in younger patients often has an inflammatory component, the contribution of acute inflammation to pain in elderly patients is unclear. Chronic pain is likely to be due principally to degenerative changes, and the inflammatory component is less important. The correct initial treatment for such patients is regular paracetamol or a standard compound analgesic such as co-codamol. The reluctance of some patients to take such treatment regularly is one factor that makes longer acting, but more toxic, non-steroidal anti-inflammatory drugs appear clinically more effective. In clinical trials non-steroidal anti-inflammatory drugs seem to offer additional benefit to only a very small proportion of patients with chronic pain.13 14 15 If non-steroidal anti-inflammatory drugs are indicated they should be prescribed initially for short courses at the minimum effective dose, and low dose ibuprofen should be the first choice.

    Publicity of the adverse effects of non-steroidal anti-inflammatory drugs may be starting to influence prescribing: prescribing data from primary care in England and Wales show little or no growth in the prescription of these drugs since 1992, although overall prescribing rates are rising at over 4% a year.1 Further reducing and rationalising the prescription of non-steroidal anti-inflammatory drugs to elderly patients is an important public health target for primary care as we approach the millennium.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.

    Log in

    Log in through your institution


    * For online subscription