This site is intended for health professionals only

At the heart of general practice since 1960

pul jul aug2020 cover 80x101px
Read the latest issue online

Independents' Day

Race linked to stroke survival

With the options for treatment

fast disappearing, Dr Iain Gilchrist updates advice

for GPs

In the past few months, following the withdrawal of rofecoxib (Vioxx), there has been a deluge of bad news for GPs and their patients suffering from arthritis. Initial hopes that the problems were isolated to rofecoxib alone, and not common to other cox-2 drugs, proved shortlived. More recently, researchers have turned their attention to the older, traditional NSAIDs, such as ibuprofen, naproxen and diclofenac.

Far from these drugs being given a clean bill of health with regard to causing cardiovascular events, it

appears they may be just as liable as celecoxib to give rise to cardiac events, although rofecoxib still appears to have a higher risk profile than other drugs.

The news that the compound analgesic, co-proxamol, was also to be withdrawn because of its potentially fatal toxicity in small overdose, provided another problem for GPs and their patients (see box 1).

These events pose several questions:

·Why has it taken so long for evidence of cardiovascular side-effects of cox-2s and NSAIDs to emerge?

·What options are available for treating arthritic pain?

·What should we be advising patients who are taking anti-inflammatories?

Other than to comment that a review of the method of continuing surveillance of medications seems long overdue, and the Yellow Card system seems to have long outlived its usefulness, I shall be focusing on the latter two points.

The latest advice on cox-2s from the European Medicines Agency was published on June 27 this year and is shown in the box 2.

And just this week the EMEA, prompted to review non-selective NSAIDs by two studies raising safety concerns, concluded no changes were necessary.

What should we advise patients taking anti-inflammatories?

The first question to answer is whether they need to be taking anti-inflammatories, or whether their pain will be controlled by other measures.

Due consideration should be given to non-pharmacological treatment of arthritic pain. As an example, patients should be encouraged to exercise regularly to help maintain muscle strength.

This seems to be particularly beneficial in OA of the knee. A recent study3 has also shown benefit from weight loss in reducing symptoms in patients with knee OA symptoms. Both these measures will also be of benefit to the cardiovascular health of the patient. More use could also be made of physiotherapy.

Unfortunately, in many areas there is such a long wait for NHS treatment that this is often not a feasible option, unless the patient elects to be treated privately.

Using local injections

Local injections, either of steroids or hyaluronans, may be beneficial in some circumstances. While the main use of steroid injections is in inflammatory joint disease, it has been shown they provide short-term pain relief for up to one month compared with placebo in patients with OA.

Similarly, it has been shown that a series of hyaluronan injections can reduce pain, and improve function, in patients with OA for up to six months.

While local treatments may have their own associated side-effects, there should be no deleterious effect on the cardiovascular or gastrointestinal systems.

Many GPs undertake joint and soft tissue injections in their surgeries. The techniques are taught on some courses, such as the Diploma in Primary Care Rheumatology at the University of Bath.

Alternatively, many local rheumatologists are happy to instruct interested GPs, to avoid the long waiting times involved before patients can be seen at hospital.

Pressure could also be applied to the PCT to set up a GPwSI service, if there is sufficient demand to provide a quick and locally-based service.

Adjunctive therapies

Due consideration should be given to adjunctive therapies.


The beneficial effects of glucosamine sulphate have been increasingly recognised lately, both with regard to pain relief over a three-year period, and with reduced joint space narrowing seen on X-ray.

The effective dose used in trials in OA knee has been 1,500mg daily.

Although not classified as a drug, it can be prescribed on an FP10, although it would seem prudent to ask your PCT to include it on its drug formulary.

Side-effects are at placebo level.

Fish oil

This is a popular remedy for many patients. Although there is some evidence of its effectiveness in rheumatoid arthritis, there is little or no evidence of its effectiveness in OA. Many GPs, however, are very happy to keep an open mind on its effectiveness when asked by patients about its use.


As a traditional remedy for 'rheumatics', many patients have found benefit from local applications of creams and rubs. Capsaicin cream has been found to be more effective than placebo cream. It is used as a 0.025% cream to treat OA, and is generally more expensive than local NSAIDs, which have also been shown to be effective.

Treatment options for

arthritic pain

·Paracetamol This should be the mainstay of analgesic therapy. Many

patients will state that they have tried paracetamol as an OTC therapy, and many people will say it has proved ineffective. But many patients will have not taken this drug in the full analgesic dose of 1g four times daily, and this dose should be tried before moving to the next stage of the analgesic ladder.

·Compound analgesics The next step up is the use of compounds of paracetamol with a weak opioid, such as codeine or dihydrocodeine. These compounds increase the risk of side-effects, such as constipation, drowsiness, and other CNS side-effects, which limits their usefulness.

·Strong opioids The risk of side-effects is increased, and the risk of addiction also need to be considered.

·Anti-inflammatories In the minds of some people, it seems unduly restrictive to have an absolute contraindication to the use of cox-2s in patients with established disease, especially as there is accumulating evidence that traditional NSAIDs may be no safer.

There may well be a role, although

much smaller than before, for the use of these drugs in patients with severe inflammatory joint disease, such as rheumatoid arthritis, ankylosing spondylitis and gout, despite the existence of concurrent vascular disease.

These patients will require careful counselling, and doubtless some patients will decide to cease taking anti-inflammatories.

The dilemma is what to do if the patient makes an informed choice to continue with the anti-inflammatory, despite the presence of a contraindication.

To prescribe, or not to prescribe, is the question that will face the doctor.

Each case will need to be decided on an individual basis between doctor and fully-informed patient (see the two GP dilemmas below).

Iain Gilchrist is a member of the Primary Care Rheumatology Society and a GP in Hatfield Heath, Essex

1. A turbulent year for arthritis analgesia

September 2004

Rofecoxib withdrawn from market


CSM advises patients on a cox-2 who have ischaemic heart or cerebrovascular disease be switched to alternative treatment as soon as possible

Summary data on two clinical trials for celecoxib indicate increased cardiovascular risk; National Prescribing Centre advises GPs not to switch patients on rofecoxib to another cox-2 because of growing safety concerns

January 2005

Committee on Safety of Medicines announces gradual withdrawal of co-proxamol


European Medicines Agency begins review of cardiovascular safety of conventional NSAIDs

MHRA suspends use of valdecoxib


Three studies from the UK, US and Canada suggest traditional NSAIDs pose significant risks

EMEA issues judgment on

cox-2s restricting their use


The EMEA concludes its review of non-selective NSAIDs and recommends no changes are necessary

2. Latest advice on cox-2 drugs from the European Medicines Agency

·Cox-2 drugs must not be used in patients with established IHD, stroke, and peripheral vascular disease

·Caution should be exercised in patients with risk factors for heart disease

·Use the lowest possible dose for the shortest period of time

·Beware of sensitivity reactions and severe and potentially fatal skin reactions, especially in the first month of use

GP dilemma 1

A 60-year-old male, with a 30-year history of rheumatoid arthritis, treated with DMARDs and indometacin for many years. Co-morbidities of hiatus hernia for 20 years, heart failure for 15 years, thought to be due to aortic valve disease and cardiomyopathy, rather than IHD. Co-prescription of PPI, ACE inhibitor and diuretics.

Indometacin stopped three years ago because of onset of anaemia, unable to tolerate symptoms without NSAID, prescribed celecoxib, but unable to tolerate this drug because of dizziness. Prescribed rofecoxib, effective and tolerated well. Had aortic valve replacement one year ago, then rofecoxib withdrawn. Patient would be unable to work if not prescribed anti-inflammatory, prescribed meloxicam, which he found less effective than rofecoxib, but gave sufficient pain relief, and tolerated well.

In this case, there is a clear need for this man to continue on effective anti-inflammatory medication. He requires gastro-protection because of his previous anaemia when on indometacin and his hiatus hernia, despite an otherwise normal endoscopy. He does not require aspirin prophylaxis, as he has no evidence of IHD, having had a normal coronary arteriogram.

A 55-year-old man has a 10-year history of severe OA of his hands. Also 10-year history of angina, treated with angioplasty nine years ago. No gastrointestinal disorder. Treated with simvastatin for hypercholesterolaemia, aspirin, and diclofenac. Patient had previously stopped atenolol because of side-effects, and he found paracetamol ineffective in controlling the pain and stiffness. He found this more of a problem with his hobbies of golf and sailing, rather than his occupation as a medical negligence lawyer.

At recent review, we agreed he would try a paracetamol/ codeine combination as baseline analgesia, with use of diclofenac on a PRN basis for bad, or very active, days.

Diclofenac is the best choice of anti-inflammatory as it

does not interfere with the action of aspirin. If he had required gastro-protection as well, the addition of a PPI would be best in these circumstances, rather than prescribing a cox-2, and losing the cardioprotective effect of aspirin. The situation is less complex when the patient has risk factors for, but no established, vascular disease. It would seem prudent to screen all patients on anti-inflammatories, and those for whom we propose to prescribe these drugs on a possible long-term basis, for vascular risk factors, and attempting to modify them if found. At the very least, this will involve taking a history, especially with regard to smoking, and family history of vascular disease, checking the blood pressure, and arranging for lipid and glucose levels to be determined. While much of this work can be delegated to the practice nursing team (if any spare capacity), it will still involve the GP with extra workload involved in scrutinising the results, and counselling patients.

Most GP computer systems can calculate a cardiovascular risk score, based on data from the Framingham study. These automatically calculate risk for some factors, for example a diabetic patient will have their risk increased by 50 per cent. Most GPs have become accustomed to altering the risk calculation manually in other patients, eg those with a family history of early IHD, where the risk is also increased by about 50 per cent. In light of the accumulating evidence of the increased cardiovascular risk of anti-inflammatory drugs, it would seem reasonable to calculate the risk score for patients taking these drugs on a regular basis, adjusting the score to take account of the extra risk, and making clinical decisions, such as the prescription of statins, on the basis of the adjusted score.

This is analogous to prescribing an H2 blocker or PPI to reduce risk of gastro-intestinal side effects of NSAIDs in at-risk patients.


1 Johnsen SP et al. Arch Intern Med, 2005;165:978

2 Hippisley-Cox J, Coupland C. BMJ, 2005; 330:1366

3 Christensen R et al. Osteoarthritis and Cartilage, 2005;13:20

Useful websites



Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say