Medical management of osteoarthritisBMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7266.936 (Published 14 October 2000) Cite this as: BMJ 2000;321:936
- Karen Walker-Bone, clinical research fellow,
- Kassim Javaid, senior house officer in rheumatology,
- Nigel Arden, senior lecturer in rheumatology,
- Cyrus Cooper, professor of rheumatology ()
- MRC Environmental Epidemiology Unit, University of Southampton, Southampton General Hospital, Southampton SO16 6YD
- Correspondence to: C Cooper
- Accepted 25 May 2000
Osteoarthritis is a common, chronic, musculoskeletal disorder. Symptomatic osteoarthritis, particularly of the knee and hip, is the most common cause of musculoskeletal disability in elderly people. In the Western world it ranks fourth in health impact among women and eighth among men.1 Given this high prevalence, therapeutic approaches to treatment will have to be shared between primary and secondary care. A range of non-surgical interventions has been proposed as components of such a therapeutic strategy.
Osteoarthritis is a major cause of pain and disability in Western populations
The prevalence of osteoarthritis necessitates a “shared care” approach to management between general practitioners and hospital specialists
Several non-surgical interventions to alleviate pain and disability in lower limb osteoarthritis are now available:
Non-pharmacological measures (education, social support, physiotherapy, and occupational therapy)
Pharmacological measures (simple analgesics, non-steroidal anti-inflammatory drugs, COX-2 inhibitors, topical non-steroidal anti-inflammatory drugs, and capsaicin)
Intra-articular therapy: corticosteroids, hyaluronic acid derivatives, and tidal irrigation
These interventions have been evaluated to varying degrees, but they can be incorporated into an algorithm for the management of osteoarthritis
Therapeutic options in osteoarthritis
Education (patient and spouse or family)
Social support (telephone contact)
Physiotherapy (aerobic exercises, muscle strengthening, and patellar strapping)
Occupational therapy (aids and appliances, joint protection)
Transcutaneous electrical nerve stimulation (TENS)
Non-steroidal anti-inflammatory drugs
COX-2 inhibitors (cyclo-oxygenase-2 selective non-steroidal anti-inflammatory drugs)
Topical (non-steroidal anti-inflammatory drugs, capsaicin)
Systematic reviews and controlled clinical trials were located through Medline and BIDS 1991-9, searching under the key words: osteoarthritis; guidelines; glucosamine; capsaicin; physiotherapy, occupational therapy, acupuncture, drug therapy, education, intra-articular injection, heat, cold, rehabilitation, epidemiology, therapy. When available, the most recent reviews or meta-analyses are cited; if not available, individual controlled trials were included and methodological shortcomings discussed. We did not perform assessments of quality of individual reviews. Semiquantitative estimates of effectiveness (percentage improvement in pain or function in active group less percentage improvement in control group) were calculated for individual studies. Our focus was to bring together a diverse literature on an important clinical problem and offer a pragmatic approach to patient care.
Patient education (table 1)
A meta-analysis of 10 trials that contrasted patient education with the therapeutic effects of non-steroidal anti-inflammatory drugs confirmed a significant beneficial effect of education on joint pain but not on disability.8 The method was only around 20% as effective as non-steroidal anti-inflammatory drugs, but there was some evidence for a synergistic effect of both interventions. Any member of the care team may provide education in several forms (for example, literature, audiocassette, computer); available packages explain the disease and its management, emphasising the role of weight reduction and exercise. There is now strong evidence that formal patient education should form part of the management of osteoarthritis.
Social support (table 1)
In patients with osteoarthritis of the knee controlled studies have shown that regular telephone contact from a healthcare worker produces significant improvement in pain and functional status.9 Furthermore, education of family members can improve their ability to provide social support, which also benefits the patient.6 These specific effects complement the generally observed improvements in wellbeing and reduced use of health care associated with social support networks.
Physical therapy is a mainstay of the treatment of osteoarthritis. Two main approaches are used by physiotherapists: muscle strengthening programmes specific for certain joints and general aerobic conditioning (table 2). Both of these regimens have been clearly shown to improve pain and disability in osteoarthritis of the knee.15 A single study has shown that the technique of medial taping in patellofemoral osteoarthritis reduces pain.16 Physical measures such as diathermy and ultrasound have limited value.15 In contrast, three trials of trancutaneous electrical nerve stimulation (TENS) suggest modest pain relief when compared with placebo stimulation.15 It is estimated that osteoarthritis constitutes 50% of the workload of traditional acupuncturists. The use of acupuncture is supported by case series and uncontrolled studies, but trials that have compared random needling with acupuncture have failed to show measurable benefit for true acupuncture.17 Ideally, all newly diagnosed patients with osteoarthritis of the hip or knee should be seen by a physiotherapist.
Although occupational therapy provides a means of educating patients and social support, there are few evaluations of specific interventions such as the provision of walking aids, orthoses, and splints in controlled studies. In a single trial of patients with osteoarthritis of the hand the combination of a hand exercise programme, provision of splintage, and non-steroidal anti-inflammatory drugs improved disability in 49% of treated patients, but the study could not dissect the relative benefits of occupational therapy from those of the drugs.18 Despite the absence of formal controlled trials of many occupational therapeutic interventions, there is ample historical and anecdotal evidence of their effectiveness in clinical practice. The precise indications for referral to occupational therapy in patients with osteoarthritis remain to be delineated.
Analgesics, non-steroidal anti-inflammatory drugs, and cyclo-oxygenase-2 (COX-2) inhibitors
The evidence supporting use of analgesics and non-steroidal anti-inflammatory drugs in osteoarthritis was recently reviewed.19 Paracetamol is safe and effective. There is a slight benefit from the addition of dextropropoxyphene, but this is counterbalanced by the broader range of adverse effects. Several short term studies (under six months) have shown that non-steroidal anti-inflammatory drugs are more effective than placebo in reducing pain and improving function, but there have been few studies that have lasted longer than two years.20 Inference from these is difficult as adherence rates with non-steroidal anti-inflammatory drugs are poor because of adverse effects, while those with paracetamol are poor because of suboptimal pain relief. As many as 20-30% of all admissions to hospital and deaths from peptic ulcer disease in elderly people may be related to use of non-steroidal anti-inflammatory drugs.21 There is evidence that misoprostol and proton pump inhibitors reduce the risk of serious upper gastrointestinal injury induced by non-steroidal anti-inflammatory drugs. Adjunctive use of H2 blockers has been shown only to reduce the incidence of duodenal ulceration. The cost utility of prophylactic use of any of these agents, however, is controversial.20 It is recommended that non-steroidal anti-inflammatory drugs are initiated only after consideration of side effects and counselling of the patient; the prescription should be reviewed every six months.
Relative contraindications to starting treatment with non-steroidal anti-inflammatory drugs
Exercise caution in:
Those aged >65 years
Patients with a history of peptic ulcer disease
Concomitant treatment with corticosteroids and anticoagulants
Patients with cardiovascular disease
Heavy alcohol drinkers
Exercise caution in:
Those aged >65 years
Patients with hypertension
Patients with congestive cardiac failure
Concomitant medication with angiotensin converting enzyme inhibitors and diuretics
Cyclo-oxygenase, an enzyme involved in the conversion of arachidonic acid to prostaglandins, exists in two isoforms: COX-1, a constitutive isoform, predominates in the stomach and produces cytoprotective prostaglandins, while COX-2, an inducible isoform predominantly involved in the inflammatory cascade, gives rise to articular pain, swelling, and stiffness. Novel therapeutic agents have been developed that act as specific inhibitors of the cyclo-oxygenase-2 isoform (COX-2 inhibitors). Although published data on these drugs remain scarce,22 trials have shown similar efficacy to that of non-steroidal anti-inflammatory drugs in the treatment of osteoarthritis but with gastrointestinal toxicity comparable with that of placebo. The early evidence points towards greater safety than non-steroidal anti-inflammatory drugs alone, but comparisons are not yet available with combinations of non-steroidal anti-inflammatory drugs and cytoprotective therapy. A meta-analysis of all eight double blind, randomised, controlled trials of rofecoxib, published after completion of our literature search, suggested a relative risk of 0.51 for serious clinical upper gastrointestinal events among users of the COX-2 inhibitors when compared with conventional non-steroidal anti-inflammatory drugs.23 These drugs may be widely prescribed for patients with osteoarthritis (and are licensed for this indication), but the most cost effective strategy for their use is far from clear; non-selective non-steroidal anti-inflammatory drugs with or without co-prescription of cytoprotection may still have a role.
Topical treatment is an additional option for patients with osteoarthritis who have inadequate pain relief or who cannot tolerate systemic therapy. The two best evaluated topical agents in the treatment of the disorder are non-steroidal anti-inflammatory drugs and capsaicin. A recent meta-analysis concluded that 65% of patients allocated to active treatment with topical non-steroidal anti-inflammatory drugs had a good response compared with only 30% of patients receiving placebo.24 Although the component trials were often small and of variable quality there is reasonably strong evidence to conclude that topical non-steroidal anti-inflammatory drugs are effective and safe for patients with osteoarthritis.
Capsaicin is a naturally occurring compound that reversibly depletes the stores of the neurotransmitter substance P from sensory nerve endings; it thereby attenuates the transmission of painful stimuli from the peripheral nerve fibres to higher centres. A meta-analysis of the three placebo controlled trials of capsaicin in osteoarthritis reported that the agent is well tolerated and has significantly greater analgesic effects than placebo.25
Intra-articular corticosteroids are widely used in the management of patients with osteoarthritis of the knee, most commonly in those who have appreciable effusion or other signs of active inflammation. Several small randomised controlled trials confirm superior short term efficacy to intra-articular placebo in this setting,20 the additional benefits lasting two to four weeks. There are, however, important and maintained responses to the intra-articular placebo injections and arthrocentesis incorporated in these studies, such that the group treated with corticosteroids often show sustained benefit over baseline for several months. There is good evidence to support the judicious use of intra-articular corticosteroids in patients with knee osteoarthritis, but because of the potential for multiple intra-articular injections to accelerate cartilage damage, they should not comprise the only treatment of patients with chronic, stable osteoarthritis.
Hyaluronic acid (table 3)
Hyaluronic acid is a linear polysaccharide found naturally in synovial fluid, where it is thought to facilitate shock absorption and lubrication. In people with osteoarthritis there is a reduced concentration of hyaluronic acid, resulting in low viscosity synovial fluid and an increase in cartilage loading.35 Several preparations of hyaluronan are currently available for the treatment of osteoarthritis, the main differences being in their molecular weight and regimens for administration. The results of most randomised controlled trials suggest superior pain relief to placebo and equivalent relief to corticosteroid injections but with a greater duration of action. 32 36 37 The high molecular weight preparations seem to produce greater benefit than the low molecular weight preparations, although this observation needs confirmation in a parallel group randomised controlled trial. A substantial proportion of patients (up to 20%) experience a joint flare after injection, which, although transient, may cause considerable discomfort.
Although there is reasonable evidence that intra-articular hyaluronans are effective in patients with knee osteoarthritis, more information is required before their exact place in the management algorithm can be defined. The cost effectiveness of their use in clinical practice needs to be examined; the profile of patients most likely to benefit and the optimal regimen for repeat treatment courses need to be defined. Studies are in progress to clarify these issues.
Irrigation of the knee joint with saline by using a wide bore needle emerged as a potential treatment for osteoarthritis after clinical reports supported the value of arthroscopic lavage. A single controlled trial has shown considerable improvement after this procedure when compared with standard medical management.38 A second trial that compared the use of tidal irrigation with formal arthroscopic lavage suggested similar improvements in pain and function at three months, but the presence of a meniscal tear predicted a better response to arthroscopic intervention.39
All the pharmacological interventions described hitherto aim to relieve pain and thereby improve function in osteoarthritis. To date, no measures have been convincingly shown to modify the rate of structural change in cartilage or subchondral bone, which constitute the underlying disease process. Several putative chondroprotective agents or ones that may modify structure have been proposed, including chondroitin and glucosamine compounds, other glycosaminoglycan derivatives found in mammalian articular cartilage, and tetracycline. Clinical trials provide some justification for the use of chondroitin and glucosamine preparations but only for their analgesic or anti-inflammatory effects.40 The issue as to whether glucosamine sulphate is capable of attenuating cartilage loss in patients with early knee osteoarthritis is currently under investigation.
Osteoarthritis is a major cause of pain and disability in the general population. Currently, most patients with osteoarthritis are managed in primary care. The therapeutic options for managing osteoarthritis have expanded considerably in recent years, although most currently available treatments are palliative. A detailed recent review of non-surgical methods41 found that education, exercise, systemic analgesics, non-steroidal anti-inflammatory drugs, and topical agents were likely to be beneficial; the review questioned the value of intra-articular treatment.
The management of patients with osteoarthritis should ideally be multidisciplinary and include both education and physiotherapy. Patients do best if they are empowered in their own management. The figure outlines a possible management schedule for patients with knee osteoarthritis, although this must be tailored to fit the individual patient and will vary for different joint sites. The future holds promise for drugs that may genuinely modify structure, but these will require careful evaluation so that they may be appropriately positioned in the management algorithm. Finally, the publication of evidence based clinical guidelines42 for the management of this disorder is urgently awaited.
We are grateful to Professors Maxime Dougados, Michael Doherty, and Paul Dieppe for their critical evaluation of this review. KW-B is in receipt of an ARC Clinical Research Fellowship. The manuscript was prepared by Mrs Gill Strange.
Competing interests None declared.