Osteoarthritis

Overview 
Theory 
Diagnosis 
Management 
Follow up 
Resources 

The main goal of treatment in OA is to control joint pain and stiffness, thereby improving function.

Management includes non-pharmacological approaches (patient education, self-management, and exercise programmes), pharmacological therapies, and surgery. Therapy must be individualised; different treatments can be combined.

[Figure caption and citation for the preceding image starts]: Managing osteoarthritis in primary care (key: PPI - proton pump inhibitor)Adapted from NICE CG177 (2014) Osteoarthritis: care and management in adults [Citation ends]. com.bmj.content.model.Caption@11b7ae77

It is important to acknowledge the role of the placebo effect when considering any treatment in OA, as there is evidence that placebo is an effective treatment for people with OA patients.[86][87]

The effect size of different types of placebo varies: for example, intra-articular placebo injections and topical placebo treatment might have a greater effect than oral placebos.[88]

Non-pharmacological approaches

All patients should start treatment for OA with non-pharmacological approaches.[7][73] These include patient education, self-management, exercise programmes (with reassurance that exercise, e.g., resistance training, tai chi, yoga, and water-based exercise, is not harmful to the joints), and they may also benefit from cognitive behavioural therapy in combination with physiotherapy.[7][73][89][90][91][92][93][94][95] [ ]

Physiotherapy and occupational therapy, as well as manual mobilisation, are also recommended.[7][73][96][97][98][99]

Exercise and weight loss programmes

Exercise is recommended for all patients with OA, though there is considerably more evidence for the use of exercise in the treatment of knee and hip OA than for hand OA.[7] Balance exercises or tai chi are recommended for patients with OA of the knee and/or hip, and yoga is suggested as an alternative for patients with OA of the knee.[7]

Weight loss is recommended for patients with OA of the knee and/or hip who are overweight.[7]

Two Cochrane reviews conclude that exercise programmes have a small to moderate beneficial effect on pain and function for patients with knee and hip OA.[100][101] [ ] However, the benefit from physiotherapy on hip OA is unclear.[102] [ ] One meta-analysis showed a modest effect on pain, though no improvement in self-reported function for exercise in patients with OA of the hip.[103] Further meta-analyses reported that 14% more patients with hip OA responded to exercise therapy, compared with placebo, and that hip abductor muscle strengthening exercises significantly improved knee pain and other functional outcomes for patients with knee OA.[104][105]

Evidence from randomised controlled trials (RCTs) suggests that quadricep strengthening exercises and weight loss are effective in controlling the pain of knee OA.[106][107] Subsequent meta-analyses demonstrate that hip strengthening exercises are an effective rehabilitation treatment for patients with OA of the knee.[105][108][109]

Exercise can improve quality of life by reducing pain and increasing function for patients with OA, especially those who are overweight or obese.[110] A combination of diet and exercise has been shown to reduce pain and increase muscle mass in patients with OA, and that diet alone or in combination with exercise can improve function.[111][112]

Osteoarthritis core treatments

A primary care physician who leads research for Arthritis UK discusses core treatments for osteoarthritis, including education, advice and support, weight loss, and exercise.

Assistive devices

The American College of Rheumatology (ACR) recommends:[7]

Use of a walking stick for patients with knee and/or hip OA in one or more joints, which is causing a sufficiently large impact on ambulation, joint stability, or pain to warrant their use
Tibiofemoral knee braces for patients with OA of the knee, in one or both knees, which is causing a sufficiently large impact on ambulation, joint stability, or pain to warrant their use
Hand orthoses for patients with OA of the first carpometacarpal (CMC) joint, or in joints apart from the first CMC joint of the hand
Patellofemoral braces for patients with patellofemoral knee OA.

The ACR does not recommend modified shoes, or lateral and medial wedged insoles for patients with knee and/or hip OA.[7]

The National Institute of Health and Care Excellence (NICE) in the UK recommends:[73]

Walking aids, such as walking sticks, for people with lower limb OA
Insoles, braces, tape, splints, or supports are not routinely recommended to OA patients, unless
- there is joint instability or abnormal biomechanical loading AND
- therapeutic exercise is ineffective or unsuitable without the addition of an aid or device AND
- the addition of an aid or device is likely to improve movement.

One network meta-analysis reported that lateral wedge insoles in combination with knee bracing reduce peak knee adduction moment in patients with tibiofemoral OA, while gait training influenced both knee adduction angular impulse and knee adduction moment, so it is recommended for reducing biomechanical risk factors.[113]

There is conflicting evidence for the use of orthoses and/or braces for medial knee OA. There is evidence to suggest that lateral wedge insoles do not reduce pain or improve functionality in patients with medial knee OA, but conversely that lateral wedge insoles with arch support significantly improved pain and physical function in patients with knee OA.[114][115][116]

Knee valgus bracing has been demonstrated as an effective intervention to improve the quality of life and reduce pain during daily activities for patients with medial knee OA.[117] However, evidence suggests that valgus knee bracing may only be effective in the short term.[118][119][120][121] [ ]

Combining both a knee brace and lateral wedge insoles has been shown to improve pain and function in patients with medial knee OA.[122]

Unloader shoes do not appear to confer benefit in medial knee OA.[123]

Patellar bracing or taping for patellofemoral pain can be considered. One RCT suggests the use of a knee brace may be helpful in reducing pain and bone marrow lesions in patellofemoral OA.[124] Results of one meta-analysis reported that a multimodal physiotherapy intervention that included taping significantly reduced pain in the short term for patients with patellofemoral OA.[125]

One meta-analysis found that splinting in patients with thumb and carpometacarpal (CMC) joint OA reduced pain and improved function in the medium term (3-12 months), but not the short term.[126]

Dietary supplements

Glucosamine and chondroitin sulfate are not recommended for the management of patients with OA; decisions regarding the use of these agents should be discussed with patients.[7][73] Despite this recommendation, glucosamine and chondroitin sulfate are commonly used by people with OA. Modest efficacy and low risk may explain the popularity of these supplements among patients.

Both agents have been associated with modest pain reduction in patients with knee OA, and are considered safe.[127][128][129][130][131][132] [ ] However, many trials are of low quality.[130]

Results of studies on the efficacy of glucosamine or chondroitin vary. One meta-analysis found that glucosamine or chondroitin sulfate reduced pain in patients with knee OA individually, but found no additional benefit associated with combination treatment, whereas subsequent evidence suggests that combination treatment is effective for the treatment of knee OA, compared with placebo.[129][133]

The inconsistencies between the labelling and actual contents of many dietary supplements should be considered; prescription-grade preparations should be sought.[132][134]

Acupuncture and transcutaneous electrical nerve stimulation (TENS)

The ACR recommends acupuncture for patients with knee, hip, and/or hand OA.[7] However, NICE in the UK does not recommend acupuncture for the management of OA.[73] TENS is not recommended for the treatment of patients with OA due to insufficient evidence of benefit.[7][73]

Evidence suggests that acupuncture may benefit patients with knee OA.[135][136][137] However, evidence of short-term benefit is based on low- to very-low-quality evidence, and may not be clinically important, when compared with control treatments.[138]

One Cochrane review concluded that acupuncture does not appear to reduce pain or improve function relative to sham acupuncture in people with hip OA.[139] However, subsequent meta-analysis suggest that acupuncture is reduced pain and improves function in patients with OA of the knee, and may be used as an adjunctive treatment.[140][141]

The results of a Cochrane review reported that there is a lack of evidence to support the use of TENS to treat patients with OA of the knee, but there is also evidence to suggest that acupuncture significantly reduced pain, and improved walking ability in patients with OA of the knee.[142][143]

Pharmacological treatment

Pharmacological management is required if non-pharmacological approaches do not adequately control symptoms, although these can be used on an as-needed basis in some patients.

If pharmacological treatments are needed, they should be used alongside non-pharmacological treatments and to support therapeutic exercise, at the lowest effective dose for the shortest time possible.[73]

Topical analgesia

Topical non-steroidal anti-inflammatory drugs (NSAIDs) are recommended for patients with OA of the hand or knee, and can be considered for other affected joints.[7][73] Other topical analgesics include capsaicin and methylsalicylate.[7] The ACR recommends topical capsaicin for patients with OA of the knee, but not for patients with OA of the hand.[7]

Topical NSAIDs effectively relieve pain in adults with knee or hand OA within 2 weeks of daily application. [ ] The results of systematic reviews and meta-analyses report that topical NSAIDs are the most effective topical analgesic for pain relief for OA patients compared with oral NSAIDs, cyclo-oxygenase-2 [COX-2] inhibitors, and opioids, and that diclofenac transdermal patches may be the most effective and safest topical NSAID for pain relief.[144][145]

Systematic reviews suggest that topical NSAIDs are relatively safe for the management of pain associated with OA.[146][147][148][149] However, confirmation of the cardiovascular safety of topical NSAIDs requires further study.[146]

One Cochrane review found that arnica gel may improve symptoms as effectively as a topical NSAID, but with a potentially worse adverse effect profile.[150] In the same review, capsicum-extract gel (capsaicinoids ≤0.05%) did not significantly improve pain or function compared with placebo.[150] However, results of a network meta-analysis suggest that topical capsaicin may be as effective as topical NSAIDs at reducing pain in patients with OA.[151]

Oral analgesia

Guidance on oral analgesia for OA treatment varies, check local guidance.

In the US, paracetamol, oral NSAIDs, tramadol, and duloxetine (an antidepressant with analgesic properties) are recommended for patients with knee, hip, and/or hand OA.[7] Opioids other than tramadol may be considered when other treatments are ineffective.[7]

In the UK, oral NSAIDs are recommended as first-line oral pharmacological treatment, taking into account potential gastrointestinal, renal, liver, and cardiovascular toxicity, and any risk factors the patient may have, including age, pregnancy, current medication, and comorbidities.[73][73] Gastroprotection should be considered for patients on long-term NSAID therapy, especially those at risk of gastrointestinal bleeding.[73] Paracetamol or weak opioids are not routinely offered to OA patients unless they are used infrequently for short-term pain relief, or all other oral analgesia is contraindicated, not tolerated or ineffective.[73]

Paracetamol

Studies have demonstrated that paracetamol has a small to modest benefit for patients with OA of the hip or knee, and is statistically inferior to all other drug categories for the management of OA pain (oral NSAIDs, topical NSAIDs, COX-2 inhibitors, and opioids).[144][152][153]

As such paracetamol alone may not have a role in the treatment of hip or knee OA, irrespective of the dose used, but may be added for rescue analgesia, or if local therapies alone do not control symptoms.[134][154]

With this evidence of limited efficacy, and with more data available regarding the potential adverse reactions of paracetamol, careful consideration should taken about the use of paracetamol for the treatment of OA.[152][154][155]

Osteoarthritis: benefits of paracetamol

A primary care physician who leads research for Arthritis UK discusses the benefits of paracetamol for patients with hip and knee pain due to osteoarthritis.

NSAIDs

Oral NSAIDs are more effective than paracetamol for the management of OA pain; however, they are associated with gastrointestinal (GI) and renal toxicity.[144][156][157] Gastroprotection should be offered to patients on long-term NSAID therapy, especially those at risk of GI bleeding.[73] Evidence suggests that proton-pump inhibitors (PPIs) provide better protection against NSAID-induced peptic ulcer disease and gastritis compared with H2 antagonists.[158] Misoprostol is a prostaglandin E1 analogue and is another option for gastroprotection, but diarrhoea is a common adverse effect, and the drug is less well tolerated than PPIs.[159][160][161]

Diclofenac or etoricoxib (not available in some countries) may be the most effective NSAID for the treatment of pain in knee and hip OA, but potential benefit must be weighed against adverse effects, and may not be appropriate for patients with comorbidities or for long-term use.[154][145]

Selective COX-2 inhibitors may be used as an alternative to non-selective NSAIDs. They are associated with reduced risk of GI adverse effects compared with non-selective NSAIDs, but similar renal toxicity.[162][163] COX-2 inhibitors are effective for the management of pain associated with knee and hip OA, and may have a role in patients at increased risk for GI adverse effects.[134][154] However, COX-2 inhibitors do not confer an advantage with respect to GI symptoms when compared with placebo, or NSAID and PPI used concomitantly for gastroprotection.[164][165] The incidence of upper GI adverse effects did not differ between patients with knee OA who were treated with fixed-dose combination naproxen and esomeprazole or with celecoxib; the former reported significantly more heartburn-free days than those on celecoxib.[166]

Evidence suggests that NSAID use substantially contributes to the association between OA and cardiovascular disease (CVD), with increased risk reaching significance as early as 4 weeks into treatment.[167][168] Several patient characteristics may be associated with increased CVD risk when taking an NSAID, such as age >80 years, history of CVD, rheumatoid arthritis, chronic obstructive pulmonary disease, renal disease, and hypertension.[169] One meta-analysis suggested that diclofenac and ibuprofen were associated with increased cardiovascular risk, while naproxen and celecoxib were not.[170] However, similar incident rates of cardiovascular events have been reported for ibuprofen, celecoxib, and naproxen.[171] GI and cardiovascular safety profiles of individual oral NSAIDs differ, and careful patient selection is required to maximise the risk:benefit ratio.[134] The lowest effective dose of NSAID should be used to minimise adverse effects.

Opioids

It should be noted that the potential clinical benefit of opioid treatment, regardless of preparation or dose, does not outweigh the harm opioid treatment may cause in patients with OA.[145] Opioids provide minimal relief of OA symptoms, and are known to cause discomfort in many patients. Clinicians should give careful consideration to the utility of opioids in the management of OA.[172]

Oral and transdermal opioids can decrease pain intensity and improve function in patients with OA of the knee or hip compared with placebo, but the observed benefits were small (12% absolute improvement in mean pain compared with placebo [various pain scales]; number needed to benefit of 10).[173] No studies of tramadol contributed to these results.

A subsequent meta-analysis reported that opioids did not demonstrate a clinically relevant reduction in pain or disability compared with placebo in patients with OA of the hip or knee in at 4-24 weeks. Number needed to treat for an additional dropout due to side effects was 5 (95% CI 4 to 7).[174]

Evidence suggests that tramadol is generally well tolerated and can be combined with paracetamol and/or NSAIDs.[175] However, tramadol alone or in combination with paracetamol is unlikely to have an important benefit on mean pain or function in patients with OA.[175][176]

Duloxetine

Duloxetine is an antidepressant with analgesic properties.

Results from one systematic review suggest that duloxetine may be effective for the treatment of chronic pain associated with OA, with a number needed to benefit (clinically meaningful outcome at study end compared with placebo) of 7.[177]

Indirect comparisons between duloxetine and a number of post-first-line oral treatments for OA, including selective COX-2 inhibitors and opioids, found no difference in the total WOMAC composite scores (an inclusive set of OA outcomes) after approximately 12 weeks of treatment.[178] Some analyses suggested that etoricoxib may be superior to duloxetine.[178]

Evidence from subsequent systematic reviews found that duloxetine moderately reduces pain compared with placebo, in patients with knee OA.[179][180][181][182]

Commonly observed adverse effects reported among patients with OA treated with duloxetine include nausea, fatigue, constipation, and dry mouth.[179] There is a possible increased serotonergic effect if given with tramadol.

Acute exacerbation of symptoms despite regular analgesia

Intra-articular corticosteroid injections and intra-articular viscosupplementation are useful, particularly in the knee, for acute exacerbations of OA or when NSAIDs are contraindicated or not tolerated. These interventions can be used in addition to the non-pharmacological therapies and analgesia.

Intra-articular corticosteroids and intra-articular viscosupplementation should be compared with injectable placebo to determine the incremental effect size in reducing pain and improving function that can be attributed to the therapeutic agent. In meta-analyses, administration of an intra-articular normal saline placebo injection to patients with knee or hip OA resulted in clinically meaningful improvements in pain and function at 6 months.[183][184]

Not all placebos are the same; effect size varies with the method of delivering the drug.

Intra-articular corticosteroid injections

The ACR recommends intra-articular corticosteroid injections for patients with knee and/or hip OA, but only conditionally recommends this treatment for patients with OA of the hand.[7] In the UK, intra-articular corticosteroid injections are only recommended when other pharmacological treatments are ineffective or unsuitable, or to support therapeutic exercise.[73]

Trials comparing intra‐articular corticosteroid injections with sham or non‐intervention controls are often small and of low methodological quality.[185][186]

Intra-articular corticosteroid injections reduced pain and improved function in patients with OA of the knee at 6 weeks compared with placebo.[187] However, it appears that intra-articular corticosteroid injections do not reduce joint pain for patients with hand or temporomandibular OA compared with placebo.[188][189]

It is unclear how long the benefit of intra-articular corticosteroids lasts in patients with OA. Results from meta-analyses vary, with reports of continued efficacy from 1 to 12 weeks in patients with OA of the hip.[185][186][190][191] However, intra-articular corticosteroid may increase the risk of rapidly destructive hip disease, especially at higher doses.[192]

Meta-analysis of individual patient data suggests that patients with severe knee pain at baseline may derive greater short-term benefit (reduction in pain up to 4 weeks) from intra‐articular corticosteroid injection than patients with less severe pain.[193]

Intra-articular triamcinolone every 12 weeks for 2 years failed to significantly reduce OA knee pain compared with intra-articular saline (-1.2 vs. -1.9; between-group difference -0.6, 95% CI -1.6 to 0.3) in a double-blind RCT.[194] Triamcinolone was associated with significantly greater cartilage volume loss than saline (mean change in index compartment cartilage thickness of -0.21 mm vs. -0.10 mm; between-group difference -0.11 mm, 95% CI -0.20 to -0.03), but the clinical significance of this finding is unclear.[194]

Time-limited adverse effects of intra-articular injection include post-injection pain, swelling, and post-injection flare. Intra-articular injection of corticosteroid was not associated with loss of joint space at 1- and 2-year follow-up in a placebo-controlled randomised trial of patients with knee arthritis.[195] Similarly, in meta-analysis intra-articular corticosteroids for knee OA had no effect on joint space narrowing beyond that of control interventions.[185]

Evidence suggests that recurrent intra-articular corticosteroid injections often provide inferior (or non-superior) symptom relief compared with other injectables (including placebo) at 3 months and beyond in patients with OA.[196]

Intra-articular viscosupplementation

Guidelines do not recommend intra-articular hyaluronic acid injections for the management of OA.[7][73]

Despite this recommendation, it is commonly used for the management of symptomatic knee arthritis; studies variously report modest or no benefit.[197][198][199]

One literature review concludes that intra-articular hyaluronic acid should be considered as a treatment for patients with OA, tailored by disease stage and patient phenotype, despite recommendations to the contrary from international guidelines.[200]

One meta-analysis found that intra-articular viscosupplementation with hyaluronan or hylan derivatives is effective in the management of OA of the knee; improvement from baseline during the 5- to 13-week post-injection period was 28% to 54% for pain and 9% to 32% for function.[199] The analyses suggested that different hyaluronan/hylan products exert differential therapeutic effects, and that response is time dependent.[199]

Analysing data only from placebo-controlled trials with low risk of bias, one meta-analysis indicated that intra-articular hyaluronic acid provides a modest, but real, benefit for patients with OA of the knee (pain intensity standardised mean difference [SMD] -0.21, 95% CI -0.32 to -0.10; function at 3 months SMD -0.12, 95% CI -0.22 to -0.02).[201]

However, in subsequent meta-analyses, intra-articular injection of hyaluronic acid was not associated with a clinically important difference in pain for patients with OA of the knee compared with placebo, but may increase the risk of serious adverse effects.[197][202]

Surgery

Patients with OA pain that persists despite multiple treatment modalities and which substantially impacts their quality of life should be referred and considered for joint placement surgery.

Total knee replacement followed by non-surgical treatment resulted in significantly greater pain relief and functional improvement after 12 months than non-surgical treatment alone (Knee Injury and Osteoarthritis Outcome Score [KOOS4] 32.5 vs. 16.0; adjusted mean difference 15.8, 95% CI 10.0 to 21.5) in an RCT of patients with moderate-to-severe knee OA who were eligible for unilateral total knee replacement.[203] Total knee replacement was associated with more serious adverse events.[203]

Unipartmental (partial) knee arthroplasty has been demonstrated to provide pain relief and satisfactory activity level for patients aged 60 years or younger. The results of one meta-analysis reported that 96.5% of implants survived at 10-year follow-up.[204]

There is no role of partial meniscectomy for meniscal tear in knee OA based on an RCT.[205]

Arthroscopic surgery is not effective for knee OA.[33][206][207] Clinical guidelines do not recommend the use of arthroscopic surgery in knee OA.[7][73]

BMJ Rapid Recommendations: arthroscopic surgery for degenerative knee arthritis and meniscal tears Opens in new window MAGICapp: recommendations, evidence summaries and consultation decision aids Opens in new window

[Figure caption and citation for the preceding image starts]: BMJ Rapid Recommendations: arthroscopic surgery for degenerative knee arthritis and meniscal tearsSiemieniuk RAC, et al. BMJ.2017 May 10;357:j1982 [Citation ends]. com.bmj.content.model.Caption@a1f57b2

In patients with primary glenohumeral OA with an intact rotator cuff, total shoulder arthroplasty significantly improved postoperative patient-reported outcome measures (PROMs) compared with hemiarthroplasty.[208]

Evidence suggests that denervation may be an effective treatment for OA of the hand, especially for proximal inter-phalangeal (PIP) and trapeziometacarpal (TMC) joints.[209][210]

Arthroplasty, trapeziectomy, and arthrodesis are options for thumb OA.[211] One meta-analysis concluded that there remains uncertainty about which procedure offers the best functional outcome and safety profile to treat OA of the thumb, the results of the systematic review suggest trapeziectomy with ligament reconstruction and tendon interposition yielded good postoperative range of movement, while arthrodesis demonstrated a high rate of moderate-severe complications.[212]

Osteoarthritis: referral for surgery

A primary care physician who leads research for Arthritis UK discusses when joint replacement surgery should be offered to patients with osteoarthritis.

Preoperative interventions and preparation

Pre-surgical obesity is associated with worse clinical outcomes following hip or knee arthroplasty in patients with OA; weight loss prior to elective surgery for knee or hip OA may improve post-surgical outcomes and reduce the risk of patient harm.[213]

Low- to moderate-quality evidence suggests that preoperative exercise reduces pain in patients with hip or knee OA prior to joint replacement, and may improve activity after hip replacement.[214]

Many orthopaedists recommend delaying joint replacement surgery for at least 3 months following intra-articular corticosteroid injection to reduce the risk of postoperative peri-prosthetic infections.[215][216]

[Figure caption and citation for the preceding image starts]: Osteoarthritis treatment flowchartCreated by BMJ Knowledge Centre [Citation ends]. com.bmj.content.model.Caption@5f346c26

Use of this content is subject to our disclaimer