Advertisement

OARSI recommendations for the management of hip and knee osteoarthritis, Part II: OARSI evidence-based, expert consensus guidelines

      Summary

      Purpose

      To develop concise, patient-focussed, up to date, evidence-based, expert consensus recommendations for the management of hip and knee osteoarthritis (OA), which are adaptable and designed to assist physicians and allied health care professionals in general and specialist practise throughout the world.

      Methods

      Sixteen experts from four medical disciplines (primary care, rheumatology, orthopaedics and evidence-based medicine), two continents and six countries (USA, UK, France, Netherlands, Sweden and Canada) formed the guidelines development team. A systematic review of existing guidelines for the management of hip and knee OA published between 1945 and January 2006 was undertaken using the validated appraisal of guidelines research and evaluation (AGREE) instrument. A core set of management modalities was generated based on the agreement between guidelines. Evidence before 2002 was based on a systematic review conducted by European League Against Rheumatism and evidence after 2002 was updated using MEDLINE, EMBASE, CINAHL, AMED, the Cochrane Library and HTA reports. The quality of evidence was evaluated, and where possible, effect size (ES), number needed to treat, relative risk or odds ratio and cost per quality-adjusted life years gained were estimated. Consensus recommendations were produced following a Delphi exercise and the strength of recommendation (SOR) for propositions relating to each modality was determined using a visual analogue scale.

      Results

      Twenty-three treatment guidelines for the management of hip and knee OA were identified from the literature search, including six opinion-based, five evidence-based and 12 based on both expert opinion and research evidence. Twenty out of 51 treatment modalities addressed by these guidelines were universally recommended. ES for pain relief varied from treatment to treatment. Overall there was no statistically significant difference between non-pharmacological therapies [0.25, 95% confidence interval (CI) 0.16, 0.34] and pharmacological therapies (ES=0.39, 95% CI 0.31, 0.47). Following feedback from Osteoarthritis Research International members on the draft guidelines and six Delphi rounds consensus was reached on 25 carefully worded recommendations. Optimal management of patients with OA hip or knee requires a combination of non-pharmacological and pharmacological modalities of therapy. Recommendations cover the use of 12 non-pharmacological modalities: education and self-management, regular telephone contact, referral to a physical therapist, aerobic, muscle strengthening and water-based exercises, weight reduction, walking aids, knee braces, footwear and insoles, thermal modalities, transcutaneous electrical nerve stimulation and acupuncture. Eight recommendations cover pharmacological modalities of treatment including acetaminophen, cyclooxygenase-2 (COX-2) non-selective and selective oral non-steroidal anti-inflammatory drugs (NSAIDs), topical NSAIDs and capsaicin, intra-articular injections of corticosteroids and hyaluronates, glucosamine and/or chondroitin sulphate for symptom relief; glucosamine sulphate, chondroitin sulphate and diacerein for possible structure-modifying effects and the use of opioid analgesics for the treatment of refractory pain. There are recommendations covering five surgical modalities: total joint replacements, unicompartmental knee replacement, osteotomy and joint preserving surgical procedures; joint lavage and arthroscopic debridement in knee OA, and joint fusion as a salvage procedure when joint replacement had failed. Strengths of recommendation and 95% CIs are provided.

      Conclusion

      Twenty-five carefully worded recommendations have been generated based on a critical appraisal of existing guidelines, a systematic review of research evidence and the consensus opinions of an international, multidisciplinary group of experts. The recommendations may be adapted for use in different countries or regions according to the availability of treatment modalities and SOR for each modality of therapy. These recommendations will be revised regularly following systematic review of new research evidence as this becomes available.

      Key words

      Osteoarthritis (OA) is the most common type of arthritis and the major cause of chronic musculoskeletal pain and mobility disability in elderly populations worldwide
      • Peat G.
      • McCarney R.
      • Croft P.
      Knee pain and osteoarthritis in older adults: a review of community burden and current use of primary health care.
      . Knee and hip pain are the major causes of difficulty in walking and climbing stairs in the elderly in Europe
      • Dawson J.
      • Linsell L.
      • Zondervan K.
      • Rose P.
      • Randall T.
      • Carr A.
      • et al.
      Epidemiology of hip and knee pain and its impact on overall health status in older adults.
      and the USA
      • Dunlop D.D.
      • Manheim L.M.
      • Song J.
      • Chang R.W.
      Arthritis prevalence and activity limitations in older adults.
      and as many as 40% of people over the age of 65 in the community in the United Kingdom suffer symptoms associated with knee or hip OA
      • Dawson J.
      • Linsell L.
      • Zondervan K.
      • Rose P.
      • Randall T.
      • Carr A.
      • et al.
      Epidemiology of hip and knee pain and its impact on overall health status in older adults.
      .
      Treatment of OA of the knee and hip is directed towards
      • Reducing joint pain and stiffness.
      • Maintaining and improving joint mobility.
      • Reducing physical disability and handicap.
      • Improving health-related quality of life.
      • Limiting the progression of joint damage.
      • Educating patients about the nature of the disorder and its management.
      More than 50 modalities of non-pharmacological, pharmacological and surgical therapy for knee and hip OA are described in the medical literature
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      .
      Over the years a number of National and Regional Guidelines have been developed to assist physicians, allied health professionals and patients in their choice of therapy for the management of knee and hip OA, but internationally agreed and universally applicable guidelines for the management of these global disorders have been lacking.
      In September 2005 the Osteoarthritis Research International (OARSI) appointed an international, multidisciplinary committee of experts with a remit to produce up to date, evidence-based, globally relevant, consensus recommendations for the management of knee and/or hip OA in 2007. The first part of the work of this committee was to undertake a critical appraisal of all existing evidence-based and consensus guidelines for the treatment of knee and/or hip OA and a systematic review of the recent research evidence. The results of this critical appraisal and systematic review were published recently
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . This second part of the report contains the current OARSI evidence-based, expert consensus recommendations for the treatment of knee and/or hip OA.

      Scope and purpose

      The guidelines are intended to provide concise, patient-focussed, up to date, evidence-based, expert consensus recommendations for the management of hip and knee OA, which are globally relevant.

      Target users

      The guidelines have been developed to provide assistance to physicians and allied health care professionals who deal with patients with OA hip and knee in both primary and secondary (specialist) care settings. The guidelines should also provide a helpful resource for patients with OA hip or knee, patient representative groups and health care funders and administrators. It is anticipated that these OARSI international core recommendations will be modified and adapted as appropriate for National and Regional use.

      Stakeholder involvement

      The guideline development committee was composed of 16 experts from four medical disciplines (primary care 2, rheumatology 11, orthopaedics 1, and evidence-based medicine 2) and six countries in Europe and North America (France, Netherlands, Sweden, UK, Canada and USA). All members of the development team participated in (1) a critical appraisal of existing treatment guidelines
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      ; (2) a Delphi exercise to generate consensus recommendations; and (3) an exercise to grade the strength of recommendation (SOR) for all modalities of therapy recommended.

      Rigour of development

      Critical appraisal of existing guidelines

      Methodological details of the systematic literature search, the inclusion/exclusion criteria, the quality and content assessment and the data analyses of all existing guidelines for the management of hip and/or knee OA published between 1945 and October 2005 can be found in the first part of this report
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The quality of the guidelines was assessed using the AGREE instrument
      The AGREE Collaboration
      Appraisal of Guidelines for Research & Evaluation (AGREE) Instrument.
      and standardised percent scores (0–100%) for scope, stakeholder involvement, rigour, clarity, applicability and editorial independence, as well as overall quality, were calculated. Treatment modalities addressed and recommended by the guidelines were summarised. Agreement (%) was estimated and the best level of evidence (LoE) to support each recommendation was extracted.

      Systematic review of the more recent evidence

      Systematic reviews of research evidence for the treatment of hip and/or knee OA up to January 2002 were available from the systematic literature review undertaken by the European League against Rheumatism (EULAR). Methodological details of the systematic literature search, the inclusion/exclusion criteria, the quality assessments and outcome measures (efficacy, side effects and cost-effectiveness) for research evidence relating to the treatment of OA hip and/or knee published between 31st January 2002 and 31st January 2006 can also be found in the first part of this report
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The quality of evidence was evaluated using the Oxman and Guyatt method for systematic reviews
      • Oxman A.D.
      • Guyatt G.H.
      Validation of an index of the quality of review articles.
      and the Jadad scale for randomised controlled trials (RCTs)
      • Jadad A.R.
      • Moore R.A.
      • Carroll D.
      • Jenkinson C.
      • Reynolds D.J.
      • Gavaghan D.J.
      • et al.
      Assessing the quality of reports of randomized clinical trials: is blinding necessary?.
      . Where possible, effect size (ES)
      • Hedges L.V.
      Fitting continues models to effect size data.
      , number needed to treat (NNT)
      • Cook R.J.
      • Sackett D.L.
      The number needed to treat: a clinically useful measure of treatment effect.
      , relative risk (RR) or odds ratio (OR)
      • Kleinbaum D.G.
      • Kuppler L.L.
      • Morgenstern H.
      Epidemiologic Research – Principles and Quantitative Methods.
      and cost per quality-adjusted life year (QALY) gained
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      were estimated. Sensitivity analyses
      • Lau J.
      • Ioannidis J.P.A.
      • Schmid C.H.
      Quantitative synthesis in systematic reviews.
      were undertaken to determine whether selected RCTs published after January 31st 2006 would alter any of the evidence-based conclusions from the critical appraisal of existing guidelines and the systematic review of the recent research evidence significantly.

      Delphi exercise to generate consensus recommendations

      Concise propositions relating to all aspects of non-pharmacological, pharmacological and surgical treatments of OA hip and/or knee were generated as follows.
      The committee of experts was divided into three subgroups:
      • Non-pharmacological: Altman, Brandt, Croft, and Doherty.
      • Pharmacological: Abramson, Bierma-Zeinstra, Dougados, and Hochberg.
      • Surgical: Arden, Hunter, Kwoh, Lohmander, and Tugwell.
      Each expert was provided with a comprehensive table of 51 potential treatment modalities together with a summary of recommendations from the critical appraisal of existing guidelines
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      (percentage of guidelines addressing modality, AGREE instrument score for quality
      The AGREE Collaboration
      Appraisal of Guidelines for Research & Evaluation (AGREE) Instrument.
      , the LoE
      • Shekelle P.G.
      • Woolf S.H.
      • Eccles M.
      • Grimshaw J.
      Clinical guidelines: developing guidelines.
      and ES for pain
      • Hedges L.V.
      Fitting continues models to effect size data.
      ) and a summary of the systematic analysis of the research evidence from 2002 to 2006
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      (Quality scores
      • Oxman A.D.
      • Guyatt G.H.
      Validation of an index of the quality of review articles.
      • Jadad A.R.
      • Moore R.A.
      • Carroll D.
      • Jenkinson C.
      • Reynolds D.J.
      • Gavaghan D.J.
      • et al.
      Assessing the quality of reports of randomized clinical trials: is blinding necessary?.
      , ES
      • Hedges L.V.
      Fitting continues models to effect size data.
      for pain, function and stiffness, the NNT
      • Cook R.J.
      • Sackett D.L.
      The number needed to treat: a clinically useful measure of treatment effect.
      , the RR/OR
      • Kleinbaum D.G.
      • Kuppler L.L.
      • Morgenstern H.
      Epidemiologic Research – Principles and Quantitative Methods.
      and the cost per QUALY
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      ). A full list of references from which the summary data had been extracted was also provided. With the exception of the co-chairs (RM and GN) and the lead researcher (WZ), who did not contribute to the primary generation of propositions in order to avoid administrative bias, each committee expert was asked to generate a comprehensive list of propositions relating to modalities of treatment in the group to which they were assigned, based on the available research evidence and their own clinical expertise. There was no limit to the number of propositions proposed for the initial master list.
      After elimination of closely similar or overlapping propositions a master list of 110 propositions relating to 54 non-pharmacological modalities of treatment, 37 pharmacological, 18 surgical and one combining non-pharmacological and pharmacological modalities was circulated to all members of the guideline development group apart from RM, GN and WZ for acceptance or rejection. The experts were also given the opportunity to suggest amalgamations and rewording of individual propositions. After four rounds of the Delphi exercise in which propositions with >60% of votes were accepted, those with <20% were rejected and those attracting between 20% and 60% of votes were taken forward for consideration following further amalgamations and minor rewording, provisional consensus was reached on 34 propositions. These were posted on the OARSI website and presented for comments and suggestions by OARSI members at a plenary session of the World Congress on OA in Prague in December 2006. After further additions, amalgamations, minor rewording and two further Delphi rounds, consensus was reached on accepting 25 carefully worded propositions (Table I). All eligible members of the committee voted at each step of the Delphi exercise.
      Table IOARSI recommendations and research evidence
      PropositionLoEES for pain (95% CI)Frequency recommended in existing guidelinesLevel of consensus (%)SOR (%) (95% CI)
      General
       1. Optimal management of OA requires a combination of non-pharmacological and pharmacological modalities.IV12/1210096 (93–99)
      Non-pharmacological modalities of treatment
       2. All patients with hip and knee OA should be given information access and education about the objectives of treatment and the importance of changes in lifestyle, exercise, pacing of activities, weight reduction, and other measures to unload the damaged joint(s). The initial focus should be on self-help and patient-driven treatments rather than on passive therapies delivered by health professionals. Subsequently emphasis should be placed on encouraging adherence to the regimen of non-pharmacological therapy.Ia (education)0.06 (0.02, 0.10)8/89297 (95–99)
      IV (adherence)
       3. The clinical status of patients with hip or knee OA can be improved if patients are contacted regularly by phone.Ia0.12 (0.00, 0.24)2/27766 (57–75)
       4. Patients with symptomatic hip and knee OA may benefit from referral to a physical therapist for evaluation and instruction in appropriate exercises to reduce pain and improve functional capacity. This evaluation may result in provision of assistive devices such as canes and walkers, as appropriate.IV5/510089 (82–96)
       5. Patients with hip and knee OA should be encouraged to undertake, and continue to undertake, regular aerobic, muscle strengthening and range of motion exercises. For patients with symptomatic hip OA, exercises in water can be effective.Ia (knee)0.52 (0.34, 0.70) aerobic21/218596 (93–99)
      IV (hip)0.32 (0.23, 0.42) strength21/21
      Ib (hip, water based)0.25 (0.02, 0.47) water based8/8
       6. Patients with hip and knee OA, who are overweight, should be encouraged to lose weight and maintain their weight at a lower level.Ia0.13 (−0.12, 0.38)13/1410096 (92–100)
       7. Walking aids can reduce pain in patients with hip and knee OA. Patients should be given instruction in the optimal use of a cane or crutch in the contralateral hand. Frames or wheeled walkers are often preferable for those with bilateral disease.IV11/1110090 (84–96)
       8. In patients with knee OA and mild/moderate varus or valgus instability, a knee brace can reduce pain, improve stability and diminish the risk of falling.Ia8/99276 (69–83)
       9. Every patient with hip or knee OA should receive advice concerning appropriate footwear. In patients with knee OA insoles can reduce pain and improve ambulation. Lateral wedged insoles can be of symptomatic benefit for some patients with medial tibio-femoral compartment OA.IV (footwear)9277 (66–88)
      Ia (insole)12/13
       10. Some thermal modalities may be effective for relieving symptoms in hip and knee OA.Ia0.69 (−0.07, 1.45)7/107764 (60–68)
       11. TENS can help with short-term pain control in some patients with hip or knee OA.Ia8/106958 (45–72)
       12. Acupuncture may be of symptomatic benefit in patients with knee OA.Ia0.51 (0.23, 0.79)5/86959 (47–71)
      Pharmacological modalities of treatment
       13. Acetaminophen (up to 4 g/day) can be an effective initial oral analgesic for treatment of mild to moderate pain in patients with knee or hip OA. In the absence of an adequate response, or in the presence of severe pain and/or inflammation, alternative pharmacologic therapy should be considered based on relative efficacy and safety, as well as concomitant medications and co-morbidities.Ia (knee)0.21 (0.02, 0.41)16/167792 (88–99)
      IV (hip)
       14. In patients with symptomatic hip or knee OA, non-steroidal anti-inflammatory drugs (NSAIDs) should be used at the lowest effective dose but their long-term use should be avoided if possible. In patients with increased GI risk, either a COX-2 selective agent or a non-selective NSAID with co-prescription of a PPI or misoprostol for gastroprotection may be considered, but NSAIDs, including both non-selective and COX-2 selective agents, should be used with caution in patients with CV risk factors.Ia (knee)0.32 (0.24, 0.39)NSAID+PPI 8/810093 (88–99)
      Ia (hip)NSAID+misoprostol 8/8
      COX-2 inhibitors 11/11
       15. Topical NSAIDs and capsaicin can be effective as adjunctives and alternatives to oral analgesic/anti-inflammatory agents in knee OA.Ia (NSAIDs)0.41 (0.22, 0.59)7/910085 (75–95)
      Ia (capsaicin)8/9
       16. IA injections with corticosteroids can be used in the treatment of hip or knee OA, and should be considered particularly when patients have moderate to severe pain not responding satisfactorily to oral analgesic/anti-inflammatory agents and in patients with symptomatic knee OA with effusions or other physical signs of local inflammation.Ib (hip)0.72 (0.42, 1.02)11/136978 (61–95)
      Ia (knee)
       17. Injections of IA hyaluronate may be useful in patients with knee or hip OA. They are characterised by delayed onset, but prolonged duration, of symptomatic benefit when compared to IA injections of corticosteroids.Ia (knee)0.32 (0.17, 0.47)8/98564 (43–85)
      Ia (hip)
       18. Treatment with glucosamine and/or chondroitin sulphate may provide symptomatic benefit in patients with knee OA. If no response is apparent within 6 months treatment should be discontinued.Ia (glucosamine)0.45 (0.04, 0.86)6/109263 (44–82)
      Ia (chondroitin)0.30 (−0.10, 0.70)2/7
       19. In patients with symptomatic knee OA glucosamine sulphate and chondroitin sulphate may have structure-modifying effects while diacerein may have structure-modifying effects in patients with symptomatic OA of the hip.Ib (knee)6941 (20–62)
      Ib (hip)
       20. The use of weak opioids and narcotic analgesics can be considered for the treatment of refractory pain in patients with hip or knee OA, where other pharmacological agents have been ineffective, or are contraindicated. Stronger opioids should only be used for the management of severe pain in exceptional circumstances. Non-pharmacological therapies should be continued in such patients and surgical treatments should be considered.Ia (week opioids)9/99282 (74–90)
      IV (strong opioids)
      IV (others)
      Surgical modalities of treatment
       21. Patients with hip or knee OA who are not obtaining adequate pain relief and functional improvement from a combination of non-pharmacological and pharmacological treatment should be considered for joint replacement surgery. Replacement arthroplasties are effective, and cost-effective interventions for patients with significant symptoms, and/or functional limitations associated with a reduced health-related quality of life, despite conservative therapy.III14/149296 (94–98)
       22. Unicompartmental knee replacement is effective in patients with knee OA restricted to a single compartment.IIb10076 (64–88)
       23. Osteotomy and joint preserving surgical procedures should be considered in young adults with symptomatic hip OA, especially in the presence of dysplasia. For the young and physically active patient with significant symptoms from unicompartmental knee OA, high tibial osteotomy may offer an alternative intervention that delays the need for joint replacement some 10 years.IIb10/1010075 (64–86)
       24. The role of joint lavage and arthroscopic debridement in knee OA are controversial. Although some studies have demonstrated short-term symptom relief, others suggest that improvement in symptoms could be attributable to a placebo effect.Ib (lavage)0.09 (−0.27, 0.44)3/310060 (47–82)
      Ib (debridement)−0.01 (−0.37, 0.35)5/6
       25. In patients with OA of the knee, joint fusion can be considered as a salvage procedure when joint replacement has failed.IV2/210069 (57–82)
      LoE: Ia: meta-analysis of RCTs; Ib: RCT; IIa controlled study without randomisation; IIb: quasi-experimental study (e.g., uncontrolled trial, one arm dose-response trial, etc.); III: observational studies (e.g., case–control, cohort, and cross-sectional studies); and IV: expert opinion.
      ES is the standard mean difference, i.e., the mean difference between a treatment and a control group divided by the SD of the difference. ES=0.2 is considered small, ES=0.5 is moderate, and ES>0.8 is large.

      Strength of recommendation (SOR)

      The SOR for each treatment proposition was based on the opinions of the guideline development group after taking into consideration the research evidence for efficacy, safety and cost-effectiveness of each treatment proposed, and the clinical expertise of the members of the guideline committee including such considerations as the experts' experience and perception of patient tolerance, acceptability and adherence to the treatment in question and their expert knowledge of any logistic issues involved in the administration of the treatment.
      Each member of the guideline development committee, except for RM, GN and WZ, was asked to indicate their SOR for each accepted proposition on a 0–100 mm visual analogue scale (VAS) after being provided with
      • (a)
        the list of accepted propositions in which the level of the research evidence for each proposition was indicated (Table I) according to the evidence hierarchy
        • Shekelle P.G.
        • Woolf S.H.
        • Eccles M.
        • Grimshaw J.
        Clinical guidelines: developing guidelines.
        (Table II),
        Table IILevel of Evidence (LoE)
        LoEType of evidence
        IaMetaanalysis of Randomized Controlled Trials
        IbAt least one Randomized Controlled Trial
        IIaAt least one well-designed controlled study, but without randomisation
        IIbAt least one well-designed quasi-experimental study
        IIIAt least one non-experimental descriptive study (e.g., comparative, correlation or case–controlled study)
        IVExpert committee reports, opinions and/or experience of respected authorities
      • (b)
        the results of the critical appraisal of existing guidelines
        • Zhang W.
        • Moskowitz R.W.
        • Nuki G.
        • Abramson S.
        • Altman R.D.
        • Arden N.
        • et al.
        OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
        ,
      • (c)
        a summary of the systematic analysis of the research evidence from 2002 to 2006 (Ref. 
        • Zhang W.
        • Moskowitz R.W.
        • Nuki G.
        • Abramson S.
        • Altman R.D.
        • Arden N.
        • et al.
        OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
        , Table 5) including details of quality scores, ES for pain, function and stiffness, the NNT, the RR or OR and the cost per QALY for each modality of treatment proposed where these were available, and
      • (d)
        a first draft of this manuscript.
      Mean and standard error of the mean (s.e.m.) for the SOR for each proposition were calculated, with and without recusals for voting on individual propositions by individual members of the committee, where there was any possibility of a potential conflict of interest, and the results were expressed as means with 95% confidence limits.

      OARSI recommendations

      After six rounds of the Delphi exercise there was expert consensus for 25 recommendations for the treatment of hip and knee OA. These are summarised in Table I together with the level of evidence (LoE) supporting them, the ES for pain relief [ESpain 95% confidence interval (CI)], the extent of consensus (%) and the SOR (mean±2 s.e.m.) for each proposition. The treatment propositions recommended in Table I are grouped as general, non-pharmacological, pharmacological and surgical without ranking the recommendations for the order in which the treatments should be offered.

        General recommendations

      • 1.
        Optimal management of OA requires a combination of non-pharmacological and pharmacological modalities.
        SOR: 96% (95% CI 93–99)
      Combination of pharmacological and non-pharmacological treatments is frequently employed in clinical practise and is universally recommended in 12/12 existing guidelines for the management of hip and/or knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . Although there was 100% consensus and strong recommendation for combining pharmacological and non-pharmacological therapies following the Delphi exercise, this recommendation lacks evidence from RCTs with appropriate factorial design. It is largely based on expert opinion (LoE IV) and uncontrolled observations of additional benefit in RCTs and meta-analyses (MAs) of trials of non-pharmacological modalities of therapy (e.g., exercise
      • Petrella R.J.
      • Bartha C.
      Home based exercise therapy for older patients with knee osteoarthritis: a randomised controlled trial.
      • Roddy E.
      • Zhang W.
      • Doherty M.
      Aerobic walking or strengthening exercise for osteoarthritis of the knee? A systematic review.
      , weight reduction
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      • Messier S.P.
      • Loeser R.F.
      • Miller G.D.
      • Morgan T.M.
      • Rejeski W.J.
      • Sevick M.A.
      • et al.
      Exercise and dietary weight loss in overweight and obese older adults with knee osteoarthritis: the arthritis, diet, and activity promotion trial.
      , and education
      • Warsi A.
      • LaValley M.P.
      • Wang P.S.
      • Avorn J.
      • Solomon D.H.
      Arthritis self-management education programs: a meta-analysis of the effect on pain and disability.
      ) where all patients were receiving pharmacological treatment with analgesics or non-steroidal anti-inflammatory drugs (NSAIDs).

        Non-pharmacological modalities of treatment

      • 2.
        All patients with hip and knee OA should be given information access and education about the objectives of treatment and the importance of changes in lifestyle, exercise, pacing of activities, weight reduction, and other measures to unload the damaged joint(s). The initial focus should be on self-help and patient-driven treatments rather than on passive therapies delivered by health professionals. Subsequently emphasis should be placed on encouraging adherence to the regimen of non-pharmacological therapy.
        SOR: 97% (95% CI 95–99)
      Provision of information and overall patient education about the objectives of treatment and the importance of changes in lifestyle, exercise, pacing of activities, weight reduction and other measures to unload damaged joints is supported by two MAs
      • Warsi A.
      • LaValley M.P.
      • Wang P.S.
      • Avorn J.
      • Solomon D.H.
      Arthritis self-management education programs: a meta-analysis of the effect on pain and disability.
      • Chodosh J.
      • Morton S.C.
      • Mojica W.
      • Maglione M.
      • Suttorp M.J.
      • Hilton L.
      • et al.
      Meta-analysis: chronic disease self-management programs for older adults.
      (LoE Ia), but the ES for pain relief is small (0.06 95% CI 0.02, 0.10)
      • Chodosh J.
      • Morton S.C.
      • Mojica W.
      • Maglione M.
      • Suttorp M.J.
      • Hilton L.
      • et al.
      Meta-analysis: chronic disease self-management programs for older adults.
      and RCTs with an appropriate factorial design to assess the efficacy of individual components of the education programme have not been undertaken. Attempts to identify which components of self-management programmes contribute most to their efficacy by meta-regression analysis were unsuccessful
      • Chodosh J.
      • Morton S.C.
      • Mojica W.
      • Maglione M.
      • Suttorp M.J.
      • Hilton L.
      • et al.
      Meta-analysis: chronic disease self-management programs for older adults.
      • Warsi A.
      • Wang P.S.
      • LaValley M.P.
      • Avorn J.
      • Solomon D.H.
      Self-management education programs in chronic disease: a systematic review and methodological critique of the literature.
      • Chodosh J.
      • Morton S.C.
      • Suttorp M.J.
      • Shekelle P.J.
      Self-management education for osteoarthritis.
      . The recommendation that initial focus should be on self-help and patient-driven treatments rather than on passive therapies delivered by health professionals is based on expert opinion, common sense and economic considerations alone (LoE IV). There is, however, evidence from a number of RCTs of exercise therapy
      • WH Jr., Ettinger
      • Burns R.
      • Messier S.P.
      • Applegate W.
      • Rejeski W.J.
      • Morgan T.
      • et al.
      A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis. The Fitness, Arthritis and Seniors Trial (FAST).
      • Rejeski W.J.
      • Brawley L.R.
      • Ettinger W.
      • Morgan T.
      • Thompson C.
      Compliance to exercise therapy in older participants with knee osteoarthritis: implications for treating disability.
      • O'Reilly S.C.
      • Muir K.R.
      • Doherty M.
      Effectiveness of home exercise on pain and disability from osteoarthritis of the knee: a randomised, controlled trial.
      • Belza B.
      • Topolski T.
      • Kinne S.
      • Patrick D.L.
      • Ramsey S.D.
      Does adherence make a difference? Results from a community-based aquatic exercise program.
      • Thomas K.S.
      • Muir K.R.
      • Doherty M.
      • Jones A.C.
      • O'Reilly S.C.
      • Bassey E.J.
      Home-based exercise programme for knee pain and knee osteoarthritis: randomised controlled trial.
      (LoE Ib) to support the recommendation that subsequent emphasis should be placed on encouraging adherence to the regimen of non-pharmacological therapy.
      • 3.
        The clinical status of patients with hip or knee OA can be improved if patients are contacted regularly by phone.
        SOR: 66% (95% CI 57–75)
      The best evidence to suggest that monthly telephone contact by lay personnel aimed at promoting self-care for patients with OA knee could be associated with improvements in joint pain and physical function for up to a year comes from an RCT in 439 OA patients
      • Weinberger M.
      • Tierney W.M.
      • Booher P.
      • Katz B.P.
      Can the provision of information to patients with osteoarthritis improve functional status? A randomized, controlled trial.
      . Subsequent subgroup analysis showed that regular telephone contact was associated with pain relief (ES=0.65, P<0.01) even in a small group of 40 patients whose medical treatment with drugs and physical therapy remained stable
      • Rene J.
      • Weinberger M.
      • Mazzuca S.A.
      • Brandt K.D.
      • Katz B.P.
      Reduction of joint pain in patients with knee osteoarthritis who have received monthly telephone calls from lay personnel and whose medical treatment regimens have remained stable.
      , and telephone contact did not influence psycho-social outcomes such as morale, satisfaction with care, adherence to medication or social support
      • Weinberger M.
      • Tierney W.M.
      • Booher P.
      • Katz B.P.
      The impact of increased contact on psychosocial outcomes in patients with osteoarthritis: a randomized, controlled trial.
      . Overall the ES for pain relief and maintenance of physical function may be much smaller. While there are no published MAs of trials of telephone intervention alone, Warsi's MA of arthritis self-management programmes
      • Warsi A.
      • LaValley M.P.
      • Wang P.S.
      • Avorn J.
      • Solomon D.H.
      Arthritis self-management education programs: a meta-analysis of the effect on pain and disability.
      included three trials in patients with knee OA in which telephone contact was part of the package
      • WH Jr., Ettinger
      • Burns R.
      • Messier S.P.
      • Applegate W.
      • Rejeski W.J.
      • Morgan T.
      • et al.
      A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis. The Fitness, Arthritis and Seniors Trial (FAST).
      • Mazzuca S.A.
      • Brandt K.D.
      • Katz B.P.
      • Chambers M.
      • Byrd D.
      • Hanna M.
      Effects of self-care education on the health status of inner city patients with osteoarthritis of the knee.
      • Keefe F.J.
      • Caldwell D.S.
      • Williams D.A.
      • Gil K.M.
      • Mitchell D.
      • Martinez S.
      • et al.
      Pain coping skills training in the management of osteoarthritis knee pain. II Follow-up results.
      . Notwithstanding the difficulty of assessing the efficacy of individual components of the self-management strategy, two of these studies
      • Mazzuca S.A.
      • Brandt K.D.
      • Katz B.P.
      • Chambers M.
      • Byrd D.
      • Hanna M.
      Effects of self-care education on the health status of inner city patients with osteoarthritis of the knee.
      • Keefe F.J.
      • Caldwell D.S.
      • Williams D.A.
      • Gil K.M.
      • Mitchell D.
      • Martinez S.
      • et al.
      Pain coping skills training in the management of osteoarthritis knee pain. II Follow-up results.
      demonstrated much smaller, non-significant, effects on pain. The estimated ES
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      for the three trials was similar to the pooled ES for pain relief in 17 self-management programmes (ES=0.12, 95% CI 0.00–0.24)
      • Warsi A.
      • LaValley M.P.
      • Wang P.S.
      • Avorn J.
      • Solomon D.H.
      Arthritis self-management education programs: a meta-analysis of the effect on pain and disability.
      . The proposition that the clinical status of patients with hip OA can be improved if patients are contacted regularly by phone is based on expert opinion alone (LoE IV).
      • 4.
        Patients with symptomatic hip and knee OA may benefit from referral to a physical therapist for evaluation and instruction in appropriate exercises to reduce pain and improve functional capacity. This evaluation may result in provision of assistive devices such as canes and walkers, as appropriate.
        SOR: 89% (95% CI 82–96)
      The recommendation to refer patients with symptomatic hip or knee OA to a physical therapist is mainly supported by expert opinion (LoE IV). Referral to a physical therapist was strongly recommended by 100% of the expert panel and is also recommended in 5/5 of existing guidelines where referral for physiotherapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The recommendation to refer patients with symptomatic knee OA for physical therapy is supported by the results of three RCTs
      • Fransen M.
      • Crosbie J.
      • Edmonds J.
      Physical therapy is effective for patients with osteoarthritis of the knee: a randomized controlled clinical trial.
      • Deyle G.D.
      • Henderson N.E.
      • Matekel R.L.
      • Ryder M.G.
      • Garber M.B.
      • Allison S.C.
      Effectiveness of manual physical therapy and exercise in osteoarthritis of the knee. A randomized, controlled trial.
      • Deyle G.D.
      • Allison S.C.
      • Matekel R.L.
      • Ryder M.G.
      • Stang J.M.
      • Gohdes D.D.
      • et al.
      Physical therapy treatment effectiveness for osteoarthritis of the knee: a randomized comparison of supervised clinical exercise and manual therapy procedures versus a home exercise program.
      . One demonstrated significant short-term (8 weeks) improvements in pain, physical function and health-related quality of life
      • Fransen M.
      • Crosbie J.
      • Edmonds J.
      Physical therapy is effective for patients with osteoarthritis of the knee: a randomized controlled clinical trial.
      . Another showed improvements in WOMAC indices up to 1 year after referral for a 4 week treatment programme by a physical therapist
      • Deyle G.D.
      • Henderson N.E.
      • Matekel R.L.
      • Ryder M.G.
      • Garber M.B.
      • Allison S.C.
      Effectiveness of manual physical therapy and exercise in osteoarthritis of the knee. A randomized, controlled trial.
      ; and a third demonstrated improved clinical outcomes over and above a programme of home exercises
      • Deyle G.D.
      • Allison S.C.
      • Matekel R.L.
      • Ryder M.G.
      • Stang J.M.
      • Gohdes D.D.
      • et al.
      Physical therapy treatment effectiveness for osteoarthritis of the knee: a randomized comparison of supervised clinical exercise and manual therapy procedures versus a home exercise program.
      (LoE Ib). However two other RCTs of multimodal physiotherapy programmes, including patellar taping and exercises, showed no persistent benefits when compared with standard treatment without physical therapy
      • Quilty B.
      • Tucker M.
      • Campbell R.
      • Dieppe P.
      Physiotherapy, including quadriceps exercises and patellar taping, for knee osteoarthritis with predominant patello-femoral joint involvement: randomized controlled trial.
      or simulated placebo physical therapy treatment modalities
      • Bennell K.L.
      • Hinman R.S.
      • Metcalf B.R.
      • Buchbinder R.
      • McConnell J.
      • McColl G.
      • et al.
      Efficacy of physiotherapy management of knee joint osteoarthritis: a randomized, double-blind, placebo controlled trial.
      . There are no published RCTs of referral of patients with symptomatic hip OA for multimodal physical therapy.
      • 5.
        Patients with hip and knee OA should be encouraged to undertake, and continue to undertake, regular aerobic, muscle strengthening and range of motion exercises. For patients with symptomatic hip OA, exercises in water can be effective.
        SOR: 96% (95% CI 93–99)
      The recommendation that patients with OA knee should be encouraged to undertake regular aerobic walking exercises and home-based quadriceps muscle strengthening exercises is a core recommendation in 21/21 published guidelines
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      and is supported by a systematic review and MA of 13 RCTs
      • Roddy E.
      • Zhang W.
      • Doherty M.
      Aerobic walking or strengthening exercise for osteoarthritis of the knee? A systematic review.
      (LoE Ia). Pooled ESs for pain relief are in the moderate range for both aerobic (ES=0.52, 95% CI 0.34, 0.70) and muscle strengthening exercises (ES=0.32, 95% CI 0.23, 0.42) and pooled ESs for self-reported disability are 0.46 (95% CI 0.25, 0.67) for aerobic exercise and 0.32 (95% CI 0.23, 0.41) for quadriceps strengthening exercises. By contrast the recommendation that patients with hip OA continue to undertake regular aerobic, muscle strengthening and range of motion exercises is largely based on clinical expertise (LoE IV)
      • Roddy E.
      • Zhang W.
      • Doherty M.
      • Arden N.K.
      • Barlow J.
      • Birrell F.
      • et al.
      Evidence-based recommendations for the role of exercise in the management of osteoarthritis of the hip or knee—the MOVE consensus.
      . Evidence for pain relief (ES=0.25, 95% CI 0.02, 0.47)
      • Cochrane T.
      • Davey R.C.
      • Matthes Edwards S.M.
      Randomised controlled trial of the cost-effectiveness of water-based therapy for lower limb osteoarthritis.
      and improvement in stiffness (ES=0.17, 95% CI 0.05, 0.39)
      • Cochrane T.
      • Davey R.C.
      • Matthes Edwards S.M.
      Randomised controlled trial of the cost-effectiveness of water-based therapy for lower limb osteoarthritis.
      in patients with symptomatic hip OA following exercise in water comes from two RCTs
      • Cochrane T.
      • Davey R.C.
      • Matthes Edwards S.M.
      Randomised controlled trial of the cost-effectiveness of water-based therapy for lower limb osteoarthritis.
      • Stener-Victorin E.
      • Kruse-Smidje C.
      • Jung K.
      Comparison between electro-acupuncture and hydrotherapy, both in combination with patient education and patient education alone, on the symptomatic treatment of osteoarthritis of the hip.
      (LoE Ib).
      • 6.
        Patients with hip and knee OA, who are overweight, should be encouraged to lose weight and maintain their weight at a lower level.
        SOR: 96% (95% CI 92–100)
      Encouragement to lose weight and maintain weight at a lower level in overweight patients with lower limb OA was strongly recommended by all members of the guideline development group (100% Table I) and is a core recommendation in 13/14 existing guidelines for the management of lower limb OA where this modality of therapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . At the time of completing the systematic review of the published research evidence before 31st January 2006 the recommendation was supported by the results of two high quality RCTs
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      • Messier S.P.
      • Loeser R.F.
      • Miller G.D.
      • Morgan T.M.
      • Rejeski W.J.
      • Sevick M.A.
      • et al.
      Exercise and dietary weight loss in overweight and obese older adults with knee osteoarthritis: the arthritis, diet, and activity promotion trial.
      (LoE Ib). In patients with knee OA the ESs for relief of pain (ES=0.13, 95% CI −0.12, 0.38)
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      • Messier S.P.
      • Loeser R.F.
      • Miller G.D.
      • Morgan T.M.
      • Rejeski W.J.
      • Sevick M.A.
      • et al.
      Exercise and dietary weight loss in overweight and obese older adults with knee osteoarthritis: the arthritis, diet, and activity promotion trial.
      , stiffness (0.36 95% CI −0.08, 0.80)
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      and functional improvement (0.69 95% CI 0.24, 1.14)
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      were small to moderate with an NNT of 3 (95% CI 2, 9)
      • Christensen R.
      • Astrup A.
      • Bliddal H.
      Weight loss: the treatment of choice for knee osteoarthritis? A randomized trial.
      for a decrease in WOMAC scores of >50%, 8 weeks after commencing a low energy diet (3.4 MJ/day). The recommendation is further supported by the publication of a recent systematic review and MA
      • Christensen R.
      • Bartels E.M.
      • Astrup A.
      • Bliddal H.
      Effect of weight reduction in obese patients diagnosed with knee osteoarthritis: a systematic review and meta-analysis.
      of four RCTs with data on 454 patients with OA knee (LoE Ia). The pooled ESs for improvements in pain and physical disability are confirmed as small (0.20 95% CI 0, 0.39 and 0.23 95% CI 0.04, 0.42, respectively), with a mean weight reduction of 6.1 kg (range 4.7–7.6 kg). Meta-regression analysis demonstrated significant improvement in disability with weight loss>5% or at a rate of >0.24%/week. There are no published RCTs to confirm comparable benefits from weight loss in patients with hip OA. The recommendation that patients with hip OA should be encouraged to lose weight and maintain their weight at a lower level is based on expert opinion (LoE IV) and evidence of a relationship between obesity and hip OA in case–control studies
      • Lievense A.M.
      • Bierma-Zeinstra S.M.
      • Verhagen A.P.
      • van Baar M.E.
      • Velhaar J.A.
      • Koes B.W.
      Influence of obesity on the development of osteoarthritis of the hip: a systematic review.
      .
      • 7.
        Walking aids can reduce pain in patients with hip and knee OA. Patients should be given instruction in the optimal use of a cane or crutch in the contralateral hand. Frames or wheeled walkers are often preferable for those with bilateral disease.
        SOR: 90% (95% CI 84–96)
      Although there are no RCTs to support their use there was complete expert consensus for the proposition that walking aids can reduce pain in patients with hip and knee OA (LoE IV), and for the recommendation that patients should be given instruction in the optimal use of a cane or crutch in the contralateral hand. This is supported by kinematic studies of knee moments of force following the use of a cane in the contralateral hand in patients with knee OA
      • Chan G.N.Y.
      • Smith A.W.
      • Kirtley C.
      • Tsang W.W.N.
      Changes in knee moments with contralateral versus ipsilateral cane usage in females with knee osteoarthritis.
      , and earlier studies of the biomechanics of the hip following the use of a stick in the contralateral hand in patients with hip OA
      • Blount W.P.
      Don't throw away the cane.
      .There are data that show that up to 40% of patients with hip or knee OA own a cane
      • Van der Esch M.
      • Heijmans M.
      • Dekker J.
      Factors contributing to the possession and use of walking aids among persons with rheumatoid arthritis and osteoarthritis.
      and sticks or canes are recommended for patients with symptomatic knee OA in 11/11 existing guidelines
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      .
      • 8.
        In patients with knee OA and mild/moderate varus or valgus instability, a knee brace can reduce pain, improve stability and diminish the risk of falling.
        SOR: 76% (95% CI 69–83)
      Evidence that pain, stiffness and physical function are significantly improved using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and the McMaster Toronto arthritis patient preference questionnaire (MACTAR) with knee braces in patients with knee OA comes from a Cochrane review
      • Brouwer R.W.
      • Jakma T.S.
      • Verhagen A.P.
      • Verhaar J.A.
      • Bierma-Zeinstra S.M.
      Braces and orthoses for treating osteoarthritis of the knee.
      (LoE Ia) and a single RCT
      • Kirkley A.
      • Webster-Bogaert S.
      • Litchfield R.
      • Amendola A.
      • Macdonald A.
      • McCalden R.
      • et al.
      The effect of bracing on varus gonarthrosis.
      which compared the use of a valgus brace+medical treatment with a neoprene sleeve+medical treatment and medical treatment alone. Assessment at 6 months showed greater improvement in WOMAC scores with use of the valgus brace than the neoprene sleeve. Knee braces are recommended in 8/9 existing guidelines for the management of knee OA where this modality of treatment was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      .
      • 9.
        Every patient with hip or knee OA should receive advice concerning appropriate footwear. In patients with knee OA insoles can reduce pain and improve ambulation. Lateral wedged insoles can be of symptomatic benefit for some patients with medial tibio-femoral compartment OA.
        SOR: 77% (95% CI 66–88)
      The use of lateral wedged insoles for patients with medial tibio-femoral compartment OA is recommended in 12/13 existing guidelines for the management of knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The proposition that lateral wedged insoles can provide symptomatic benefit for patients with medial tibio-femoral compartment OA, as well as decreasing lateral thrust in the knee
      • Ogata K.
      • Yasunaga M.
      • Nomiyama H.
      The effect of wedged insoles on the thrust of osteoarthritic knees.
      , is supported by three observational studies
      • Ogata K.
      • Yasunaga M.
      • Nomiyama H.
      The effect of wedged insoles on the thrust of osteoarthritic knees.
      • Sasaki T.
      • Yasuda K.
      Clinical evaluation of the treatment of osteoarthritic knees using a newly designed wedged insole.
      • Keating E.M.
      • Faris P.M.
      • Ritter M.A.
      • Kane J.
      Use of lateral heel and sole wedges in the treatment of medial osteoarthritis of the knee.
      , but not by three RCTs
      • Toda Y.
      • Segal N.
      • Kato A.
      • Yamamoto S.
      • Irie M.
      Effect of a novel insole on the subtalar joint of patients with medial compartment osteoarthritis of the knee.
      • Maillefert J.F.
      • Hudry C.
      • Baron G.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis: a prospective randomized controlled study.
      • Pham T.
      • Maillefert J.F.
      • Hudry C.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis. A two-year prospective randomized controlled study.
      . Despite the fact that there was no symptomatic benefit (WOMAC pain, joint stiffness, and physical functioning subscales) at 6 months
      • Maillefert J.F.
      • Hudry C.
      • Baron G.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis: a prospective randomized controlled study.
      or 2 years
      • Pham T.
      • Maillefert J.F.
      • Hudry C.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis. A two-year prospective randomized controlled study.
      in a prospective RCT of laterally wedged insoles in 156 patients with medial femoro-tibial OA, NSAID usage was reduced and compliance was better in the treatment group. This was accepted by the investigators
      • Maillefert J.F.
      • Hudry C.
      • Baron G.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis: a prospective randomized controlled study.
      • Pham T.
      • Maillefert J.F.
      • Hudry C.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis. A two-year prospective randomized controlled study.
      and a systematic review
      • Brouwer R.W.
      • Jakma T.S.
      • Verhagen A.P.
      • Verhaar J.A.
      • Bierma-Zeinstra S.M.
      Braces and orthoses for treating osteoarthritis of the knee.
      as evidence supporting clinical benefit (LoE Ia). No structural protection was observed in this study after 2 years
      • Pham T.
      • Maillefert J.F.
      • Hudry C.
      • Kieffert P.
      • Bourgeois P.
      • Lechevalier D.
      • et al.
      Laterally elevated wedged insoles in the treatment of medial knee osteoarthritis. A two-year prospective randomized controlled study.
      . The recommendation that every patient with hip or knee OA should receive advice concerning appropriate footwear is based on expert opinion alone (LoE IV). There have been no controlled trials of footwear in patients with hip OA and no controlled trials to support the hypothesis
      • Dieppe P.
      • Buckwalter J.A.
      Management of limb joint osteoarthritis.
      that sports shoes or other footwear with shock absorbing soles provide symptomatic benefit in patients with lower limb OA (hip or knee) by reducing impact loads.
      • 10.
        Some thermal modalities may be effective for relieving symptoms in hip and knee OA.
        SOR: 64% (95% CI 60–68)
      Heat and cryotherapy are used very widely in the management of patients with OA. Heat can be administered by a variety of techniques including diathermy and the application of heat packs or immersion in warm water or wax baths, while cryotherapy is usually administered by application of ice packs or massage with ice. Thermotherapy of one kind or another is recommended in 7/10 existing guidelines where these modalities were considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . Supporting evidence is very limited. A single systematic review
      • Brosseau L.
      • Yonge K.A.
      • Robinson V.
      • Marchand S.
      • Judd M.
      • Wells G.
      • et al.
      Thermotherapy for treatment of osteoarthritis.
      (LoE Ia) analysed two RCTs of ice massage in 100 patients with knee OA
      • Yurtkuran M.
      • Kocagil T.
      TENS, electroacupuncture and ice massage: comparison of treatment for osteoarthritis of the knee.
      and ice packs or short wave diathermy in two groups of 15 and 17 patients with knee OA
      • Clarke R.G.
      • Willis L.A.
      • Stenner L.
      • Nichols P.J.R.
      Evaluation of physiotherapy in the treatment of osteoarthritis of the knee.
      . Massaging with ice for 20 min×5/week for 2 weeks resulted in clinically significant (29%) improvement in quadriceps strength (ES=1.03, 95% CI 0.44, 1.62) but had no clinically significant effect on the range of movement or on walking
      • Yurtkuran M.
      • Kocagil T.
      TENS, electroacupuncture and ice massage: comparison of treatment for osteoarthritis of the knee.
      . Application of ice packs×3/week for 3 weeks was followed by some improvement in pain (weighted mean difference, WMD −2.70 95% CI −5.52, 0.12)
      • Clarke R.G.
      • Willis L.A.
      • Stenner L.
      • Nichols P.J.R.
      Evaluation of physiotherapy in the treatment of osteoarthritis of the knee.
      , but this was not statistically significant. Short wave diathermy was not followed by any improvement in pain after 3 weeks and there was no evidence of clinical benefit following either modality of thermotherapy after 3 months
      • Clarke R.G.
      • Willis L.A.
      • Stenner L.
      • Nichols P.J.R.
      Evaluation of physiotherapy in the treatment of osteoarthritis of the knee.
      . There have been no controlled trials of thermotherapy for patients with hip OA.
      • 11.
        Transcutaneous electrical nerve stimulation (TENS) can help with short-term pain control in some patients with hip or knee OA.
        SOR: 58% (95% CI 45–72)
      TENS is a recommended treatment for relief of pain in 8/10 existing guidelines for the management of knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . Evidence for efficacy available to the OARSI treatment guidelines development group was summarised in a Cochrane systematic review published in 2000
      • Osiri M.
      • Welch V.
      • Brosseau L.
      • Shea B.
      • McGowan J.
      • Tugwell P.
      • et al.
      Transcutaneous electrical nerve stimulation for knee osteoarthritis.
      and a systematic review published in 2004
      • Brosseau L.
      Efficacy of transcutaneous electrical nerve stimulation for osteoarthritis of the lower extremities: a meta-analysis.
      (NNT=2, 95% CI 1, 5) (LoE Ia). The short-term efficacy of 2–4 weeks treatment with TENS in providing clinically significant pain relief in patients with knee OA has been subsequently confirmed in a recent systematic review and MA of seven RCTs involving 425 patients
      • Bjordal J.M.
      • Johnson M.I.
      • Lopes-Martins R.A.B.
      • Bogen B.
      • Chow R.
      • Llunggren A.E.
      Short-term efficacy of physical interventions in osteoarthritic knee pain. A systematic review and meta-analysis of randomised placebo-controlled trials.
      . Dose-dependent inhibition of nociceptive nerve transmission at a segmental level may provide a physiological rationale
      • Sluka K.A.
      • Walsh D.
      Transcutaneous electrical nerve stimulation: basic science mechanisms and clinical effectiveness.
      for the efficacy of TENS, and no serious adverse effects of therapy have been reported
      • Bjordal J.M.
      • Johnson M.I.
      • Lopes-Martins R.A.B.
      • Bogen B.
      • Chow R.
      • Llunggren A.E.
      Short-term efficacy of physical interventions in osteoarthritic knee pain. A systematic review and meta-analysis of randomised placebo-controlled trials.
      .
      • 12.
        Acupuncture may be of symptomatic benefit in patients with knee OA.
        SOR: 59% (95% CI 47–71)
      Acupuncture is recommended as a modality of therapy for the symptomatic treatment of patients with OA knee or hip in 5/8 existing guidelines in which it was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      , and its recommendation achieved a 69% consensus following the Delphi exercise. A summary of the evidence for its clinical efficacy in lower limb joint OA which was available to the OARSI treatment guideline development group showed moderate ESs for pain (ES=0.51, 95% CI 0.23, 0.79), stiffness (ES=0.41, 95% CI 0.13, 0.69) and function (ES=0.51, 95% CI 0.23, 0.79) with an NNT of 4 (95% CI 3, 9) for clinically significant relief of pain
      • Witt C.
      • Selim D.
      • Reinhold T.
      • Jena S.
      • Brinkhaus B.
      • Liecker B.
      • et al.
      Cost-effectiveness of acupuncture in patients with headache, low back pain and osteoarthritis of the hip and the knee.
      (LoE Ib). An earlier (2001) systematic review of the evidence for the efficacy of acupuncture in OA knee which included seven RCTs and 393 patients suggested that real acupuncture was more effective than sham acupuncture for relief of pain (LoE Ia) but the evidence with regard to improvement in function was inconclusive
      • Ezzo J.
      • Hadhazy V.
      • Birch S.
      • Lixing L.
      • Kaplan G.
      • Hochberg M.
      • et al.
      Acupuncture for osteoarthritis of the knee: a systematic review.
      . A very recent RCT in 352 patients with knee OA showed very small, statistically significant, improvements in pain intensity in patients 2 and 6 weeks following true acupuncture but the addition of acupuncture to a course of advice and exercises delivered by physiotherapists provided no additional improvement in the WOMAC index pain subscale at 6 months
      • Foster N.E.
      • Thomas E.
      • Barlas P.
      • Hill J.C.
      • Young J.
      • Mason E.
      • et al.
      Acupuncture as an adjunct to exercise based physiotherapy for osteoarthritis of the knee: randomised controlled trial.
      .

        Pharmacological modalities of treatment

      • 13.
        Acetaminophen (paracetamol) (up to 4 g/day) can be an effective initial oral analgesic for treatment of mild to moderate pain in patients with knee or hip OA. In the absence of an adequate response, or in the presence of severe pain and/or inflammation, alternative pharmacologic therapy should be considered based on relative efficacy and safety, as well as concomitant medications and co-morbidities.
        SOR: 92% (95% CI 88–99)
      Acetaminophen (paracetamol) is a core recommendation for use as an analgesic in 16/16 existing guidelines for the management of hip or knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . Current European (EULAR) recommendations for the management of hip
      • Zhang W.
      • Doherty M.
      • Arden N.
      • Bannwarth B.
      • Bijlsma J.
      • Gunther K.P.
      • et al.
      EULAR evidence based recommendations for the management of hip osteoarthritis: report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT).
      and knee
      • Jordan K.M.
      • Arden N.K.
      • Doherty M.
      • Bannwarth B.
      • Bijlsma J.W.
      • Dieppe P.
      • et al.
      EULAR recommendations 2003: an evidence based approach to the management of knee osteoarthritis: Report of a Task Force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT).
      OA suggest that, because of its safety and efficacy, doses of up to 4 g/day should be the oral analgesic of first choice for mild/moderate pain, and if successful, should be used as the preferred long-term oral analgesic. However, in recent years both the efficacy
      • Case J.P.
      • Baliunas A.J.
      • Black J.A.
      Lack of efficacy of acetaminophen in treating symptomatic knee osteoarthritis: a randomized, double-blind, placebo controlled comparison trial wit diclofenac sodium.
      and the safety
      • Garcia Roderiguez L.A.
      • Hernandez-Diaz S.
      Risk of upper gastrointestinal complications among users of acetaminophen and non-steroidal anti-inflammatory drugs.
      • Rahme E.
      • Pettitt D.
      • LeLorier J.
      Determinants and sequelae associated with utilization of acetaminophen versus traditional nonsteroidal antiinflammatory drugs in an elderly population.
      of long-term use of acetaminophen at this dose have been questioned. Evidence for efficacy available to the OARSI treatment guideline development committee was summarised in a Cochrane systematic review
      • Towheed T.E.
      • Hochberg M.C.
      • Judd M.G.
      • Wells G.
      Acetaminophen for osteoarthritis.
      largely based on a single RCT published before July 2002 and an MA of 10 RCTs published in 2004
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      with data from 1712 OA patients (LoE Ia). Efficacy was confirmed but the ES was small (ES=0.21, 95% CI 0.02, 0.4)
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      . A more recently updated Cochrane systematic review published in 2006
      • Towheed T.E.
      • Maxwell L.
      • Judd M.G.
      • Catton M.
      • Wells G.
      Acetaminophen for osteoarthritis.
      included data from 5986 patients in 15 RCTs (7 vs placebo and 10 vs NSAIDs). Acetaminophen was superior to placebo in 5/7 trials and pooled analysis of data on overall pain using multiple methods showed a statistically significant, but very small, reduction in pain (ES=0.13, 95% CI 0.04, 0.22) which is of questionable clinical significance. The NNT to achieve an improvement in pain ranged from 2 (1, 2)
      • Towheed T.E.
      • Hochberg M.C.
      • Judd M.G.
      • Wells G.
      Acetaminophen for osteoarthritis.
      in the earlier systematic review to 4–16
      • Towheed T.E.
      • Maxwell L.
      • Judd M.G.
      • Catton M.
      • Wells G.
      Acetaminophen for osteoarthritis.
      in the later MA. There was no significant difference in toxicity between acetaminophen and placebo in these short-term trials (RR=1.02, 95% CI 0.89, 1.87)
      • Towheed T.E.
      • Maxwell L.
      • Judd M.G.
      • Catton M.
      • Wells G.
      Acetaminophen for osteoarthritis.
      . The evidence for possible gastrointestinal (GI) and renal toxicity with long-term treatment with acetaminophen 4 g/day, reviewed in the first part of this report
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      , remains equivocal. The RR for upper GI bleeding or perforation ranged from RR 1.2 (95% CI 0.8, 1.7)
      • Lewis S.C.
      • Langman M.J.S.
      • Laporte J.-R.
      • Matthres N.S.
      • Rawlins M.D.
      • Wiholm B.-E.
      Dose–response relationships between individual nonaspirin nonsteroidal anti-inflammatory drugs (NSAIDs) and serious upper gastrointestinal bleeding: a meta-analysis based on individual patient data.
      in a MA of three case–control studies with individual patient data to RR 3.6 (95% CI 2.60, 5.10)
      • Garcia Roderiguez L.A.
      • Hernandez-Diaz S.
      Risk of upper gastrointestinal complications among users of acetaminophen and non-steroidal anti-inflammatory drugs.
      in a case–control study using the UK General Practice Research Database; and the RR of renal insufficiency ranged from RR 0.83 (95% CI 0.50, 1.39) in one cohort study (CS)
      • Rexrode K.M.
      • Buring J.E.
      • Glynn R.J.
      • Stampfer M.J.
      • Youngman L.D.
      • Gaziano J.M.
      Analgesic use and renal function in men.
      to RR 2.5 (95% CI 1.7, 2.6) in a case–control comparison
      • Fored C.M.
      • Ejerblad E.
      • Lindblad P.
      • Fryzek J.P.
      • Dickman P.W.
      • Signorello L.B.
      • et al.
      Acetaminophen, aspirin, and chronic renal failure.
      .
      • 14.
        In patients with symptomatic hip or knee OA, non-steroidal anti-inflammatory drugs (NSAIDs) should be used at the lowest effective dose but their long-term use should be avoided if possible. In patients with increased GI risk, either a COX-2 selective agent or a non-selective NSAID with co-prescription of a proton pump inhibitor (PPI) or misoprostol for gastroprotection may be considered, but NSAIDs, including both non-selective and COX-2 selective agents, should be used with caution in patients with cardiovascular (CV) risk factors.
        SOR: 93% (95% CI 88–99)
      The use of oral NSAIDs with misoprostol or a PPI for gastroprotection is recommended in 8/8 existing guidelines for the management of hip or knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      and the use of selective COX-2 inhibitors is recommended in all 11 of the guidelines where this modality of therapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . A telephone survey of 1149 patients with OA in the UK in 2003 revealed that only 15% were taking paracetamol, while 32% were taking non-selective NSAIDs and 18% COX-2 selective drugs for analgesia
      TNS/arthritis care survey.
      . There is evidence that NSAIDs can be effective in reducing pain in patients with OA knee and hip (LoE Ia). A 2004 MA of 23 short-term, placebo-controlled RCTs of NSAIDs, including COX-2 selective agents in >10000 patients with knee OA, showed that the ES for pain reduction was 0.32 (95% CI 0.24, 0.39)
      • Bjordal J.M.
      • Ljunggren A.E.
      • Klovning A.
      • Slordal L.
      Non-steroidal anti-inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials.
      . However in 10 trials that did not exclude non-responders where the outcomes were more homogeneous the ES for pain reduction was smaller (ES=0.23, 95% CI 0.15, 0.31)
      • Bjordal J.M.
      • Ljunggren A.E.
      • Klovning A.
      • Slordal L.
      Non-steroidal anti-inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials.
      . Evidence that NSAIDs are superior to acetaminophen for pain relief in patients with lower limb joint OA is available from another 2004 MA of RCTs
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      (ES=0.20, 95% CI 0.10, 0.30). The clinical response rate was higher (RR=1.24, 95% CI 1.08, 1.41) and the number of patients preferring NSAIDs to acetaminophen was considerably greater (RR=2.46, 95% CI 1.51, 4.12)
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      . The ESs for pain relief in short-term trials are, however, less than 0.4, which has been suggested as the minimum to be of any clinical importance
      • Angst F.
      • Aesclimann A.
      • Michel B.A.
      • Stucki G.
      Minimal clinically important rehabilitation effects in patients with osteoarthritis of the lower extremities.
      .
      There is abundant evidence that NSAIDs are associated with more adverse effects than acetaminophen in short-term trials. The 2004 MA
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      showed that NSAIDs were associated with GI discomfort more frequently than acetaminophen (RR=1.35, 95% CI 1.05, 1.75) and this was confirmed in the more recent Cochrane systematic review of short-term RCTs (RR=1.47, 95% CI 1.08, 2.00)
      • Towheed T.E.
      • Maxwell L.
      • Judd M.G.
      • Catton M.
      • Wells G.
      Acetaminophen for osteoarthritis.
      . More importantly NSAIDs can cause serious GI complications such as peptic ulcers, perforations and bleeds (PUBS) and this risk increases with age, concurrent use of other medications, and probably with the duration of therapy
      • Tramer M.R.
      • Moore R.A.
      • Reynolds D.J.
      • McQuay H.J.
      Quantitative estimation of rare adverse events which follow a biological progression: a new model applied to chronic NSAID use.
      . A MA of severe upper GI complications of NSAIDs showed an OR of 5.36 (95% CI 1.79, 16.1) in 16 NSAID vs placebo trials in 4431 patients and a pooled OR for PUBS of 3.0 (95% CI 2.5, 3.7) in 23 case–control studies in 25,732 patients
      • Ofman J.J.
      • MacLean C.H.
      • Straus W.L.
      • Morton S.C.
      • Berger M.L.
      • Roth E.A.
      • et al.
      A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs.
      . The pooled RR of PUBS from nine cohort studies representing >750,000 person years of drug exposure was 2.7 (95% CI 2.1, 3.5)
      • Ofman J.J.
      • MacLean C.H.
      • Straus W.L.
      • Morton S.C.
      • Berger M.L.
      • Roth E.A.
      • et al.
      A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs.
      . The recommendation that in patients with increased GI risk, either a COX-2 selective agent or a non-selective NSAID with co-prescription of a PPI or misoprostol for gastroprotection should be considered is supported by evidence from a systematic review of 112 RCTs which included nearly 75,000 patients
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      (LoE Ia). The RRs for symptomatic ulcers and serious GI complications with these different strategies are shown in Table III. There was no evidence for similar gastroprotection with H2 receptor antagonists and treatment with misoprostol is associated with an increased risk of diarrhoea (RR=1.81, 95% CI 1.52, 2.61)
      • Capurso L.
      • Koch M.
      Prevention of NSAID-induced gastric lesions: H2 antagonists or misoprostol? A meta-analysis of controlled clinical studies.
      and the GI protection that is associated with the use of COX-2 selective agents is largely lost when low-dose aspirin is administered concurrently for CV prophylaxis
      • Schnitzer T.J.
      • Burmester G.R.
      • Mysler E.
      • Hochberg M.C.
      • Doherty M.
      • Ehrsam E.
      • TARGET Study Group
      • et al.
      Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial.
      .
      Table IIIRelative risk of GI adverse events associated with NSAIDs and strategies for their prevention
      Intervention
      Compared with placebo/non-exposure unless otherwise stated.
      Adverse eventsRR/OR (95% CI)Evidence
      AcetaminophenGI discomfort0.80 (0.27, 2.37)MA of RCTs
      • Zhang W.
      • Jones A.
      • Doherty M.
      Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials.
      GI perforation/bleed3.60 (2.60, 5.10)CC
      • Garcia Roderiguez L.A.
      • Hernandez-Diaz S.
      Risk of upper gastrointestinal complications among users of acetaminophen and non-steroidal anti-inflammatory drugs.
      GI bleeding1.2 (0.8, 1.7)MA of CCs
      • Lewis S.C.
      • Langman M.J.S.
      • Laporte J.-R.
      • Matthres N.S.
      • Rawlins M.D.
      • Wiholm B.-E.
      Dose–response relationships between individual nonaspirin nonsteroidal anti-inflammatory drugs (NSAIDs) and serious upper gastrointestinal bleeding: a meta-analysis based on individual patient data.
      NSAIDsGI perforation/ulcer/bleed5.36 (1.79, 16.10)MA of RCTs
      • Ofman J.J.
      • MacLean C.H.
      • Straus W.L.
      • Morton S.C.
      • Berger M.L.
      • Roth E.A.
      • et al.
      A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs.
      2.70 (2.10, 3.50)MA of CSs
      • Ofman J.J.
      • MacLean C.H.
      • Straus W.L.
      • Morton S.C.
      • Berger M.L.
      • Roth E.A.
      • et al.
      A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs.
      3.00 (2.70, 3.70)MA of CCs
      • Ofman J.J.
      • MacLean C.H.
      • Straus W.L.
      • Morton S.C.
      • Berger M.L.
      • Roth E.A.
      • et al.
      A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs.
      Topical NSAIDsGI events0.81 (0.43, 1.56)MA of RCTs
      • Lin J.
      • Zhang W.
      • Jones A.
      • Doherty M.
      Efficacy of topical NSAIDs in the treatment of osteoarthritis: a meta-analysis of randomized controlled trials.
      GI bleed/perforation1.45 (0.84, 2.50)CC
      • Evans J.M.M.
      • McMahon A.D.
      • McGilchrist M.M.
      • White G.
      • Murray F.E.
      • McDevitt D.G.
      • et al.
      Topical non-steroidal anti-inflammatory drugs and admission to hospital for upper gastrointestinal bleeding and perforation: a record linkage case–control study.
      H2 blocker+NSAID vs NSAIDSerious GI complications0.33 (0.01, 8.14)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      Symptomatic ulcers1.46 (0.06, 35.53)
      PPI+NSAID vs NSAIDSerious GI complications0.46 (0.07, 2.92)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      Symptomatic ulcer0.09 (0.02, 0.47)
      Misoprostol+NSAID vs NSAIDSerous GI complications0.57 (0.36, 0.91)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      Symptomatic ulcers0.36 (0.20, 0.67)
      Diarrhoea1.81 (1.52, 2.61)MA of RCTs
      • Capurso L.
      • Koch M.
      Prevention of NSAID-induced gastric lesions: H2 antagonists or misoprostol? A meta-analysis of controlled clinical studies.
      COX-2 inhibitors vs NSAIDSerious GI complications0.55 (0.38, 0.80)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      Symptomatic ulcers0.49 (0.38, 0.62)
      RR: relative risk; OR: odds ratio; CI: confidence interval; GI: gastrointestinal; NSAID: non-steroidal anti-inflammatory drug; H2 blockers: histamine type 2 receptor antagonists.
      Compared with placebo/non-exposure unless otherwise stated.
      What is the evidence to support the recommendation that NSAIDs, including both non-selective and COX-2 selective agents, should be used with caution in patients with CV risk factors? Following the withdrawal of the COX-2 selective NSAID rofecoxib in 2004 because of an increased RR of thrombotic CV events including myocardial infarction and stroke in a colorectal adenoma chemoprevention trial
      • Bresalier R.S.
      • Sandler R.S.
      • Quan H.
      • Bolognese J.A.
      • Oxenius B.
      • Horgan K.
      • et al.
      Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial.
      a number of RCTs and systematic reviews of the CV safety of other COX-2 selective and non-selective NSAIDs have been undertaken
      • Hernandez-Diaz S.
      • Varas-Lorenzo C.
      • Garcia Rodriguez L.A.
      Non-steroidal antiinflammatory drugs and the risk of acute myocardial infarction.
      • Caldwell B.
      • Aldington S.
      • Weatherall M.
      • Shirtcliffe P.
      • Beasley R.
      Risk of cardiovascular events and celecoxib: a systematic review and meta-analysis.
      • Juni P.
      • Nartey L.
      • Reichenbach S.
      • Sterchi R.
      • Dieppe P.A.
      • Egger M.
      Risk of cardiovascular events and rofecoxib: cumulative meta-analysis.
      • Aldington S.
      • Shirtcliffe P.
      • Weatherall M.
      • Beasley R.
      Increased risk of cardiovascular events with parecoxib/valdecoxib: a systematic review and meta-analysis.
      . Table IV shows the RRs for CV events in patients treated with COX-2 selective and non-selective NSAIDs. While the increased risk of CV adverse events with rofecoxib was confirmed, similar CV toxicity was not seen consistently with celecoxib or valdecoxib and the overall CV risk associated with COX-2 selective inhibitors was not significantly greater than that associated with conventional non-selective NSAIDs (RR=1.19, 95% CI 0.80, 1.75)
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      . This has been borne out in the more recent 2006 systematic review and MA of atherothrombotic complications of COX-2 selective and non-selective NSAIDs
      • Kearney P.M.
      • Baigent C.
      • Godwin J.
      • Halls H.
      • Emberson J.R.
      • Patrono P.
      Do selective cyclo-oxygenase inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials.
      . The incidence of serious vascular events was 1% per annum in patients treated with COX-2 selective agents compared with 0.9% in those on traditional NSAIDs (RR=1.16, 95% CI 0.97, 1.38)
      • Kearney P.M.
      • Baigent C.
      • Godwin J.
      • Halls H.
      • Emberson J.R.
      • Patrono P.
      Do selective cyclo-oxygenase inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials.
      . There was, however, some heterogeneity in risk among the traditional NSAIDs with a modest increase in risk of CV events with ibuprofen (RR=1.51, 95% CI 0.96, 2.37) and diclofenac (RR=1.63, 95% CI 1.12, 2.37) but not with naproxen (RR=0.92, 95% CI 0.67, 1.26)
      • Kearney P.M.
      • Baigent C.
      • Godwin J.
      • Halls H.
      • Emberson J.R.
      • Patrono P.
      Do selective cyclo-oxygenase inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials.
      . The current advice

      EMEA/CHMP/410051/2006.

      from the European Agency for the Evaluation of Medicinal Products (EMEA) is that COX-2 selective NSAIDs are contraindicated in patients with ischaemic heart disease or stroke and that prescribers should exercise caution when prescribing COX-2 inhibitors for patients with risk factors for heart disease, such as hypertension, hyperlipidaemia, diabetes and smoking, as well as for patients with peripheral arterial disease. In the USA all marketed prescription NSAIDs, both non-selective and COX-2 selective carry a boxed warning about their potential for causing CV events and GI bleeding.
      Table IVRelative risks of CV and renal adverse events associated with COX-2 selective and non-selective NSAIDs
      Intervention
      Compared with placebo/non-exposure unless otherwise stated.
      Adverse eventsRR/OR (95% CI)Evidence
      AcetaminophenRenal failure0.83 (0.50, 1.39)CS
      • Rexrode K.M.
      • Buring J.E.
      • Glynn R.J.
      • Stampfer M.J.
      • Youngman L.D.
      • Gaziano J.M.
      Analgesic use and renal function in men.
      2.5 (1.7, 3.6)CC
      • Fored C.M.
      • Ejerblad E.
      • Lindblad P.
      • Fryzek J.P.
      • Dickman P.W.
      • Signorello L.B.
      • et al.
      Acetaminophen, aspirin, and chronic renal failure.
      NSAIDsMyocardial infarction1.09 (1.02, 1.15)MA of CSs
      • Hernandez-Diaz S.
      • Varas-Lorenzo C.
      • Garcia Rodriguez L.A.
      Non-steroidal antiinflammatory drugs and the risk of acute myocardial infarction.
      H2 blocker+NSAID vs NSAIDSerious CV or renal events0.53 (0.08, 3.46)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      PPI+NSAID vs NSAIDSerious CV or renal events0.78 (0.10, 6.26)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      Misoprostol+NSAID vs NSAIDSerious CV or renal events1.78 (0.26, 12.07)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
      COX-2 inhibitors
      Coxibs vs NSAIDSerious CV or renal events1.19 (0.80, 1.75)MA of RCTs
      • Hooper L.
      • Brown T.J.
      • Elliott R.
      • Payne K.
      • Roberts C.
      • Symmons D.
      The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review.
       CelecoxibMyocardial infarction2.26 (1.0, 5.1)MA of RCTs
      • Caldwell B.
      • Aldington S.
      • Weatherall M.
      • Shirtcliffe P.
      • Beasley R.
      Risk of cardiovascular events and celecoxib: a systematic review and meta-analysis.
      0.97 (0.86, 1.08)MA of CSs & CCs
      • Hernandez-Diaz S.
      • Varas-Lorenzo C.
      • Garcia Rodriguez L.A.
      Non-steroidal antiinflammatory drugs and the risk of acute myocardial infarction.
       RofecoxibMyocardial infarction2.24 (1.24, 4.02)MA of RCTs
      • Juni P.
      • Nartey L.
      • Reichenbach S.
      • Sterchi R.
      • Dieppe P.A.
      • Egger M.
      Risk of cardiovascular events and rofecoxib: cumulative meta-analysis.
      1.27 (1.12, 1.44)MA of CSs and CCs
      • Hernandez-Diaz S.
      • Varas-Lorenzo C.
      • Garcia Rodriguez L.A.
      Non-steroidal antiinflammatory drugs and the risk of acute myocardial infarction.
       ValdecoxibCV events2.3 (1.1, 4.7)MA of RCTs
      • Aldington S.
      • Shirtcliffe P.
      • Weatherall M.
      • Beasley R.
      Increased risk of cardiovascular events with parecoxib/valdecoxib: a systematic review and meta-analysis.
      CV: cardiovascular; please see the footnotes of Table III for other abbreviations.
      Compared with placebo/non-exposure unless otherwise stated.
      The EMEA also advises
      • Prescribers and patients should continue to use NSAIDs at the lowest effective dose for the shortest duration to control symptoms.
      • Prescribers should continue to chose any NSAID on the basis of the overall safety profile of the product, as set out in the product information, and the patient's individual risk factors.
      • Prescribers should not switch between NSAIDs without careful consideration of the overall safety profile of the products and the patient's individual risk factors, as well as the patient's preferences.
      • 15.
        Topical NSAIDs and capsaicin can be effective as adjunctives and alternatives to oral analgesic/anti-inflammatory agents in knee OA.
        SOR: 85% (95% CI 75–95)
      Topical NSAIDs are widely used as adjunctive or alternative therapy by patients with OA knee and are recommended in 7/9 existing guidelines where this modality of therapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . A MA of 13 RCTs, including 1983 patients with hand as well as knee OA, undertaken in 2004 confirmed that topical NSAIDs were superior to placebo in relieving pain and stiffness and in improving function (LoE Ia)
      • Lin J.
      • Zhang W.
      • Jones A.
      • Doherty M.
      Efficacy of topical NSAIDs in the treatment of osteoarthritis: a meta-analysis of randomized controlled trials.
      . Efficacy for pain relief was only apparent in the first 2 weeks of treatment with ESs of 0.41 (95% CI 0.16, 0.66) in week 1 and 0.40 (95% CI 0.15, 0.65) in week 2 but topical NSAIDs are less effective than oral NSAIDs in the first week of treatment. The NNT for topical NSAIDs was only 3 (95% CI 2, 4) but placebo effects may be large with all topical therapies. The ESs for improvement in stiffness and in function were 0.49 (95% CI 0.17, 0.80) and 0.36 (95% CI 0.24, 0.48), respectively. However the MA showed evidence of statistically significant asymmetry of a funnel plot
      • Egger M.
      • Smith J.D.
      • Schneider M.
      • Minder C.
      Bias in meta-analysis detected by a simple graphical test.
      suggesting the possibility of publication bias with under reporting of negative studies and consequent overestimation of the benefits of topical NSAIDs. This MA provided no trial evidence to support long-term use of topical NSAIDs in knee OA but there was some heterogeneity of efficacy between preparations and a more recent MA did demonstrate a small pooled effect (ESpain=0.28, 95% CI 0.14, 0.42)
      • Biswal S.
      • Medhi B.
      • Panhi P.
      Long-term efficacy of nonsteroidal anti-inflammatory drugs in knee osteoarthritis: meta-analysis of randomized placebo controlled trials.
      . Overall topical NSAIDs are safe with no more side effects than placebo
      • Lin J.
      • Zhang W.
      • Jones A.
      • Doherty M.
      Efficacy of topical NSAIDs in the treatment of osteoarthritis: a meta-analysis of randomized controlled trials.
      . GI side effects are less likely than they are with oral NSAIDs
      • Lin J.
      • Zhang W.
      • Jones A.
      • Doherty M.
      Efficacy of topical NSAIDs in the treatment of osteoarthritis: a meta-analysis of randomized controlled trials.
      • Egger M.
      • Smith J.D.
      • Schneider M.
      • Minder C.
      Bias in meta-analysis detected by a simple graphical test.
      and there was no evidence that they could be a cause of upper GI perforation or bleeds in a large case–control study
      • Evans J.M.M.
      • McMahon A.D.
      • McGilchrist M.M.
      • White G.
      • Murray F.E.
      • McDevitt D.G.
      • et al.
      Topical non-steroidal anti-inflammatory drugs and admission to hospital for upper gastrointestinal bleeding and perforation: a record linkage case–control study.
      (Table III). However local reactions such as itching, burning and rashes are more frequent
      • Lin J.
      • Zhang W.
      • Jones A.
      • Doherty M.
      Efficacy of topical NSAIDs in the treatment of osteoarthritis: a meta-analysis of randomized controlled trials.
      .
      Topical capsaicin creams contain a lipophilic alkaloid extracted from chilli peppers which activates and sensitises peripheral c-nociceptors by binding and activating the transient receptor potential vanilloid type 1 (TRPV1) cation channel
      • Baron R.
      Capsaicin and nociception: from basic mechanisms to novel drugs.
      . Paradoxically, although the application of capsaicin to the skin causes burning pain at the site of application, it can also be an effective topical analgesic which is recommended as an alternative or adjunctive treatment for knee OA in 8/9 existing treatment guidelines where this modality of therapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . Evidence for the efficacy of topical capsaicin (0.025% cream×4 daily) in patients with knee OA is supported by an MA of RCTs of topical capsaicin in the treatment of chronic painful conditions
      • Zhang W.Y.
      • Li Wan Po A.
      The effectiveness of topically applied capsaicin. A meta-analysis.
      (LoE Ia). This included a single placebo-controlled trial in 70 patients with knee OA
      • Deal C.L.
      • Schnitzer T.J.
      • Lipstein E.
      • Seibold J.R.
      • Stevens R.M.
      • Levy M.D.
      • et al.
      Treatment of arthritis with topical capsaicin: a double-blind trial.
      as well as two RCTs in patients with hand OA. The mean reduction in pain was 33% with an NNT of 4 (95% CI 3, 5) after 4 weeks of therapy but adequate blinding is not possible in trials with this agent. Treatment with topical capsaicin is safe but 40% of patients are troubled by local burning, stinging or erythema.
      • 16.
        Intra-articular (IA) injections with corticosteroids can be used in the treatment of hip or knee OA, and should be considered particularly when patients have moderate to severe pain not responding satisfactorily to oral analgesic/anti-inflammatory agents and in patients with symptomatic knee OA with effusions or other physical signs of local inflammation.
        SOR: 78% (95% CI 61–95)
      IA injections of corticosteroids have been widely used as adjunctive therapy in the treatment of patients with knee OA for more than 50 years
      • Miller J.H.
      • White J.
      • Norton T.H.
      The value of intra-articular injections in osteoarthritis of the knee.
      , and are recommended as a treatment option in 11/13 of existing treatment guidelines where this modality of therapy was considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The efficacy of IA steroid injections in patients with knee OA is well supported by evidence from a 2005 Cochrane systematic review
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Intraarticular corticosteroid for treatment of osteoarthritis of the knee.
      (LoE Ia), subsequently updated in 2006
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Intraarticular corticosteroid for treatment of osteoarthritis of the knee.
      , which examined data from 13 placebo-controlled RCTs. The ES for relief of pain was in the moderate range (ES=0.72, CI 0.42, 1.02) with an NNT of 4 (95% CI 2, 11) at 2 and 3 weeks after injection but function was not significantly improved (ES=0.06, 95% CI −0.17, 0.30) and evidence for relief of pain 4 and 24 weeks post-injection was lacking
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Intraarticular corticosteroid for treatment of osteoarthritis of the knee.
      . Some RCTs have demonstrated better outcomes in patients with synovial effusions
      • Gaffney K.
      • Ledingham J.
      • Perry J.D.
      Intra-articular triamcinolone hexacetonide in knee osteoarthritis: factors influencing the clinical response.
      but others have not found that clinical signs of inflammation or the presence of a joint effusion
      • Dieppe P.A.
      • Sathapatayavongs B.
      • Jones H.E.
      • Bacon P.A.
      • Ring E.F.
      Intra-articular steroids in osteoarthritis.
      • Jones A.
      • Doherty M.
      Intra-articular corticosteroids are effective in osteoarthritis but there are no clinical predictors of response.
      are predictors of a good clinical response; suggesting that IA steroid injections should not be restricted to patients with physical signs of inflammation and/or joint effusion. A single RCT
      • Valtonen E.J.
      Clinical comparison of triamcinolonehexacetonide and betamethasone in the treatment of osteoarthrosis of the knee joint.
      in 42 patients with knee OA with signs of inflammation showed that IA injections of 20 mg of triamcinolone hexacetonide were superior to 6 mg of a betamethasone acetate/bisodium phosphate combination for the number of patients reporting pain reduction up to 4 weeks after injection (RR=2, 95% CI 1.10, 3.63) but the number of head to head comparisons between different IA corticosteroid preparations is too few to support any evidence-based recommendations for a particular preparation.
      By contrast the evidence to support the recommendation for IA steroid injection in patients with OA hip is mainly limited to two RCTs
      • Flanagan J.
      • Casale F.F.
      • Thomas D.L.
      • Desai K.B.
      Intra-articular injection for pain relief in patients awaiting hip replacement.
      • Kullenberg B.
      • Runesson R.
      • Tuvhag R.
      • Olsson C.
      • Resch S.
      Intra-articular corticosteroid injection: pain relief in osteoarthritis of the hip?.
      (LoE Ib) and two uncontrolled cohort studies
      • Plant M.J.
      • Borg A.A.
      • Dziedzic K.
      • Saklatvala J.
      • Dawes P.T.
      Radiographic patterns and response to corticosteroid hip injection.
      • Robinson P.
      • Keenan A.M.
      • Conaghan P.G.
      Clinical effectiveness and dose–response of image-guided intra-articular corticosteroid injection for hip osteoarthritis.
      . In one RCT an IA injection combining bupivacaine and triamcinolone did not give better pain relief than IA injections of saline after 1 month (RR=1.18, 95% CI 0.68, 2.15) or after 3 months (RR=0.61, CI 0.23, 1.60); and the combination containing IA steroid was not better than injections of local anaesthetic alone in patients with OA awaiting hip joint replacement
      • Flanagan J.
      • Casale F.F.
      • Thomas D.L.
      • Desai K.B.
      Intra-articular injection for pain relief in patients awaiting hip replacement.
      . A second RCT in 80 patients with severe symptomatic OA hip compared the effects of fluoroscopically controlled IA injection of 80 mg triamcinolone hexacetonide or 1% mepivacaine and demonstrated significant reduction in pain and improved mobility after 3 weeks and 3 months in the steroid treated patients but not in those treated with IA injections of local anaesthetic
      • Kullenberg B.
      • Runesson R.
      • Tuvhag R.
      • Olsson C.
      • Resch S.
      Intra-articular corticosteroid injection: pain relief in osteoarthritis of the hip?.
      .
      No serious adverse events were reported as a consequence of IA steroid injections in 1973 patients in 28 controlled trials in patients with OA knee
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Intraarticular corticosteroid for treatment of osteoarthritis of the knee.
      . Potential side effects include post-injection flares of pain, crystal synovitis, haemarthrosis, joint sepsis and steroid articular cartilage atrophy, as well as systemic corticosteroid effects such as fluid retention or aggravation of hypertension or diabetes mellitus. Emphasis has been placed on the importance of accurate placement of IA injections to maximise benefit and reduce the risk of adverse effects such as fat necrosis and para-articular tissue atrophy
      • Jones A.
      • Regan M.
      • Ledingham J.
      • Pattrick M.
      • Manhire A.
      • Doherty M.
      Importance of placement of intra-articular steroid injections.
      . There are limited data at present to indicate how frequently it is safe to administer IA steroid injections to patients with OA hip or knee. Most experts recommend caution regarding too-frequent use; repeat injections more than four times annually are generally not recommended.
      • 17.
        Injections of IA hyaluronate may be useful in patients with knee or hip OA. They are characterised by delayed onset, but prolonged duration, of symptomatic benefit when compared to IA injections of corticosteroids.
        SOR: 64% (95% CI 43–85)
      Hyaluronic acid is a large molecular weight glycosaminoglycan which is a constituent of synovial fluid in normal and osteoarthritic joints. IA injection of hyaluronan (HA), with relatively high and low molecular weight averages, is widely used, and recommended in 8/9 existing guidelines as a useful therapeutic modality for treating patients with OA knee as a viscosupplement or pharmaceutical
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      , despite considerable ongoing controversy with regard to its efficacy, cost-effectiveness and benefit to risk ratio. The evidence available to the OARSI treatment guidelines development group from the critical appraisal of existing guidelines and the systematic review of the research evidence from 2002 to January 2006 was derived from two systematic reviews published in 2003
      • Lo G.H.
      • LaValley M.
      • McAlindon T.
      • Felson D.T.
      Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis.
      and 2005
      • Arrich J.
      • Piribauer F.
      • Mad P.
      • Schmid D.
      • Klaushofer K.
      • Mullner M.
      Intra-articular hyaluronic acid for the treatment of osteoarthritis of the knee: systematic review and meta-analysis.
      (LoE Ia). The pooled ES for reduction in pain at 2–3 months following at least three IA injections at weekly intervals in 22 placebo-controlled RCTs was 0.32 (95% CI 0.17, 0.47). There was, however, significant heterogeneity between studies with inconclusive data to suggest that the higher molecular weight HA preparations may be more effective
      • Lo G.H.
      • LaValley M.
      • McAlindon T.
      • Felson D.T.
      Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis.
      . An asymmetric funnel plot and a positive Egger test also suggested the possibility of publication bias; and the identification of two unpublished trials with a pooled ES of 0.07 (95% CI −0.15, 0.28) further suggested that the overall ES might have been overestimated
      • Lo G.H.
      • LaValley M.
      • McAlindon T.
      • Felson D.T.
      Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis.
      . The 2005 MA found no evidence of improvement in function in pooled results from nine placebo-controlled RCTs which included joint function as an outcome (ES=0.00, 95% CI −0.23, 0.23) and no effects on pain during movement compared with saline injections that were judged to be clinically meaningful at any time point after treatment
      • Arrich J.
      • Piribauer F.
      • Mad P.
      • Schmid D.
      • Klaushofer K.
      • Mullner M.
      Intra-articular hyaluronic acid for the treatment of osteoarthritis of the knee: systematic review and meta-analysis.
      . Two further systematic reviews of IA injections of HA in patients with OA knee were published in 2006
      • Medina J.M.
      • Thomas A.
      • Denegar C.R.
      Knee osteoarthritis: should your patient opt for hyaluronic acid injection?.
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Viscosupplementation for the treatment of osteoarthritis of the knee.
      . One MA of seven placebo-controlled RCTs which used the WOMAC or Lequesne indexes as outcome measurements found small but significant improvements in the Lequesne index, but not in the WOMAC scales for self-reported pain or disability up to 6 months after treatment
      • Medina J.M.
      • Thomas A.
      • Denegar C.R.
      Knee osteoarthritis: should your patient opt for hyaluronic acid injection?.
      . A more comprehensive industry-sponsored Cochrane review which included an MA of 40 placebo-controlled trials with five different commercially available HA products found statistically significant improvements in pain on weight bearing when results were pooled (WMDs of −8, −13, −9 and −3 at 1–4, 5–13, 14–26 and 45–52 weeks, respectively), but improvements from baseline to the maximum at 5–13 weeks varied from 28% to 54% for pain and from 9% to 32% for function with different products
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Viscosupplementation for the treatment of osteoarthritis of the knee.
      . In 10 trials comparing IA HA injections with IA corticosteroids there were no significant differences 4 weeks after injection but IA HA was shown to be more effective 5–13 weeks post-injection for one or more of a number of outcome variables (WOMAC OA index, Lequesne index, pain, range of flexion, and number of responders)
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Intraarticular corticosteroid for treatment of osteoarthritis of the knee.
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Viscosupplementation for the treatment of osteoarthritis of the knee.
      . No major safety issues were detected
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Viscosupplementation for the treatment of osteoarthritis of the knee.
      but in placebo-controlled trials minor adverse events such as transient pain at the injection site occurred slightly more frequently in patients treated with IA HA (RR=1.08, 95% CI 1.01, 1.15)
      • Arrich J.
      • Piribauer F.
      • Mad P.
      • Schmid D.
      • Klaushofer K.
      • Mullner M.
      Intra-articular hyaluronic acid for the treatment of osteoarthritis of the knee: systematic review and meta-analysis.
      . A recent study
      • Campbell J.
      • Bellamy N.
      • Gee T.
      Differences between systematic reviews/meta-analyses of hyaluronic acid/hyaluronan//hylan in osteoarthritis of the knee.
      used the decision algorithm proposed by Jadad et al.
      • Jadad A.R.
      • Cook D.J.
      • Browman G.P.
      A guide to interpreting discordant systematic reviews.
      and the GRADE (Grades of Recommendation Assessment, Development and Evaluation)
      • The GRADE Working Group
      Grading quality of evidence and strength of recommendations.
      system to explore the reasons for discordant conclusions in six published systematic reviews of IA HA for the treatment of OA knee
      • Lo G.H.
      • LaValley M.
      • McAlindon T.
      • Felson D.T.
      Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis.
      • Arrich J.
      • Piribauer F.
      • Mad P.
      • Schmid D.
      • Klaushofer K.
      • Mullner M.
      Intra-articular hyaluronic acid for the treatment of osteoarthritis of the knee: systematic review and meta-analysis.
      • Medina J.M.
      • Thomas A.
      • Denegar C.R.
      Knee osteoarthritis: should your patient opt for hyaluronic acid injection?.
      • Bellamy N.
      • Campbell J.
      • Robinson V.
      • Gee T.
      • Bourne R.
      • Wells G.
      Viscosupplementation for the treatment of osteoarthritis of the knee.
      • Wang C.T.
      • Lin J.
      • Chang C.J.
      • Lin Y.T.
      • Hou S.M.
      Therapeutic effects of hyaluronic acid on osteoarthritis of the knee. A meta-analysis of randomized controlled trials.
      • Modowai A.
      • Ferrer M.
      • Choi H.K.
      • Castle J.A.
      Hyaluronic acid injections relieve knee pain.
      . The reasons for inconsistency identified included inclusion of different controlled trials as a result of different search strategies and selection criteria, differences in the outcome measures and time points selected for extraction; and different statistical methods for data synthesis, which resulted in conflicting estimates of therapeutic effect
      • Campbell J.
      • Bellamy N.
      • Gee T.
      Differences between systematic reviews/meta-analyses of hyaluronic acid/hyaluronan//hylan in osteoarthritis of the knee.
      . There is much less research evidence to support the proposition that IA injections of HA can be a useful treatment in patients with hip OA. Three quasi systematic reviews have examined the results of a number uncontrolled clinical trials and case series
      • Conrozier T.
      • Vignon E.
      Is there evidence to support the inclusion of viscosupplementation in the treatment paradigm for patients with hip osteoarthritis?.
      • Fernandez Lopez J.C.
      • Ruana-Ravina A.
      Efficacy and safety of intraarticular hyaluronic acid in the treatment of hip osteoarthritis: a systematic review.
      • Van den Bekerom M.P.J.
      • Lamme B.
      • Sermon A.
      • Mulier M.
      What is the evidence for viscosupplementation in the treatment of patients with hip osteoarthritis? Systematic review of the literature.
      , a single comparison of injection of a low or high molecular weight HA
      • Tikiz C.
      • Unlu Z.
      • Sener A.
      • Efe M.
      • Tuzun C.
      Comparison of the efficacy of lower and higher molecular weight viscosupplementation in the treatment of hip osteoarthritis.
      , and a single, double blind, three armed RCT in 101 patients with hip OA in which IA injections of a low molecular weight HA preparation were compared with IA saline and IA corticosteroid injections
      • Quistgaard E.
      • Christensen R.
      • Torp-Pedersen S.
      • Bliddal H.
      Intra-articular treatment of hip osteoarthritis: a randomized trial of hyaluronic acid, corticosteroid and isotonic saline.
      . In the randomised comparison of three injections of high and low molecular weight HA given at weekly intervals under fluoroscopic control there were significant improvements of approximately 40% in VAS, WOMAC and Lequesne index scores 1, 3 and 6 months after treatment but no significant differences at any of the time points between the two groups
      • Tikiz C.
      • Unlu Z.
      • Sener A.
      • Efe M.
      • Tuzun C.
      Comparison of the efficacy of lower and higher molecular weight viscosupplementation in the treatment of hip osteoarthritis.
      . However in the placebo-controlled trial in which three injections of HA, corticosteroid or saline were given with ultrasound guidance at 2 weekly intervals, there were no significant differences between the HA treated, corticosteroid treated or saline treated groups in pain on walking, WOMAC or Lequesne indices 14, 28 or 90 days after the course of injections
      • Quistgaard E.
      • Christensen R.
      • Torp-Pedersen S.
      • Bliddal H.
      Intra-articular treatment of hip osteoarthritis: a randomized trial of hyaluronic acid, corticosteroid and isotonic saline.
      . Responses at 14 days applying OARSI response criteria were 53% in patients treated with HA, 56% in the corticosteroid treated group and 33% in the placebo-treated patients. At 28 days 53% responded to HA, 66% to corticosteroids and 44% to placebo
      • Quistgaard E.
      • Christensen R.
      • Torp-Pedersen S.
      • Bliddal H.
      Intra-articular treatment of hip osteoarthritis: a randomized trial of hyaluronic acid, corticosteroid and isotonic saline.
      .
      • 18.
        Treatment with glucosamine and/or chondroitin sulphate may provide symptomatic benefit in patients with knee OA. If no response is apparent within 6 months treatment should be discontinued.
        SOR: 63% (95% CI 44–82)
      The aminosugar glucosamine and the glycosaminoglycan chondroitin sulphate are both naturally occurring constituents of cartilage proteoglycans that are very widely used as ‘nutritional supplements’ by patients with OA
      • Barnes P.M.
      • Powell-Griner E.
      • McFann K.
      • Nahin R.L.
      Complementary and alternative medicine use among adults: United States 2002.
      . A crystalline preparation of glucosamine sulphate is approved as a medicinal product for the treatment of OA in many countries in Europe, Asia and Latin America. Glucosamine sulphate is recommended in 6/10 existing guidelines for the management of hip or knee OA, but chondroitin sulphate in only 2/7 guidelines where these modalities of therapy were considered
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      , and there is continuing controversy as to the efficacy of these agents as symptom modifying drugs.
      Evidence available to the guideline development committee concerning the efficacy and safety of glucosamine was mainly derived from the 2005 update of the Cochrane systematic review and MA
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      and an earlier MA published in 2003
      • Richy F.
      • Bruyere O.
      • Ethgen O.
      • Cucherat M.
      • Henrotin Y.
      • Reginster J.-Y.
      Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis.
      (LoE Ia). In comparisons with placebo, pooled analysis of 20 RCTs involving 2570 patients with knee OA showed a 28% improvement in pain (ES=0.61, 95% CI 0.28, 0.95) and a 21% improvement in function using the Lequesne index (ES=0.51, 95% CI 0.96, 0.05)
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      . However WOMAC pain, stiffness and function were not significantly changed and there was considerable heterogeneity of outcomes in different trials. With such marked heterogeneity, pooling of results may not be appropriate and estimates of overall ESs may be misleading. The possible reason(s) for the variation in outcomes also requires an explanation. In 10 placebo-controlled RCTs in which the Rottapharm preparation of glucosamine sulphate 1500 mg daily was used there were significant improvements in pain (ES=1.31, 95% CI 0.64, 1.99) and function (ES=0.51, 95% CI 0.05, 0.96) while there were no significant improvements in WOMAC pain or function indices in the pooled results of RCTs that used other glucosamine formulations
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      . Analysis of the eight RCTs in which allocation concealment was considered adequate also failed to show drug efficacy for relief of pain or improvement in function using the WOMAC index
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      . A more recent systematic review was undertaken specifically to try and identify the factors that might be responsible for the heterogeneity of outcomes in trials of glucosamine
      • Vlad S.C.
      • La Valley M.P.
      • McAlindon T.E.
      • Felson D.T.
      Glucosamine for pain in osteoarthritis. Why do trial results differ?.
      . In 15 RCTs which met the inclusion criteria the summary of ES for pain relief was 0.35 (95% CI 0.14, 0.56) but there was a considerable variation in outcomes attributable to differences between studies, rather than to chance
      • Higgins J.P.
      • Simon G.T.
      • Deeks J.J.
      • Altman R.D.
      Measuring inconsistency in meta-analyses.
      , with an I2 of 80%
      • Vlad S.C.
      • La Valley M.P.
      • McAlindon T.E.
      • Felson D.T.
      Glucosamine for pain in osteoarthritis. Why do trial results differ?.
      (i.e., 80% of the inconsistency could be attributed to the true differences between studies). An Egger test and funnel plot
      • Egger M.
      • Smith J.D.
      • Schneider M.
      • Minder C.
      Bias in meta-analysis detected by a simple graphical test.
      did not suggest publication bias and there were no clear indications that the heterogeneity was attributable to differences in trial design, trial quality, the number of dropouts or differences in intention to treat analyses, but the differences in adequacy of the allocation concealment detected in the Cochrane review
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      were confirmed. The most striking differences, however, seemed to be related to the glucosamine preparation that was used. The ES for trials which used glucosamine sulphate was 0.44 (95% CI 0.18, 0.70) compared with 0.06 (95% CI −0.08, 0.20) for those that used glucosamine hydrochloride, and the ES for trials utilising the Rottapharm preparation of glucosamine sulphate was 0.55 (95% CI 0.29, 0.82) compared with an ES of 0.11 (95% CI −0.16, 0.38) for trials with other products. The possibility of industry bias as an additional or alternative explanation for the heterogeneity of outcomes between glucosamine trials was also suggested, but not substantiated
      • Vlad S.C.
      • La Valley M.P.
      • McAlindon T.E.
      • Felson D.T.
      Glucosamine for pain in osteoarthritis. Why do trial results differ?.
      . In the first part of this report it was shown that sensitivity analysis following addition of the data from two large multicentre RCTs which were published after the close of the systematic review in January 2006; the NIH sponsored Glucosamine/chondroitin Arthritis Intervention Trial (GAIT)
      • Clegg D.O.
      • Reda D.J.
      • Harris C.L.
      • Klein M.A.
      • O'Dell J.R.
      • Hooper M.M.
      • et al.
      Glucosamine, chondroitin sulphate and the two in combination for painful knee osteoarthritis.
      in which glucosamine hydrochloride was used, and the Glucosamine Unum in Die Efficacy (GUIDE) trial
      • Herrero-Beaumont G.
      • Roman Ivorra J.A.
      • del Carmen Trabado M.C.
      • Blanco F.J.
      • Benito P.
      • Martin-Mola E.
      • et al.
      Glucosamine sulphate in the treatment of knee osteoarthritis symptoms: a randomized, double-blind, placebo-controlled study using acetaminophen as a side comparator.
      in which glucosamine sulphate 1500 mg daily was employed, to the main body of trial outcomes, did not alter the ESs for pain efficacy significantly
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . The NNT for treatment of knee OA with glucosamine sulphate is 5 (95% CI 4, 7)
      • Richy F.
      • Bruyere O.
      • Ethgen O.
      • Cucherat M.
      • Henrotin Y.
      • Reginster J.-Y.
      Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis.
      and treatment is not associated with any serious adverse effects
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      .
      The evidence supporting the recommendation that chondroitin sulphate may provide symptomatic benefit in patients with knee OA is also conflicting. At the time of the Delphi exercise the evidence for efficacy of chondroitin sulphate was supported by two MAs published in 2000
      • Leeb B.F.
      • Schweizer M.
      • Montag K.
      • Smolen J.
      A meta-analysis of chondroitin sulphate in the treatment of osteoarthritis.
      • McAlindon T.E.
      • La Valley M.P.
      • Gulin J.P.
      • Felson D.T.
      Glucosamine and chondroitin for treatment of osteoarthritis: a systematic quality assessment and meta-analysis.
      and a third one in 2003
      • Richy F.
      • Bruyere O.
      • Ethgen O.
      • Cucherat M.
      • Henrotin Y.
      • Reginster J.-Y.
      Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis.
      (LoE Ia). Analysis of eight RCTs involving 755 patients showed a moderate ES for pain reduction (ES=0.52, 95% CI 0.37, 0.67) with an NNT of 5 (4, 7) and no evidence of serious side effects
      • Richy F.
      • Bruyere O.
      • Ethgen O.
      • Cucherat M.
      • Henrotin Y.
      • Reginster J.-Y.
      Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis.
      . However, as shown in the first part of this report
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      , sensitivity analysis following addition of the data from the GAIT study
      • Clegg D.O.
      • Reda D.J.
      • Harris C.L.
      • Klein M.A.
      • O'Dell J.R.
      • Hooper M.M.
      • et al.
      Glucosamine, chondroitin sulphate and the two in combination for painful knee osteoarthritis.
      to the main body of trial outcomes reduced the ES for pain reduction significantly (ES=0.30, 95% CI −0.10, 0.70) and suggested that treatment with chondroitin sulphate was not significantly more effective than placebo. This was also the conclusion of the most recent systematic review and MA
      • Reichenbach S.
      • Sterchi R.
      • Scherer M.
      • Trelle S.
      • Burgi E.
      • Burgi U.
      • et al.
      Meta-analysis: chondroitin for osteoarthritis of the knee or hip.
      . In their analysis of 20 trials involving 3846 patients the ES for pain relief was large (ES=0.75, 95% CI 0.50, 0.99) but they identified very marked heterogeneity of outcomes between trials with an I2 of 92%. Small trials with poor quality features such as uncertain concealment of allocation and a failure to analyse results on an intention to treat basis showed larger effects in favour of chondroitin than did the remaining trials
      • Reichenbach S.
      • Sterchi R.
      • Scherer M.
      • Trelle S.
      • Burgi E.
      • Burgi U.
      • et al.
      Meta-analysis: chondroitin for osteoarthritis of the knee or hip.
      . Similar caveats had been raised in one of the earlier MAs
      • McAlindon T.E.
      • La Valley M.P.
      • Gulin J.P.
      • Felson D.T.
      Glucosamine and chondroitin for treatment of osteoarthritis: a systematic quality assessment and meta-analysis.
      . When Reichenbach et al. restricted the analysis to three recent trials with large sample sizes and an intention to treat analysis
      • Clegg D.O.
      • Reda D.J.
      • Harris C.L.
      • Klein M.A.
      • O'Dell J.R.
      • Hooper M.M.
      • et al.
      Glucosamine, chondroitin sulphate and the two in combination for painful knee osteoarthritis.

      Kahan A. STOPP (study on osteoarthritis progression prevention): a new two year trial with chondroitin 4 & 6 sulphate (CS). www.ibsa-ch.com/eular 2006 amsterdam vignon-2pdf (accessed on 18 September 2006).

      • Mazieres B.
      • Hucher M.M.H.
      • Zam M.M.Z.
      Chondroitin sulphate in the treatment for knee osteoarthritis: a randomized, double-blind, multi center, placebo-controlled trial.
      ; the ES for pain reduction was only 0.03 (95% CI −0.07, 0.13) with an I2 of 0%
      • Reichenbach S.
      • Sterchi R.
      • Scherer M.
      • Trelle S.
      • Burgi E.
      • Burgi U.
      • et al.
      Meta-analysis: chondroitin for osteoarthritis of the knee or hip.
      . However this restricted analysis included one study with an exceptionally high placebo response rate
      • Clegg D.O.
      • Reda D.J.
      • Harris C.L.
      • Klein M.A.
      • O'Dell J.R.
      • Hooper M.M.
      • et al.
      Glucosamine, chondroitin sulphate and the two in combination for painful knee osteoarthritis.
      , one study that was only published as an abstract

      Kahan A. STOPP (study on osteoarthritis progression prevention): a new two year trial with chondroitin 4 & 6 sulphate (CS). www.ibsa-ch.com/eular 2006 amsterdam vignon-2pdf (accessed on 18 September 2006).

      and only 40% of all trial patients. The pooled RR for adverse events in an MA of 12 placebo-controlled trials was 0.99 (95% CI 0.76, 1.21)
      • Reichenbach S.
      • Sterchi R.
      • Scherer M.
      • Trelle S.
      • Burgi E.
      • Burgi U.
      • et al.
      Meta-analysis: chondroitin for osteoarthritis of the knee or hip.
      .
      • 19.
        In patients with symptomatic knee OA glucosamine sulphate and chondroitin sulphate may have structure-modifying effects while diacerein may have structure-modifying effects in patients with symptomatic OA of the hip.
        SOR: 41% (95% CI 20–62)
      Evidence that glucosamine sulphate 1500 mg/day may have structure-modifying effects in patients with knee OA comes from two placebo-controlled RCTs involving 414 patients
      • Reginster J.Y.
      • Deroisy R.
      • Rovati L.C.
      Long-term effects of glucosamine sulphate on osteoarthritis progression: a randomized, placebo-controlled clinical trial.
      • Pavelka K.
      • Gatterova J.
      • Olejarova M.
      • Machacek S.
      • Giocovelli G.
      • Rovati L.C.
      Glucosamine sulphate use and delay of progression of knee osteoarthritis.
      and two systematic reviews and MAs
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      • Richy F.
      • Bruyere O.
      • Ethgen O.
      • Cucherat M.
      • Henrotin Y.
      • Reginster J.-Y.
      Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis.
      (LoE Ia). In one trial there was no radiographic loss of joint space width (JSW) in the medial compartment of the tibio-femoral joint after 3 years (mean −0.06 mm, 95% CI −0.22, 0.09) in the treated patients compared with progressive loss in the placebo group (mean −0.31 mm, 95% CI −0.48, −0.13)
      • Reginster J.Y.
      • Deroisy R.
      • Rovati L.C.
      Long-term effects of glucosamine sulphate on osteoarthritis progression: a randomized, placebo-controlled clinical trial.
      . The pooled results of both trials showed an ES=0.24 (95% CI 0.04, 0.43)
      • Towheed T.E.
      • Maxwell L.
      • Anastassiades T.P.
      • Shea B.
      • Houpt J.
      • Robinson V.
      • et al.
      Glucosamine therapy for treating osteoarthritis.
      .
      The proposition that chondroitin sulphate (800 mg/day) may also have structure-modifying effects is supported by an MA of five placebo-controlled RCTs. The difference in changes over 2 years between chondroitin and placebo demonstrated a small effect in favour of chondroitin: 0.16 mm on minimum JSW (95% CI 0.08, 0.24) and 0.23 mm on mean JSW (95% CI 0.09, 0.37)
      • Reichenbach S.
      • Sterchi R.
      • Scherer M.
      • Trelle S.
      • Burgi E.
      • Burgi U.
      • et al.
      Meta-analysis: chondroitin for osteoarthritis of the knee or hip.
      (LoE Ia).
      The evidence to support the proposition that diacerein may have structure-modifying effects in patients with hip OA comes from a single 3-year placebo-controlled RCT in 507 patients with primary hip OA
      • Dougados M.
      • Nguyen M.
      • Berdah L.
      • Mazieres B.
      • Vignon E.
      • Lequesne M.
      Evaluation of the structure-modifying effects of diacerein in hip osteoarthritis: ECHODIAH, a three year, placebo-controlled trial. Evaluation of the chondromodulating effect of diacerein in OA of the hip.
      and a systematic review and MA
      • Fidelix T.S.
      • Soares B.G.
      • Treviani V.F.
      Diacerein for osteoarthritis.
      (LoE Ia). In patients who completed 3 years of therapy with diacerein 50 mg twice daily the rate of joint space narrowing was [mean±standard deviation (SD)] 0.18±0.25 mm/year vs 0.23±0.23 mm/year with placebo (P=0.042)
      • Dougados M.
      • Nguyen M.
      • Berdah L.
      • Mazieres B.
      • Vignon E.
      • Lequesne M.
      Evaluation of the structure-modifying effects of diacerein in hip osteoarthritis: ECHODIAH, a three year, placebo-controlled trial. Evaluation of the chondromodulating effect of diacerein in OA of the hip.
      . Similar structure-modifying effects were not evident in a 1-year placebo-controlled RCT in patients with knee OA
      • Pham T.
      • Le Henanff A.
      • Ravaud P.
      • Dieppe P.
      • Paolozzi L.
      • Dougados M.
      Evaluation of the symptomatic and structural efficacy of a new hyaluronic acid compound, NRD 101, in comparison with diacerein and placebo in a 1 year randomised controlled study in symptomatic knee osteoarthritis.
      .
      • 20.
        The use of weak opioids and narcotic analgesics can be considered for the treatment of refractory pain in patients with hip or knee OA, where other pharmacological agents have been ineffective, or are contraindicated. Stronger opioids should only be used for the management of severe pain in exceptional circumstances. Non-pharmacological therapies should be continued in such patients and surgical treatments should be considered.
        SOR: 82% (95% CI 74–90)
      The use of opioid analgesics is recommended in 9/9 existing treatment guidelines for the management of hip or knee OA
      • Zhang W.
      • Moskowitz R.W.
      • Nuki G.
      • Abramson S.
      • Altman R.D.
      • Arden N.
      • et al.
      OARSI recommendations for the management of hip and knee osteoarthritis, Part I: Critical appraisal of existing treatment guidelines and systematic review of current research evidence.
      . A number of systematic reviews and MAs of the use of opioids for chronic non-cancer pain
      • Kalso E.
      • Edwards J.E.
      • Moore R.A.
      • McQuay H.
      Opioids in chronic non-cancer pain: systematic review of efficacy and safety.
      • Furlan A.D.
      • Sandoval J.A.
      • Mailis-Gagnon A.
      • Tunks E.
      Opioids for chronic non-cancer pain: a meta-analysis of effectiveness and side effects.
      , musculoskeletal pain
      • Abasolo L.
      • Carmona L.
      Systematic review: are major opioids effective in the treatment of musculoskeletal pain?.
      and more recently OA
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      have provided evidence of efficacy and acceptable safety in short-term trials (LoE Ia). Analysis of 18 placebo-controlled RCTs including 3244 patients with OA showed a moderate ES for reduction in pain intensity (ES=0.78, 95% CI 0.59, 0.98) but there was substantial heterogeneity between studies which was not obviously related to the opioid preparation that was used or the methodological quality of the RCTs
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      . The median duration of trials was 12 weeks (range 1.4–72 weeks)
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      . Analysis of five placebo-controlled RCTs which included 1429 OA patients receiving opioids showed a small effect on improvement in physical function (ES=0.31, 95% CI 0.24, 0.39)
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      . Benefits associated with the use of opioids were, however, limited by frequent side effects; nausea (30%), constipation (23%), dizziness (20%), somnolence (18%) and vomiting (13%)
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      . Overall 25% of patients treated with opioids withdrew from studies compared with 7% of placebo-treated patients with a number needed to harm (NNH) of 5. The withdrawal rate for strong opioids (oxymorphone, oxycodone, oxytrex, fentanyl, morphine sulphate) was 31% (NNH 4) compared with a withdrawal rate of 19% and an NNH of 9 for the weaker opioids (tramadol, tramadol/paracetamol, codeine and propoxyphene)
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      . This MA
      • Avouac J.
      • Gossec L.
      • Dougados M.
      Efficacy and safety of opioids for osteoarthritis: a meta-analysis of randomized controlled trials.
      did not allow any conclusions concerning comparisons of the efficacy or safety of opioids and other available analgesics such as paracetamol or NSAIDs because of the very limited number of head to head trials. However, another MA of opioids for chronic non-cancer pain, including OA, demonstrated that only strong opioids were significantly more effective than paracetamol or NSAIDs (ES=0.34, 95% CI 0.01, 0.67)
      • Furlan A.D.
      • Sandoval J.A.
      • Mailis-Gagnon A.
      • Tunks E.
      Opioids for chronic non-cancer pain: a meta-analysis of effectiveness and side effects.
      . A systematic review conducted a decade earlier had, however, confirmed that paracetamol–codeine combinations did provide a small (approximately 5%) but statistically significant analgesic benefit when compared with paracetamol alone, but adverse effects were more frequent (RR=2.5, 95% CI 1.5, 4.2)
      • De Craen A.J.M.
      • Di Giulio G.
      • Lampe-Schoenmaeckers A.J.E.M.
      • Kessels A.G.H.
      • Kleijnen J.
      Analgesic efficacy and safety of paracetamol–codeine combinations versus paracetamol alone: a systematic review.
      . All the systematic reviews highlight the fact that there have been no long-term trials of the use of opiates for treating patients with OA
      • Kalso E.
      • Edwards J.E.
      • Moore R.A.
      • McQuay H.
      Opioids in chronic non-cancer pain: systematic review of efficacy and safety.
      • Furlan A.D.
      • Sandoval J.A.
      • Mailis-Gagnon A.
      • Tunks E.
      Opioids for chronic non-cancer pain: a meta-analysis of effectiveness and side effects.
      • Abasolo L.
      • Carmona L.
      Systematic review: are major opioids effective in the treatment of musculoskeletal pain?.