Effectiveness of nonpharmacological and nonsurgical interventions for hip osteoarthritis: an umbrella review of high-quality systematic reviews.

An increasing number of systematic reviews are available regarding nonpharmacological and nonsurgical interventions for hip osteoarthritis (OA). The objectives of this article are to identify high-quality systematic reviews on the effect of nonpharmacological and nonsurgical interventions for hip OA and to summarize available high-quality evidence for these treatment approaches. The authors identified and screened 204 reviews. Two independent reviewers using a previously pilot-tested quality assessment form assessed the full text of 58 reviews. Six reviews were of sufficient high quality and could be included for further analyses. There was moderate-quality evidence that acupuncture and diacerein have no effect on pain and function. There was low-quality evidence that strengthening exercises and avocado/soybean unsaponifiables reduce pain and that diacerein decreases radiographic OA progression. There was insufficient high-quality evidence regarding nonpharmacological and nonsurgical interventions for hip OA, and further primary studies and reviews are needed.

O steoarthritis (OA) is a chronic joint condition, characterized by pain, disability, and impairment. The prevalence of hip OA in Western populations over 35 years of age ranges from 3% to 11%. 1 The hip is considered one of the most common weight-bearing joints affected by OA. [2][3][4] Main treatment goals are improved function, symptomatic relief, slowing disease progression, and improving quality of life. 5 Treatment for OA may vary depending on various factors, 1 and guidelines on the management of OA recommend both pharmacological and nonpharmacological approaches. 1 There is limited availability of disease-modifying drugs, and many patients use complementary and alternative medicines and therapies. An increasing number of systematic reviews are available regarding nonpharmacological and nonsurgical interventions, and in this umbrella review we summarize and grade the quality of the available evidence for these treatment approaches.
Decisions on the provision of health care are increasingly based on the available evidence. Patients, health care professionals, and researchers need information about the effectiveness of interventions in order to improve self-management strategies, to improve clinical practice, and to set priorities for research, respectively. Decisions on the reimbursement of health care are increasingly evidence-based. Thus, purchasing organizations and policymakers in health care are in need of reliable information on the effectiveness of interventions.
Summarizing systematic reviews can facilitate decision making about appropriate health care, promote evidence-based treatment, and identify areas for future research in health care. Conclusions based on a systematic review of randomized controlled trials are considered to provide the highest level of evidence about the effectiveness of an intervention.
Based on a review of literature before 2001, Chard and Dieppe concluded, "Nonpharmaceutical therapies for OA have not been researched enough for us to understand their potential benefit." 6(p256) There is, to our knowledge, no updated overview available on the effectiveness of nonpharmacological and nonsurgical interventions for hip OA. The aim of this overview is to summarize the available evidence from systematic reviews on the effectiveness of nonpharmacological and nonsurgical interventions for patients with hip OA.

Method Criteria for Including Reviews
We included systematic reviews with the primary aim of investigating the effects of nonpharmacological and nonsurgical interventions for hip OA published in the English, Dutch, or Scandinavian language. More specifically, the following inclusion criteria were used: • People with OA: Diagnosis according to the American College of Rheumatology criteria 7 or other acceptable criteria. 8 Reviews including people with various rheumatic diagnoses were accepted only if results for OA could be extracted separately.
• Interventions: All types of nonpharmacological and nonsurgical interventions. Excluded were interventions such as gene therapy, all types of invasive interventions (ie, injections or arthroscopy), therapeutic apheresis or interventions related to pharmacological or surgical interventions (eg, therapeutic exercises after total joint replacement).
• Outcomes: For the purposes of this overview, the primary outcome measures were function, pain, and stiffness. The concept of "function" is based on the International Clas-sification of Functioning, Disability and Health 9 (ICF) definition, where "function" is an umbrella term for body function, body structure, activities, and participation.

Search Strategy
We searched the Cochrane Library (Cochrane Database of Systematic Reviews and DARE), MEDLINE, EMBASE, PEDro, PsychINFO, and CINAHL from 2000 to January 2007 for "hip osteoarthritis/-arthrosis or OA." A broad computerized search strategy was developed (Appendix 1). Reference lists from retrieved reviews were examined.
Retrieved hits were assessed by 2 of the authors (EAH, RHM), who screened the titles and abstracts to identify relevant studies. If doubt occurred, one of the other authors (KBH) was consulted. Potential relevant full-text articles were read by 2 authors (EAH, RHM).

Assessment of Methodological Quality
Two authors (EAH, RHM) independently assessed the methodological quality of the reviews. Disagreement was resolved by discussion. Nine criteria were rated as "met," "unclear/ partly met," or "not met" according to a criteria list modified from the Effective Practice and Organisation of Care (EPOC) group within the Cochrane Collaboration (Appendix 2). 10 Based on a summary of these 9 criteria, an overall scientific quality was applied to each review, as follows: "minor limitations" (at least 7 of the criteria were met), "moderate limitations" (4 -6 of the criteria were met), and "major limitations" (fewer than 4 of the criteria were met). Reviews with major limitations were excluded. The quality assessments of primary studies included in the original reviews are reported in Table 1.
Principles from Grading of Recommendations Assessment, Develop- Effectiveness of Nonpharmacological and Nonsurgical Interventions for Hip OA ment, and Evaluation (GRADE) were used for an overall assessment of the quality of evidence for each intervention. 10 -12 The GRADE concept is based on an assessment of the following criteria: quality of primary studies, design of primary studies, consistency, and directness. An over-all assessment of the quality of evidence was based on a summary of these 4 criteria, as presented in Table 2.

Data Extraction and Synthesis
Data on effectiveness were extracted from the identified high-quality re-views by 2 of the authors (EAH, RHM); if doubt occurred, one of the other authors (KBH) was consulted.
The following criteria were applied when data on effects were extracted: • Adequate quantitative pooling of data in reviews was regarded as

Effectiveness of Nonpharmacological and Nonsurgical Interventions for Hip OA
more valid than a qualitative data synthesis approach. • If no direct comparisons between treatments were undertaken or no quantitative pooling of data was done, the results are reported as "no quantitative pooling," and the authors' statements were reported.
• When we found that the results were reported inconsistently in different sections of the review, the effects were extracted from the main result section.

Study Selection
The literature search identified 204 reviews on hip OA. One hundred sixty-four articles were clearly not relevant based on information from the abstract. For 58 reviews, the full text was retrieved (Tab. 3) and assessed, and 52 reviews were excluded for various reasons (Figure). In the end, we included 6 high-quality systematic reviews, which formed the basis of this umbrella review (Tab. 1). Generally, the methodological quality of the primary studies was low to moderate, often presenting conflicting results (Tab. 4).

Quality of Findings
Acupuncture. One high-quality systematic review 13 assessed the effect of acupuncture on peripheral joint OA. The conclusions were based on 3 primary studies. On the basis of the meta-analysis, there were no statistically significant results, and thus there was no evidence that acupuncture is beneficial for reducing OA pain. Mean pain reduction was 14.43 (on a 0 -100 visual analog scale [VAS]) for the intervention group and 15.31 for the sham treatment group (mean difference of Ϫ0.03, 95% confidence interval [CI]ϭϪ0.52-0.45).
Avocado/soybean unsaponifiables (ASU). Avocado/soybean unsaponifiables may reduce pain in people with chronic hip OA. The authors' conclusion is that the evidence for the beneficial effects of ASU on OA is convincing. 14 Evidence extracted from the review was based on one primary study that compared ASU with a placebo on VAS pain scores. The mean difference was Ϫ13.80 (95% CIϭϪ25.2 to Ϫ2.38, Pϭ.02) (on a 0 -100 scale) in favor of ASU. In our opinion, current available data on ASU suggest that it may provide possible beneficial effects on OA of the hip, but there is still insufficient evidence to draw firm conclusions. These data suggest that there is low-quality evidence that ASU reduces pain (based on VAS scores) in hip OA.
Diacerein. Diacerein is a symptomatic, slow-acting herbal therapy for OA. It is a registered medication in the United States, but it is considered a herbal therapy in most other countries; therefore, it was included in this umbrella review. We included 2 reviews on the effect of diacerein. 15,16 The conclusions from these reviews on the effect of diacerein on hip OA were based on 7 primary studies; however, evidence on radiographic OA progression was based on the results of one primary study. There was a statistically significant slowing of radiographic OA progression on diacerein versus a placebo (Ͼ0.50 mm during 3 years, relative riskϭ0.84 [95% CIϭ0.71-0.99], number needed to treatϭ11 [95% CIϭ6 -167]). However, the adverse effect of diarrhea (42%) was quite common 15 and, in our opinion, should not be ignored in clinical practice and further research. There is low-quality evidence that treatment with diacerein reduces radiographic OA progression and moderate-quality evidence that it has no effect on pain, impairment, or incidence of total hip replacement. 15 Exercises. Two high-quality reviews reported on the effects of exercise on hip OA. 17,18 Fransen et al 17 concluded that no optimal exercise type or dosage could be extrapolated from the review due to little available scientific evidence. Roddy et al 18 concluded that there is some evidence that strengthening exercise may be beneficial in reducing pain in people with hip OA, but that there is not enough evidence to make conclusions on the effect on disability. There also is not enough evidence to make conclusions about the effect of aerobic exercise on pain, disability, or health status. 18 Other interventions. It was not possible to extract data on hip OA Table 2.

Level Criteria
High-quality evidence One or more updated, high-quality systematic review that are based on at least 2 high-quality primary studies with consistent results Moderate-quality evidence One or more updated systematic reviews of high or moderate quality • Based on at least 1 high-quality primary study • Based on at least 2 primary studies of moderate quality with consistent results

Low-quality evidence
One or more systematic reviews of variable quality • Based on primary studies of moderate quality • Based on inconsistent results in the reviews • Based on inconsistent results in primary studies Very low-quality evidence No high-quality systematic review was identified on this topic  OA  Table 3. for chondroitin, glucosamine, or herbal therapies such as Reumalex,* capsaicin, and tipi tea. No relevant high-quality reviews were located on weight loss, thermotherapy, patient education, lifestyle changes, electrotherapy, manual therapy, or joint traction or distraction.

Discussion
This overview (umbrella review) of systematic reviews examining the effectiveness of nonpharmacological and nonsurgical interventions for hip OA is based on an extensive literature search, combined with assessment of study quality and synthesis of findings. We identified 204 poten-tially relevant manuscripts, but in the end were able to include only 6 highquality reviews. We found that there is moderate-quality evidence that acupuncture and diacerein have no effect on pain and function. There is lowquality evidence that strengthening exercises and ASU reduce pain and that diacerein decreases radiographic OA progression. Several primary studies might have been published after the reviews included in this overview, and thus their results were not captured. Further updating of reviews and more primary research might confirm our findings and upgrade the evidence. For other interventions and outcomes, the quality of evidence was assessed as low or very low, and new primary studies are needed. For sev-eral interventions for hip OA (ie, aquatic exercise; electrotherapy; glucosamine; herbal therapies such as Reumalex, capsaicin, and tipi tea; joint traction and distraction; lifestyle changes; manual therapy; patient education; thermotherapy; and weight loss), no systematic review was identified. Our conclusions were made on basis of statistically significant changes and not clinically relevant differences. Clinical evidence-based advice perhaps instead should be founded on clinically relevant change (eg, pain reduction).
To our knowledge, no such overview has been published on hip OA until now. Umbrella reviews present a synthesis of the highest-* Gerard House Ltd, 375 Capability Green, Luton, United Kingdom. As the number of published systematic reviews increases, a common finding is that more than one systematic review addresses the same interventions, and conflicts among reviews are emerging. 19 Such discordance might cause difficulties 146 reviews excluded based on abstract 58 full-text reviews retrieved 52 reviews excluded based on: low quality, same review published in different journals, reviews with updates, not possible to extract data on hip osteoarthritis 6 reviews included in umbrella review 204 potentially relevant reviews identified and screened for retrieval

Effectiveness of Nonpharmacological and Nonsurgical Interventions for Hip OA
for decision makers (including clinicians, policymakers, researchers, and patients) who rely on reviews to help them make choices among different health care interventions.
Grading quality on nonpharmacological treatment approaches for OA is challenging. Nonpharmacological evidence is systematically graded significantly lower methodologically than pharmacological evidence. 20 This suggests that it is even more difficult to include nonpharmacological studies and reviews when applying high standards for quality assessments. The type of methodological quality assessment applied determines which studies are included in the review. The number of good primary studies available at the time of the review influences the methodological quality rating, and high-quality primary studies not included in the reviews are not a part of our results.
Depending on the total quality score, we included or excluded reviews for this umbrella review. Total quality scores are presented as a result of summing all 9 quality score items. 12 The cutoff point of 4 out of a total of 9 satisfactory items 9 might be considered strict and is debatable. High-quality primary studies are warranted in order to draw substantive conclusions regarding the effectiveness of interventions. The studies should be randomized, double-blind (or at least assessor blinded), and placebo controlled. The duration of the intervention should be of adequate length, and examinations should be frequent enough to detect a difference in outcome measures. The follow-up period should be of sufficient length to assess long-term effects. Outcome measures also should be standardized, feasible, valid, reliable, and sensitive to change.
The major finding of this umbrella review is that there is insufficient high-quality evidence regarding nonpharmacological and nonsurgical interventions for hip OA. For knee OA, this seems to be a completely different issue, as there is substantially more available evidence for different treatment approaches for this joint. Many reviews include both knee and hip OA and merge the results. It is beyond the scope of this article to address interventions that have some evidence of effectiveness for knee OA but that have not been tested to date for hip OA. We found it difficult to extract data for hip OA only in most reviews. One may not directly apply findings and evidence from studies of knee OA and extrapolate them to hip OA, as the effectiveness of different therapies may be different for these joints. Thus, our lack of findings for hip OA warrants further primary studies and reviews regarding nonpharmacological and nonsurgical interventions in this area.

Conclusion
There is insufficient high-quality evidence regarding nonpharmacological and nonsurgical interventions for hip OA, and further primary studies and reviews are needed.