Consistency of advice for the conservative management of knee osteoarthritis across international clinical practice guidelines

Abstract

Aims

Knee osteoarthritis (OA) is a leading cause of pain and disability, impacting quality of life. There are various conservative (non-surgical, non-pharmacological) management approaches suggested to support those with knee OA and multiple international clinical practice guidelines to inform patient care. This study contrasts the consistency of these management recommendations across the currently available guidelines.

Methods

Systematic search of CINAHL, MEDLINE, Science Direct, and the Guideline International Network website applying PRISMA guidance. Clinical practice guidelines from representative bodies relating to the conservative management of knee OA were included. Quality assessment was completed using the International Centre for Allied Health Evidence (iCAHE) tool.

Results

The search returned 750 records, from which 13 current clinical practice guidelines from national and international medical, surgical, and professional associations were identified. Quality of guideline construction was high. Recommendations for core knee OA management were broadly consistent; entailing exercise, self-management advice and education, and weight management; however, detail as to the proposed delivery of these interventions was lacking. Notable variation was evident between guidelines recommendation for all other interventions, such as manual therapy, bracing/orthotics and acupuncture, related to how the guideline groups interpreted generally low levels of evidence.

Conclusion

There is broad consistency across clinical practice guidelines regarding core conservative management of knee OA which should direct and support patient care. Differences in evidence interpretation as to effect of alternative treatments result in different recommendations which may be confusing for patients, physiotherapists and medical personnel referring to these services.

Cite this article: Bone Jt Open 2025;6(11):1358–1370.

Introduction

Knee osteoarthritis (OA) is one of the world’s leading causes of joint pain and disability,¹ with meta-analysis highlighting a pooled global prevalence of knee OA of 22⋅9% (95% CI 19⋅8% to 26⋅1%) in individuals aged 40 years and over.² There is no cure for knee OA; however, some interventions can relieve discomfort and enhance quality of life, and support people living with this long-term condition.

Knee arthroplasty is a very successful intervention and, with the exception of recent COVID-19-related pressures on services, ever-increasing numbers of these procedures are performed year on year.³ However, only a fraction of those diagnosed with knee OA will ever undergo surgery. UK national guidelines from the National Institute for Health and Care Excellence (NICE) for the consideration of knee arthroplasty state that symptoms should be substantial and that non-surgical management should have been shown to be ineffective.⁴ Non-surgical (or conservative) management of knee OA with physiotherapy and lifestyle modification has been recommended alongside pharmacological management as first line care for people with knee OA for decades.^5-7 The detail though as to what conservative interventions are appropriate for use is more contested.

Various authoritative groups have reviewed the evidence and published clinical practice guidelines for managing knee OA.^4,8,9 Guideline implementation though is a complex process, and healthcare practitioners and patients may not follow the published evidence particularly well.¹⁰ A recent meta-ethnography found that people with OA who have positive healthcare interactions are more likely to actively manage their condition, although many patients also report negative interactions which lead to disengagement with healthcare professionals as a result.¹¹ The same review also found that general practitioners felt underqualified to offer knee OA advice, that they lacked the time to support conservative management, and that patients did not follow the advice that was offered anyway.¹¹

Further meta-analysis highlights under-utilization and considerable variation in content and delivery of nonoperative interventions for knee OA.^12-14 It has been suggested that as few as 40% of patients with knee OA undergo thorough nonoperative management.¹⁵ A lack of management in primary care results in poor symptom management for those with early joint disease and may lead to ever more surgical referrals,¹⁶ burdening an already stressed system in secondary care.

Inconsistency in provision and patient acceptability of differing management strategies and treatments¹⁷ complicates the understanding as to the constituent parts of ‘thorough nonoperative care’. It is unclear to what extent the mainstream international clinical management guidelines for knee OA are consistent or contradictory, which may have a meaningful impact on patient desires for treatment methods and around the clinical behaviour in delivering these. As such, the aim of this work was to evaluate the consistency of recommendations in international clinical practice guidelines for the conservative (non-surgical and non-pharmacological) management of knee OA.

Methods

Search protocol and study selection

A systematic literature search of current clinical practice guidelines for the conservative (non-pharmacological) management of knee OA was undertaken in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA-ScR) guidelines,¹⁸ and in line with the predetermined study protocol, which is available via the Open Science Framework [osf.io/k7sg9]. Three online databases – CINAHL, PubMed (MEDLINE), and Science Direct – were searched from inception until the 18 August 2023. The search terms and strategies included MESH terms and Boolean operators. A summary of search strategy and search terms can be found in Supplementary Table i. Further clinical practice guidelines were identified by searching the Guideline International Network (GIN) library and citation tracking from the reference lists of included papers.

Study selection

English language versions of the most recent clinical practice guidelines for conservative (non-pharmacological and non-surgical) treatment of knee OA published by official representative bodies (national or international professional groups) were considered. Titles were screened by a single author (SM), with abstracts and full text articles reviewed for inclusion by two authors independently (SM, DFH).

Quality assessment

Quality was assessed using the International Centre for Allied Health Evidence (iCAHE) quality guideline checklist,¹⁹ a tool designed and validated for the comparison of practice guidelines. The iCAHE includes 14 questions across six domains, with individual questions scored 1 or 0 resulting in a maximum overall score of 14, with higher scores reflecting better quality. Quality assessment was carried out independently by two researchers (SM, VG), and consensus agreement reached with DH as arbiter.

Data extraction

Data was extracted from the included guidelines using a bespoke Excel data extraction tool (Microsoft, USA). Data was extracted by two authors (SM, VG), and consensus agreement reached with DFH as arbiter. Information extracted included the name of the official body and location represented by the group; composition of the group that formed the guidance; the methodology used for the development of the recommendations; treatment recommendations. We grouped the recommendations offered into three categories for summary reporting; strong evidence for, uncertain; strong evidence against. Where there is debate in the included guideline this is noted in the results section. For clarity of reporting, we have also applied consistent terminology as to the interventions in this report that reflect the variable descriptions used in the included douments (e.g. we have used ‘thermal agents’ to cover the terms thermal therapy, heat/cold therapy, thermography, etc).

Statistical analysis

Due to the nature of the data sought, analysis in this study was by simple counts and percentage agreement of treatment recommendations.

Results

Guideline selection

The search returned a total of 750 records. This included 134 records from PubMed, 424 from ScienceDirect, and 192 from CINAHL. Following removal duplicates, 736 unique records were available for screening. Overall, 16 records were retrieved for full text reviews, where five clinical practice guidelines were identified that met the eligibility criteria and were included. Two further guidelines were identified via the GIN library and an additional six guidelines through citation tracking. In total, 13 clinical practice guidelines were identified. The PRISMA diagram (Figure 1) summarizes the search strategy.

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of summarized search protocol.

Guideline development and quality assessment

The 13 guidelines retrieved were the most recent recommendations by a professional governing association or society (Table I).^4,8,9,20-29 The year published ranged from 2013 to 2022. Development panels represented a variety of clinical specialities (including rheumatologists, physiotherapists, orthopaedic surgeons, manual therapists, occupational therapists, general practitioners, nurses, dietitians, pharmacists and exercise specialists, alongside wider generalists), with 5/13 including patient representatives.^{9,20,22,23,25} The methods used to develop the recommendations varied slightly between each guideline. All 13 guidelines were based on individual systematic literature searches (using different databases and timeframes), and quality of evidence assessments. Expert panel consensus voting finalized the recommendations in 7/13 guidelines,^{8,9,20,21,23,24,27} with the remaining six reporting multiple discussions and reviews to resolve recommendation formation.^{4,22,25,26,28,29} The recommendations of the 13 clinical practice guidelines are reported in Table I, with a colour coded summary diagram presented in Figure 2. Agreement/disagreement across the guidelines for each recommendation is highlighted in Table II. All 13 guidelines were of high quality. The lowest iCAHE score was 10/14^9,29 and the highest score was 13/14^8,22,26,27 (Table III).

Table I.

Summary of data extraction from all 13 guidelines including: official governing body; location; date of publication; panels of experts; methodology for developing guidelines; non-pharmacological recommendations; and non-pharmacological recommendations against treatments.

Official group, yr (region)	Panel composition	Stated methodology	Recommendation for treatments	Recommendation against treatments
National Institute for Health and Care Excellence (NICE), 2022 (UK)	Consultant cardiologist, consultant rheumatologist, general practitioner, podiatrist, physiotherapist’s, orthopaedic surgeon, radiologist, acupuncturist, manual therapist, and clinical pharmacist	Review questions developed. Systematic literature search via MEDLINE and Cochrane Database of Systematic Reviews. Quality assessed using GRADE. Summaries of presented guidelines and the strength of the research were presented to committee. Expert discussion and interpretation of research. Recommendation given appropriate strength depending on quality of evidence and committee views	Strongly recommended Therapeutic exercise Weight loss Information and support Walking aids Conditional recommendation/adjunct Manual therapy (e.g. manipulation, mobilization or soft-tissue techniques) only alongside exercise Insoles, braces, taping, splints, or supports to be used only where biomechanical correction may help and when exercise is ineffective without one of these devices	Acupuncture Dry needling Electrotherapy (including TENS, NMES, Ultrasound, Interferential therapy, Laser therapy, Pulsed short-wave therapy, neuromuscular electrical stimulation)
Osteoarthritis Research Society International (OARSI) 2019 (International)	International panel of experts with a variety of professional backgrounds: general practice, orthopaedic surgery, rheumatology, sports medicine, and physiotherapy	Key clinical questions developed by expert panel. Systematic literature search via MEDLINE, PubMed, EMBASE, Google Scholar, and Cochrane Databases. Meta-analyses completed on appropriate literature. Quality assessed using GRADE and Cochrane Risk of Bias tool. Recommendation graded either ‘in favour or against’ and strength ‘strong or conditional’. Recommendations voted on by expert panel	Strongly recommended Structured land-based exercise Mind-body exercise (Tai chi or yoga) Weight management Education about OA/self-management Conditional recommendation/adjunct Walking aids (alongside exercise) Cognitive behavioural therapy (alongside exercise in those with widespread pain and/or depression) Aquatic exercise	Massage Wedge insole Acupuncture Electrotherapy (ultrasound) Manual therapy (manipulation and mobilizations) Thermotherapy (hot) Knee braces
American College of Rheumatology (ACR) 2019 (USA)	Interprofessional panel: rheumatologists, internists, and physical and occupational therapists	Systematic literature search via OVID Medline, PubMed, EMBASE, and the Cochrane Library. Quality assessed using GRADE and Cochrane Risk of Bias tool. Expert panel and patient panel discussed the evidence and voted on recommendations	Strongly recommended Land-based exercise Aquatic exercise Self-management programmes Weight Loss Tai chi Walking aids Tibiofemoral knee brace Conditional recommendation/adjunct Thermal agents (hot and cold) Cognitive behavioural therapy Electrotherapy (ultrasound, diathermy, and radiofrequency ablation) Acupuncture Kinesiotaping Balance training Patellofemoral brace Yoga	Strongly against TENS Conditionally against Manual Therapy Massage therapy Modified shoes Wedged insoles Pulsed vibration therapy
American Academy of Orthopaedic Surgeons (AAOS) 2021 (USA)	Three members representing the American Academy of Orthopaedic Surgeons and one member (each) representing the American Orthopaedic Society for Sports Medicine., the American Association of Hip and Knee Surgeons), the Knee Society, the Academy of Family Physicians, the American Physical Therapy Association, the Arthroscopy Association of North America, the International Cartilage Repair Society, the American Medical Society for Sport Medicine, the American Orthopaedic Society of Sports Medicine, and the American Society of Regional Anaesthesia and Pain	Systematic literature search via PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials. Quality assessed using GRADE and Evidence-to-Decision Framework. Voting completed on each recommendation and adopted if majority of 60% of voting panel	Strongly recommended Land-based exercise Aquatic exercise Patient education Self-management Conditional recommendation/adjunct Walking aids Braces Neuromuscular training Weight loss Manual therapy (in addition to exercise) Massage Laser treatment Acupuncture TENS Electrotherapy (percutaneous, electrical nerve stimulation/ pulsed electromagnetic field therapy, extracorporeal shockwave therapy, denervation therapy)	Lateral wedge insoles Dry needling Free floating devices
Italian Society of Rheumatology (ISR) 2019 (Italy)	Rheumatologists, project coordinator, general practioner, physiatrist, orthopaedic surgeon, healthcare professional, and representative of patients association	Systematic approach based on the framework of the Guidelines International Network Adaption Working Group. Systematic literature search via PubMed, Medline and EMBASE. Quality assessed using Agree II Framework	Strongly recommended Patient education/self-management Land-based exercise Aquatic exercise Weight management Orthoses Walking aids Conditional recommendation/adjunct Appropriate/comfortable footwear TENS Taping Unclear recommendation Manual therapy Acupuncture	Nothing stated
European Alliance of Associations for Rheumatology (EULAR) 2013 (Europe)	Two nurses, psychologist, dietician, two occupational therapists, three physiotherapists, five rheumatologists, two orthopaedic surgeons, general practitioner, two persons representing people with hip and/or knee osteoarthritis, clinical epidemiologist, and research fellow	Systematic literature review using MEDLINE and eight other databases. Evidence graded and scored via voting on level of agreement following two task force meeting and extensive discussions	Strongly recommended Structured exercise Weight loss Behaviour changes strategies Education and self-management Aquatic exercise Appropriate footwear Conditional recommendation/adjunct Walking aids	Lateral wedged insoles
European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO) 2019 (Europe)	Rheumatologists, specialists in physical medicine and rehabilitation, clinical epidemiologists, endocrinologists, pharmacologists, orthopaedic surgeons, geriatricians, specialists in public health and health economics, research scientists, and patient representatives	Previous version guideline algorithm discussed by an expert panel and areas of interest highlighted in light of new data. Updated literature searches via MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials, and the Cochrane Database of Systematic Reviews. Quality assessed using GRADE. Summary of recommendations provided to members of working group. 75% agreement of working group vote was considered a ‘strong’ recommendation.	Strongly recommended Information and education Weight loss Structured exercise Conditional recommendation/adjunct Knee braces Insoles Walking aids Thermal agents (hot and cold) Manual therapy Bandage/taping Hydrotherapy/aquatic exercise Tai chi	Nothing stated
Turkish League Against Rheumatism (TLAR) 2018 (Turkey)	23 physicians, 22 physical medicine and rehabilitation specialists (four with rheumatology speciality), and orthopaedic surgeon	Systematic literature search via PubMed, EMBASE, Cochrane and Turkish Medical Index. Articles reviewed for quality and classified in hierarchy in level of evidence. Each item voted on by the panel and strength of recommendation determined.	Strongly recommended Weight loss Education Exercise Conditional recommendation/adjunct Tai chi Aquatic exercise Electrotherapy (ENS, interferential, ultrasound, short wave diathermy) Thermal agents (hot and cold) Balneotherapy Assistive/adaptive devices (insoles and appropriate footwear for biomechanical insufficiency)	Nothing stated
Pan American League of Associations for Rheumatology (PANLAR) 2016 (America)	48 speciality rheumatologists, other medical specialists (orthopaedics and physiatrists), and three patients	Systematic literature search via MEDLINE, PubMed, Cochrane Library and EMBASE. Quality of literature based on Jadad scale. Strength of the recommendation evaluated by the committee based on the American Heart Association Evidence Based Scoring System. Each professional expert asked to contribute independently on the key questions. Consensus reached by a variation of the Delphi technique. 3 rounds before a final document edited by Editorial committee and final texts approved by working groups.	Strongly recommended Education Exercise Conditional recommendation/adjunct Hydrotherapy Thermal agents (hot and cold) Insoles Knee braces Taping (for knee instability) Walking aids Mechanotherapy	Nothing stated
Rheumatology and Immunology Expert Committee of the Cross-Strait Medical and Health Exchange Association (CSHMA), 2020 (China)	Multidisciplinary experts in: rheumatology, orthopedics, rehabilitation, imaging, and evidence-based medicine	Survey used to highlight clinical questions of concern. Systematic literature search via Medline, Cochrane Library, Epistemonikos, China Biology Medicine (CBM), Wanfang, and China National Knowledge Infrastructure (CNKI). Quality assessed using GRADE. Evidence and preference and values of Chinese patients used to formulate recommendations.	Strongly recommended Weight control/loss Health education/self-management Exercise Yoga Tai chi Baduanjin training Aquatic exercise Conditional recommendation/adjunct Manual therapy Massage Acupuncture Avoidance of standing, kneeling, squatting and ascending stairs	Nothing stated
Royal Australian College of General Practitioners (RACGP) 2018 (Australia)	Rheumatologists, physiotherapists, education professionals, clinical advisors, GPs, consultants, exercise and rehabilitation specialists, and epidemiologists	Systematic literature search via PubMed, CINAHL and the Cochrane Library. Searches limited to RCT and systematic reviews. Included if level 1 or 2 on National Health and Medical Research Council (NHMRC) hierarchy of evidence. Quality assessed using GRADE. Working group received initial recommendations and were voted on – 70% agreement for finalised recommendations by working group.	Strongly recommended Land-based exercise Tai chi Weight management Conditional recommendation/adjunct Cognitive behavioural therapy Stationary cycling Hatha yoga Aquatic exercise/hydrotherapy Massage therapy Manual therapy Weight management + exercise Thermal agents (hot) Walking aids TENS Neutral recommendations Self-management and education Varus unloading knee braces Shoe orthotics (medial wedge insoles) Patellar taping Pulsed electromagnetic/shortwave therapy	Thermal agents (cold) Valgus unloading braces Shoe orthotics (lateral wedge insoles) Unloading shoes Kinesio taping Acupuncture Electrotherapy (shockwave, ultrasound, laser)
Department of Veterans Affairs/ Department of Defence (VA/DoD) 2020 (USA)	Physician, service chief adult reconstruction and oncology, programme director for rheumatology fellowship programme, director of clinical practice, internal medicine chief resident, pharmacy manager, orthopaedic surgeon, clinical pharmacist, dietitian, programme director of family medicine, nurse practitioner, physician pain management, two specialists in orthopaedic physical therapy, and director of army internship	Systematic literature search via Medline, EMNASE, PubMed, Cochrane Database of Systematic Review and Agency for Healthcare Research and Quality (AHRQ) website. Quality assessed using GRADE. Findings presented to working group of experts in face-to-face meeting, 3 drafts before finalised recommendations.	Strongly recommended Self-management Exercise Weight reduction Conditional recommendation/adjunct Aquatic exercise Knee bracing Walking aids Physical therapy referral Manual therapy Neutral recommendations Acupuncture Massage Light touch Meditation Tai chi Yoga TENS	Nothing stated
Royal Dutch Society for Physical Therapy (KNGF) 2020 (Netherlands)	Guideline panel: independent chair, physical therapists (from primary and secondary care settings), manual therapist, exercise therapists, representative from Dutch Arthritis Foundation, general practitioner, orthopaedic surgeon, and rheumatologist Review panel: independent chair, OA expert, physical therapists, arthritis charity/foundation representatives, geriatric care specialists, rehabilitation physicians, nurses, dietitian, and podiatrist	Multidisciplinary panel suggested with questions based on current barriers to care. Systematic literature search via PubMed, EMBASE, Web of Science, Cochrane Library Central, EMCARE and CINAHL. Quality assessed using GRADE. Multiple meetings between expert panel to develop recommendations.	Strongly recommended Weight loss Exercise Education and lifestyle advice Conditional recommendation/adjunct Walking aids Knee brace (in very specific cases) TENS (brief intervention to support exercise if hampered by severe pain) Hydrotherapy (if overweight to reduce load)	Nothing stated

RCT, randomized controlled trial.

Fig. 2.

Visual summary of the recommendations made by the 13 guidelines, where green represents the core recommendations; yellow represents treatments as adjuncts or treatments with low level evidence; red represents recommendations against this treatment; and white represents no recommendation.

Table II.

Percentage agreement across guidelines for each recommendation area.

Treatment	Recommended for, n (%)	Acceptable adjunct, n (%)	Recommended against, n (%)	Not mentioned, n (%)
Exercise	100.0 (13)		-	-
Advice/self-management	92.3 (12)	7.7 (1)	-	-
Weight management	84.6 (11)	7.7 (1)	-	7.7 (1)
Aquatic/hydrotherapy	38.5 (5)	53.8 (7)	-	7.7 (1)
Tai chi	30.8 (4)	23.1 (3)	-	46.2 (6)
Yoga	15.4 (2)	23.1 (3)	-	61.5 (8)
Baduanjin	7.7 (1)	-	-	92.3 (12)
Cognitive behaviour therapy	7.7 (1)	23.1 (3)	-	69.2 (9)
Walking aids	23.1 (3)	61.5 (8)	-	15.4 (2)
Massage	-	30.8 (4)	23.1 (3)	46.2 (6)
Manual therapy	-	61.5 (8)	23.1 (3)	15.4 (2)
Acupuncture	-	38.5 (5)	23.1 (3)	38.5 (5)
Electrotherapy	-	15.4 (2)	38.5 (5)	46.2 (6)
TENS	-	46.2 (6)	15.4 (2)	38.5 (5)
Thermotherapy	-	30.8 (4)	23.1 (3)	46.2 (6)
Braces	7.7 (1)	53.8 (7)	15.4 (2)	23.1 (3)
Insoles/footwear	7.7 (1)	30.8 (4)	30.8 (4)	30.8 (4)
Taping	-	38.5 (5)	15.4 (2)	46.2 (6)
Splints/orthotics	7.7 (1)	7.7 (1)	30.8 (4)	53.8 (7)
Avoid certain activities	-	7.7(1)	-	92.3 (12)

Table III.

Quality assessment of guidelines using the International Centre for Allied Health Evidence quality assessment tool.

	Availability			Dates			Underlying evidence				Guideline developers		Guideline purpose and users	Ease of use	TOTAL score (max 14)
Guideline	Is the guideline readily available in full text? Max (1)	Does the guideline provide a complete reference list? Max (1)	Does the guideline provide a summary of its recommendations? Max (1)	Is there a date of completion available? Max (1)	Does the guideline provide an anticipated review date?	Does the guideline provide dates for when the literature was included? Max (1)	Does the guideline provide an outline strategy used to find underlying evidence? Max (1)	Does the guideline use a hierarchy to rank the quality of the underlying evidence? Max (1)	Does the guideline appraise the quality of the evidence which underpins its recommendations? Max (1)	Does the guideline link the hierarchy and quality of underlying evidence to each recommendation?	Are the developers of the guideline clearly stated? Max (1)	Does the qualifications and expertise of the guideline developer(s) link with the purpose of the guideline and its end users? Max (1)	Are the purpose and target users of the guideline stated? Max (1)	Is the guideline readable and easy to navigate? Max (1)
NICE	1	1	1	1	0	1	1	1	1	0	1	1	1	1	12
OARSI	1	1	1	1	0	1	1	1	1	1	1	1	1	1	13
ACR	1	1	1	0	0	1	1	1	1	1	1	1	1	1	12
AAOS	1	1	1	0	1	0	1	1	1	1	1	1	1	1	12
PANLAR	1	1	1	0	0	1	1	1	1	1	0	1	1	1	11
Va/DoD	1	1	1	0	0	1	1	1	1	1	1	1	1	1	12
CSHMA	1	1	0	1	1	1	1	1	1	1	1	1	1	1	13
RACGP	1	1	1	0	1	1	1	1	1	1	1	1	1	1	13
EULAR	1	1	1	0	0	1	1	1	0	0	1	1	1	1	10
ESCEO	1	1	1	0	0	1	1	1	1	1	1	1	1	1	12
TLAR	1	1	1	0	0	1	1	1	1	1	1	1	1	1	12
KNGF	1	1	0	1	0	1	1	0	1	0	1	1	1	1	10
ISR	1	1	1	1	0	1	1	1	1	1	1	1	1	1	13

AAOS, American Academy of Orthopaedic Surgeons; ACR, American College of Rheumatology; CSHMA, Cross-Strait Medical and Health Exchange Association; ESCEO, European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases; EULAR, European Alliance of Associations for Rheumatology; ISR, Italian Society of Rheumatology; KNGF, Royal Dutch Society for Physical Therapy; NICE, National Institute for Health and Care Excellence; OARSI, Osteoarthritis Research Society International; PANLAR, Pan American League of Associations for Rheumatology; RACGP, Royal Australian College of General Practitioners; TLAR, Turkish League Against Rheumatism; Va/DoD, US Department of Veterans’ Affairs.

Treatment recommendations

Core recommendations (strong agreement): There are three core recommendations that guidelines mostly agree on as the main conservative, non-pharmacological, management of knee OA. Exercise was advised in 13/13 (100%) guidelines. Self-management advice and education in 12/13 guidelines (92.3%), with a single neutral recommendation due to perceived insufficient evidence to advise for or against.²⁷ Weight management was advised in 11/13 guidelines (84.6%). The American Academy of Orthopaedic Surgeons (AAOS)²¹ recommended weight management as an adjunct to exercise and education, and the Pan American League of Associations for Rheumatology (PANLAR)²⁵ group made no mention of weight management.

Conditional/adjunct recommendations (contested evidence)

Alternative forms of exercise: Various, specified, alternate forms of exercise were, variously, individually considered. Hydrotherapy/aquatic exercise is advised by 5/13 of the guidelines (38.5%) as a core treatment option.^9,20-22,26 A further 7/13 guidelines (53.8%) either advised water-based exercise as an adjunct or report that the evidence is weaker than for core recommendations.^{8,23-25,27-29} NICE takes no view on hydrotherapy as part of their recommendations.⁴

Tai chi is recommended as a core treatment in 4/13 guidelines (30.8%),^8,20,26,27 and as an adjunct to other forms of exercise in a further 3/13 guidelines (23.1%) due to its short-term effects.^23,24,28 Similarly, 2/13 guidelines (15.4%) recommended yoga as core treatment^8,26 and a further three (23.1%) recommended it as an adjunct.^20,27,28 The Rheumatology and Immunology Expert Committee of the Cross-Strait Medical and Health Exchange Association guideline²⁶ also recommends a traditional Chinese form of exercise known as Baduanjin. Cognitive behavioural therapy (CBT) was recommended as part of treatment by a single guideline,⁹ but as an adjunct to exercise for those with widespread pain or depression by a further 3/13 bodies (23.1%).^8,20,27

Manual adjuncts, bracing and orthoses: Bracing, insoles, orthotics, and taping are variously considered; however, the recommendations offered differ. A single guideline²⁰ recommend knee bracing as a core part of conservative management, with a further 7/13 guidelines (53.8%) recommend this as an adjunct to exercise to increase confidence of movement when joint instability is present.^{4,21-23,25,28,29} Conversely, 2/13 guidelines (15.4%) recommend against the use of braces,^8,19 while the remaining three make no mention. Few recommendations consider the use of splints/orthotics, with only one guideline²² recommending it as routine treatment, and another⁴ advising in the presence of joint instability, where exercise is ineffective without this support. Conversely, 4/13 guidelines (30.8%)^8,9,20,27 recommend strongly against the use of orthotics/splints. Shock absorbing insoles and footwear are recommended by a single guideline (7.7%),⁹ though as an adjunct intervention by a further 4/13 (30.8%) to control malalignment of the knee.^4,23-25 However, 4/13 guidelines (30.8%) recommend strongly against the use of insoles.^8,20,21,27 Taping is not strongly recommended as an intervention by any guideline, but is considered an acceptable adjunct recommendation (due to limited evidence) by 5/13 guidelines (38.5%), specifically for issues with knee instability.^{4,20,22,23,25} Futhermore, 2/13 guidelines (15.4%)^8,27 recommend against the use of taping, and the other 6/13 (46.2%) make no reference to this.^{9,21,24,26,28,29}

The use of walking aids is widely agreed on throughout the guidelines. Although only 3/13 (23.1%) specify as a core intervention,^4,20,22 a further 8/13 (61.5%) suggest this as adjunct alongside exercise to help with stability.^{8,9,21,23,25,27-29}

Manual therapies: No guideline considers manual therapy to be core conservative management for knee OA, and 3/13 (23.1%) specifically advise against this.^8,20,29 A total of 8/13 guidelines (61.5%) propose manual therapy as an adjunct intervention.^{4,21-23,25-28}

Massage is recommended as an adjunct to core treatment in 4/13 (30.8%) guidelines.^21,26-28 The same 3/13 (23.1%) which advise against manual therapy also advise against massage.^8,20,29

Acupuncture is not recommended by any group as a core treatment. 5/13 guidelines (38.5%) considered this a reasonable adjunct intervention^20-22,26,28 based on limited evidence. It is recommended against by 3/13 guidelines (23.1%).^4,8,27

Electrophysical and thermal agents: No guidelines recommend electrotherapies as a core treatment and only 2/13 (15.4%) recommend these as an adjunct.^21,24 Conversely, 5/13 guidelines (38.5%) recommend against the use of any type of electrotherapy.^4,8,20,27,29

In all, 6/13 guidelines (46.2%) separately consider, and specifically recommend the use of TENS as an adjunct to other core treatments due to its short-term pain relief.^{21,22,24,27-29} Notably, although the Royal Australian College of General Practitioners (RACGP)²⁷ and the Royal Dutch Society for Physical Therapy²⁹ advise against other types of electrotherapies, they do give conditional recommendations for the use of TENS. However, another 2/13 guidelines (15.4%) specifically reject the use of TENS.^4,20

The use thermotherapy was recognized in 4/13 guidelines (30.8%) as a useful adjunct for short-term pain relief,^20,23-25 though conversely, 3/13 (23.1%) recommended against this.^8,27,29

Discussion

Through systematic search and review we found and compared 13 current clinical practice guidelines for the conservative management of knee OA from national and international representative bodies and advisory groups. These publications review evidence as to the effectiveness of interventions for knee OA and offer recommendations as to the appropriateness of use. The recency of these documents ensures that the same underlying evidence is used to inform these recommendations; however, each guideline group has performed their own evidence search, synthesis, and consensus process, which allows for disagreement in interpretation.

The core recommendations provided across the 13 guideline documents are broadly consistent: exercise; self-management/education; and weight management are suggested to be the effective strategies. The strongest evidence supports the use of these interventions and approaches, which should form the bedrock of clinical management for those with knee OA. The clinically meaningful benefits of therapeutic exercise to address physical dysfunction, pain and quality of life for those with knee OA are well established.^30,31 Guidelines reflect this, but promote variations of exercise interventions amid a general lack of clarity as to type, dosing, duration and loading required. A comparatively detailed recommendation is provided by the RACGP,²⁷ who specifically recommend exercise performed on land, including muscle strengthening, stretching/range of motion, aerobic conditioning, neuromuscular/balance, cycling, Tai chi, and yoga. They state that exercise dosage can vary in frequency, intensity and duration, that the exercise can be delivered in a group setting or individually, either in-person or remotely via telephone or videoconference. In contrast, the AAOS²¹ guidelines simply suggest that ‘supervised exercise, unsupervised exercise, and/or aquatic exercise are recommended over no exercise to improve pain and function for treatment of knee OA’. While this ambiguity as to intervention content and delivery facilities easily tailored interventions, responsive to the varying needs and abilities of individuals, the lack of consensus at to what an appropriate exercise intervention looks like allows huge practice variation in clinical delivery of rehabilitation and the effectiveness of these interventions may be hard to contextualise.

Similarly, patient education and self-management strategies are also well-established as being effective in knee OA.³² These are recommended across the 13 guidelines; however, there is little consensus as to what education and self-management should actually look like. The US Department of Veterans’ Affairs (VA/DoD)²⁸ define self-management as ‘the active engagement of individual patients in daily tasks or activities that aim to improve their quality of life and long-term management of the health condition’. In contrast, the RACGP²⁷ take an equivocal stance on formal face-to-face self-management education programs due to an interpreted lack of evidence. However, they do suggest that clinicians should provide information to enhance understanding about OA, its prognosis and its optimal management. Providing education as to knee arthritis is not the same intervention as providing ongoing coaching-based support to build skills and confidence in coping with arthritis. The Arthritis and Musculoskeletal Alliance (ARMA) policy potion on supported self-management³³ highlights that information provision alone is unlikely to motivate behaviour change. Rather, it proposes that self-management techniques such as referrals to third sector services, mentoring, peer support, helplines, apps, training and signposting are more appropriate. NHS England further notes that ‘patient activation’ underpins the support people need to develop their capability to manage their own health, and that information is required that can be understood and acted upon alongside tailored active support.³⁴

In those that are overweight, a 5% to 10% reduction in body weight has been well documented to reduce pain levels and a dose-response relationship has been proposed by various authors.^35,36 This is well supported across the guidelines, though interestingly, the PANLAR group²⁵ make no mention of weight management and the AAOS guidelines²¹ propose weight management as an adjunct recommendation only. This is a downgrade from previous AAOS guideline versions, due to a considered inconsistency in evidence for symptom improvement. This is at notable odds to the other author groups, and a good example of differential interpretation of the same evidence base across guideline producers. Further, though supported in principle, the delivery of weight loss interventions is not well described. Both self-monitored weight management programmes and clinically supported programmes are highlighted. Although somewhat out with the scope of this particular review, we note the current excitement around GLP-1 agonists which may offer effective pharmacological weight management. Should these agents be successfully employed, the link between obesity, knee OA and pain may drive the next guideline update.

Although broad clarity is achieved across guidelines where the evidence base is consistent (exercise, self-management and weight loss), there is more controversy in the interpretation of effectiveness of well used interventions such as bracing, manual therapy, acupuncture and TENS. Corrective wearable supports have been used in knee OA for years, under the intuitive justification that modifying a mechanical varus or valgus deformity is beneficial in terms of symptom management. Recommendations regarding bracing, insoles and orthoses are the most contentious aspect across the guidelines and are the only interventions to be both strongly recommended for and against usage. The American College of Rheumatology (ACR)²⁰ state that a tibiofemoral knee brace is strongly recommended when gait is poor due to pain or instability, whereas the Osteoarthritis Research Society International (OARSI)⁸ guidance strongly recommends against bracing in any situation. Various randomized trials have been conducted in this area with inconsistent results.^37-39 Subgroup analysis hints at the greatest effect being in ‘offloader’ braces reducing valgus alignment. Reflecting this, the RCAGP²⁷ guidance advises against bracing generally, but offer a neutral, conditional, recommendation for brace use in correcting valgus deformities (stated to be based on poor quality trial evidence). Similarly, NICE⁴ take a pragmatic view, advising against the routine offering of bracing, but proposing as an adjunct intervention ‘to enhance the ability to exercise’ in specific situations of abnormal loading.

This inconsistency as to evidence interpretation is further highlighted looking at manual therapy techniques. No guideline group suggests any strong evidence of effect in knee OA, and most (8/13) offer a recommendation for this as a potential adjunct treatment,^{4,21-23,25-28} though 3/13 specifically recommend against.^8,20,29 The debate as to effectiveness is essentially centred around a single three-arm randomized controlled trial (RCT) that evaluated exercise alone vs exercise plus manual therapy vs a placebo group.⁴⁰ Short term improvements (two months) were seen in the combined exercise plus manual therapy group; however, any initial affect was lost with further follow-up. Some guideline groups felt this was enough to warrant a conditional recommendation, others that the level of evidence was insufficient. The effect of massage is similarly contested. This treatment is considered by only 7/13 the guidelines^{8,20,21,26-29} with 4/13 offering conditional support as an adjunct treatment^21,26-28 and 3/13 recommending against.^8,20,29 As with manual therapy, key data comes from a single RCT,⁴¹ which proposes a short-term effect ( four weeks) that is lost at further follow-up.

The recommendations as to acupuncture were also split. Again, no guideline recommends this as treatment with strong evidence; however, 5/13 suggested it as an acceptable adjunct treatment^20-22,26,28 while 3/13 rejected its use.^4,8,27 A variety of trials have been conducted in this sphere, though methodological issues relating to type of intervention (traditional acupuncture, electroacupuncture, or laser) and the control groups employed complicates comparison across studies. Low level evidence suggests a small (poorly quantified) effect compared to sham acupuncture in knee OA; however, both NICE⁴ and RACGP²⁷ guidelines state a lack of benefits to recommend use, and some evidence of potential harm, justify to recommend against.

Having previously been a popular intervention, TENS is now specifically advised against by both NICE⁴ and ACR²⁰ guidelines. The AAOS²¹ also recently downgraded the recommendation for TENS due to ‘inconsistencies in evidence’. Australian, Dutch, Turkish and Italian guidelines^22,24,27,29 however, all state that TENS may be useful for short-term pain relief and conditionally support its use. Limited evidence does suggest TENS may have some immediate analgesic effect, however a well-constructed recent RCT versus sham TENS concluded no improved pain in knee OA compared to placebo.⁴² NICE⁴ advise against the use of any form of electrotherapy due to poor evidence and multiple other guideline groups agree a general lack of benefit.

Strengths of this review are that we carried out a comprehensive search of online databases and the Guideline International Network. A specific limitation of this work was the necessity to search for information presented in English, which may limit the international generalization of our findings. However, we note that many of the guidelines that we were able to include are from countries where English is not the native language, highlighting the way in which research tends to be reported in the international literature, and that clinical practice guidelines from major bodies tend to be available in multiple language translations. Further the guidelines included were themselves based on evidence summaries of the international literature that those bodies carried out. It is possible though that we have missed knee OA interventions which are supported in different cultural contexts.

The guidelines identified were all of high quality as assessed by the iCAHE tool, which allowed a comparison to be made of all 13 looking at consistencies and differences between the recommendations. While there may be local healthcare system, environment, or cultural reasons supporting the delivery of specific knee OA management strategies, or reasons for individual interventions to be offered to specific patients, we highlight a problematic inconsistency in the clinical practice guidelines that should underpin these management decisions. Despite the individual guidelines being of solid methodology (all based on data drawn from systematic review with GRADE criteria applied) the consensus opinions achieved as to the evidence quality for, and thus usefulness of, intervention varied.

Though conjectural, we suggest that where there is inconclusive effectiveness evidence, differing intrinsic biases in the guideline groups may influence their recommendation. It may be that the geographical location of the guideline body, or the professional groups reflected in the panels influence the decision making through local healthcare and cultural norms. For example, surgeons and physiotherapists may be more likely to instinctively support a ‘conditional use’ recommendation as opposed to a ‘do not use’ recommendation for ‘offloader’ braces through prior experience of these. Patients like to receive passive treatments alongside self-management and exercise-based interventions⁴³ creating an incentive to provide this, particularly in settings without centralized funding. Notably, additional mechanical support through bracing is (with poor evidence of effectiveness) particularly popular among patients, whereas walking aids (with better evidence) are unpopular, due to the stigma of use.¹⁷ Complementary treatments such as acupuncture, and those with even less of an evidence-base, are also popular among arthritis sufferers, proposedly because those that provide these treatments are more accessible and more empathetic.¹⁷ Recent data suggests provision of these contentious ‘low value’ interventions persists despite poor evidence of effectiveness.⁴⁴ The provision of such interventions is at the discretion of the health provider and payer, though delivery of passive interventions with low evidence of effectiveness may affect the success of symptom management - particularly if used instead of well-evidenced active management. Although the guideline groups disagree in their recommendation as to supporting the adjunct use of, or opposition to, the ‘non-core’ treatments, they do present a unified view as to a lacking evidence base to support any intervention out with exercise, education and weight loss.

This differing interpretation as to value of conservative management recommendation is potentially confusing for those with knee OA, the clinicians, therapists, healthcare commissioners that refer for, and provide, such treatments. Personal management opinion can currently be supported through ‘shopping’ for the most favourable guideline to justify it. A unified consensus opinion across guidelines is desirable and harmonization of evidence syntheses across guideline providers would be beneficial, though is perhaps unlikely considering the varying ‘jurisdiction’ of each group. It would though be useful if future guideline developers considered the other groups evidence weighting and recommendations in the creation of their own, and offered reasons for any discrepancy. Healthcare decision making based on evidence suggests that currently only the core treatments of exercise, education and weight loss should direct the clinical management of knee OA.

Take home message

- There is broad consistency across clinical practice guidelines regarding the core conservative management of knee osteoarthritis (OA) (exercise, education, and weight loss), which should direct and support patient care.

- There is disagreement though across clincal practice guidelines as to the benefits of various other interventions for knee OA.

- Differences in evidence interpretation result in the different recommendations, which may be confusing for patients, for physiotherapists providing conservative management, and for wider medical personnel referring to these services.

Data Availability

All data generated or analyzed during this study are included in the published article and/or in the supplementary material.