The burden of non-communicable diseases, such as osteoarthritis (OA), continues to increase for individuals and society. Regrettably, in many instances, healthcare professionals fail to manage OA optimally. There is growing disparity between the strength of evidence supporting interventions for OA and the frequency of their use in practice. Physical activity and exercise, weight management and education are key management components supported by evidence yet lack appropriate implementation. Furthermore, a recognition that treatment earlier in the disease process may halt progression or reverse structural changes has not been translated into clinical practice. We have largely failed to put pathways and procedures in place that promote a proactive approach to facilitate better outcomes in OA. This paper aims to highlight areas of evidence-based practical management that could improve patient outcomes if used more effectively.
- physical activity
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Our knowledge of osteoarthritis (OA) continues to grow. No longer described as joint ‘wear and tear’ in the absence of inflammation,1 OA is characterised by chronic abnormal remodelling affecting the entire joint, the outcome of which can be structural and functional failure.2 Hip and knee OA are considerable contributors to disability worldwide. The projected trend for the disease is ominous and poses a significant challenge to society and healthcare. It is estimated that by 2040, approximately 25% of the US adult population (78.4 million people) will have diagnosed OA, 34.6 million of whom will experience activity limitations attributable to this OA.3 This mirrors what is happening in other Western countries where the burden of OA on healthcare systems is vast. With the expected increase in OA prevalence, it is reasonable to assume that this can only get worse. The continued shifting of demographic and societal factors, including population ageing, obesity and financial austerity, has negatively affected the continued (and ultimately unsustainable) growth in OA burden. In 2011, a review outlined the need to comprehensively overhaul lower limb OA management.4 Since then, the healthcare community has largely failed to address these recommendations and translate evidence into clinical practice. This paper provides a contemporary view to support the original narrative review, given the lack of improvement in care provision and the growing discrepancy between the strength of evidence supporting interventions (especially relating to primary and secondary disease prevention) and their implementation. The authors seek to re-emphasise the need to revise the way we consider musculoskeletal health, particularly OA. We outline three key evidence-based therapeutic tools that should be used in OA management (physical activity (PA) and exercise, weight management and patient education) and describe barriers preventing best practice. In addition, we present opportunities to improve patient care through research incorporating broader stakeholder engagement and strategies to facilitate its translation into practice.
Best practice in OA management (prevention and treatment)
OA management should aim to reduce pain, enhance function and improve quality of life. Three critical areas of best practice are described below.
Physical activity and exercise
PA is a vital component of a healthy lifestyle. While there is a dearth of evidence linking PA and reduced risk of OA, PA plays a prominent role in obesity management—a known risk factor for OA—and is important in the prevention and/or management of other comorbidities.5 Guidelines encourage individuals to complete a minimum of 150 mins of moderate activity per week for multiple and wide-ranging benefits, yet a high proportion of the population fails to do so. In both advanced and low-income/middle-income countries, trends suggest that substantial increases in sedentary behaviours and further reductions in overall PA levels will persist.6 Evidence suggests those with or at risk of OA seldom achieve national PA recommendations. Of more than 1000 individuals with knee OA, just 10% met the PA guidelines.7 White et al 8 found that neither pain nor radiographic changes accounted for high levels of inactivity.8
Exercise, a subcategory of PA, is the cornerstone treatment for OA and is critical in secondary and tertiary disease prevention. Therapeutic exercises aim to increase muscular strength, neuromuscular control, aerobic endurance and joint movement across the disease spectrum, including predisease. High-quality systematic reviews and meta-analyses have concluded that injury prevention protocols should include neuromuscular training combining sport-specific strength, balance and agility components.9 A Cochrane review highlighted the benefits of exercise for people with knee OA, including reduced knee pain, improved quality of life and enhanced functional capacity.10
Weight loss reduces the risk of developing symptomatic OA, likely by modifying both biomechanical forces and metabolic factors.11 Numerous studies highlight the substantial beneficial outcomes including improved joint pain, function and quality of life in those with established OA.11–13 Weight loss method may impact on the efficacy and effectiveness of the intervention. Guidelines typically recommend exercise as well as weight loss.14 15 Messier et al 12 reported the greatest benefits in pain and function when diet and exercise were combined.12 Regardless of the method (independent diet or exercise, or combined), weight loss improves pain and function associated with OA. Weight loss can also be beneficial to patients with severe OA, including those appropriate for joint arthroplasty. Obesity complicates surgical management of OA; it lengthens hospital stay postsurgery (compounded by the presence of comorbidities) and is associated with greater risk of complications and joint arthroplasty revision. Obesity is a growing concern for joint arthroplasty; one study found obesity rates among those requiring joint arthroplasty revision rose from 9.7% to 24.6% over 10 years.16
There are several important messages that must be given to patients with OA and the wider population. First, meeting the PA guidelines will not increase the risk of OA.17 Second, therapeutic exercise does not result in worsening symptoms, structural OA progression or functional decline in those with joint pain.18 Third, the multiple and extensive benefits of a healthy and active lifestyle extend far beyond musculoskeletal health and should be strongly endorsed. While reinforcing these key messages, education aims to equip patients with sufficient knowledge and motivation to take on effective self-management, irrespective of OA severity, through an interactive and engaging process. Education should ultimately stimulate positive behavioural changes that reduce risk of OA onset or progression and be complementary, or ensure adherence, to other clinical interventions. While no single strategy will universally increase exercise adherence or reduce barriers to exercise for all patients, some emerging techniques have been explored. Evidence suggests that returning for further physiotherapy after an initial exercise period and promoting behavioural graded exercise (the self-regulation of gradual increases in exercise incorporated into activities of daily living) may be beneficial.19 It has been shown that group education can help achieve clinical benefits in individuals with hip and/or knee OA. Data from the Good Life with osteoArthritis in Denmark (GLA:D) demonstrate that multimodal initiatives are feasible and can contribute towards reducing pain and increasing function.20 Individualising self-management according to the attitudes and capabilities of each patient is likely to improve outcomes. In addition, clinicians have a professional responsibility to provide accurate and comprehensive information regarding the risk/benefit of alternative treatment options (eg, medications and surgery).
Limitations of current care
Across healthcare sectors, deficiencies and barriers prevent optimal management in OA. The detrimental effects of failing to implement the key components described above are compounded by additional factors, some of which are briefly discussed in the following section.
Failure to shift focus to prevention
The two foremost risk factors for OA development—obesity and joint injury21—persist and continue to increase worldwide.
Obesity is a modifiable risk factor for OA and other chronic diseases. Unfortunately, efforts to reverse the ongoing global obesity epidemic are failing. Reports indicate more than one in three English children are overweight or obese,22 and a similar proportion in America are affected.23 It is projected that by 2030, the prevalence of obesity will increase 33%, meaning that over 40% of the American population will be obese.24 Greater body mass index (BMI), increased body weight and obesity are all associated with a heightened risk of developing knee OA.21 25 26 The attributable risk of obesity for OA-indicated knee joint replacement is 31%.26 Interactions between abnormal metabolites, body composition and inflammatory mediators are implicated in altered joint structure and milieu found in OA.27 Furthermore, a systemic association between obesity and OA is likely (eg, increased prevalence of hand OA in the obese); thus, OA may need to be considered a metabolic disorder.28 29
Government and organisational support of large-scale programmes to promote healthy weight and prevent sedentariness would likely reap long-term benefits. Controlling the rising tide of global obesity would have a phenomenal impact on the prevalence of symptomatic OA and ultimately reduce the number of joint replacements. Approximately half of all cases of symptomatic OA in the USA and over 40% of cases in the UK and Australia would be prevented if obesity were eliminated in those respective countries. This would equate to a 60% reduction in the number of knee replacements required in the USA.30 Eliminating obesity is a nigh impossible challenge, but even relatively modest reductions in body weight across a large population would have a hugely beneficial effect. As such, clinicians must aim to promote positive changes in those they can influence and health policy should reflect these concerns.
Following joint trauma, 20%–50% of people develop OA (around 12% of all OA cases).31 Injury surveillance within the general population is challenging, and available data only hint at true injury rates. Longitudinal studies show a 26% increase in the number of lower limb sports-related injuries over just a 7-year period.32 Injuries to the ACL or knee meniscus substantially increase susceptibility to OA. A systematic review identified the risk of developing moderate or severe radiographic changes following ACL injury increased by a factor of 3.84 compared with controls.33 An analysis of patients 10–15 years post-ACL reconstruction identified 71% developed radiographic knee OA (Kellgren and Lawrence grade ≥2) in the injured joint and 25% had radiographic OA in the contralateral knee. Those with radiographic OA reported significantly greater symptoms compared with those without. Cases of severe OA (Kellgren and Lawrence grade 4) were associated with more pain and symptoms, reduced levels of function and impaired quality of life.34 There is a startling (and early) risk of cartilage loss at different anatomical sites following ACL injury.35 After 7–11 years, the lateral and medial femoral condyles and the patella exhibit greatly increased risk of cartilage loss, 50-fold, 19-fold and 30-fold increases, respectively, with accelerated chondral degradation at 5–7 years postinjury. Since severe joint injuries are becoming increasingly common among adolescents, the timeframes outlined above highlight how early-onset post-traumatic OA can elongate the morbidity burden over a vast proportion of life.
The increased risk of OA following injury is not unique to the knee. Musculoskeletal injury is also associated with a fivefold increased risk of developing hip OA.36 Primary ankle OA is rare, but approximately 80% of cases develop secondary to trauma.37
Injury prevention schemes have been implemented successfully across some sports and should be encouraged and more widely adopted.38 Evidence supports the beneficial effects of injury prevention programmes when executed appropriately and consistently in specific populations. Pooled estimates report risk reductions for ACL injury at 52% and 85% for female and male athletes, respectively.39 It is logical that decreasing the number of injuries will reduce OA incidence, but long-term studies are necessary to prove this association. Programme adherence remains an issue, with gaps remaining in the knowledge of sporting communities, highlighting the need to engage and educate athletes, coaches and parents at all levels.9 40
When joint injury does occur, current management may not adequately prioritise long-term musculoskeletal health, which may contribute to increased risk of OA secondary to the injury itself. It is unclear whether clinicians accurately describe, and patients comprehend, the potential life-course consequences of injuries such as ACL rupture—particularly to young athletes—and the risks associated with a return to high-level sport. Guidelines for rehabilitation following injury provide extensive evidence relating to the period until return-to-play, but little beyond that timeframe.41 Clinicians and carers should consider the ethics of management, and whether a given course of action is in the athlete’s long-term interest.42 In addition, the detrimental consequences of injury at a young age may be compounded by other factors. Within 3–10 years of an intra-articular knee joint injury, the odds of being overweight or obese are raised compared with uninjured controls.43 The presence of excess inflammatory mediators together with the injury may expedite OA progression. Furthermore, negative psychological consequences postinjury (eg, reduced knee confidence) are common. This could be an underappreciated factor for the development of secondary OA. Supplementary exercises or incorporating psychological interventions could improve OA outcomes in this specific group and should be the focus of further research.44
Lack of early diagnostic criteria and intervention
The importance of intervention during the early stages of OA (including presymptomatic disease) is increasingly acknowledged. Early OA should present an opportunity to positively influence, and possibly reverse, changes that occur to the joint and periarticular structures.45 Failure to do so ultimately contributes to worse clinical outcomes (figure 1); however, a strong evidence base for the management of early OA is lacking.
Defining and diagnosing early OA remains challenging, as symptoms can be sporadic or absent and radiological evaluation is often unhelpful, with a disconnect between symptoms and radiographic findings. No standard definition exists for early OA despite efforts to provide specific classification criteria.46 Poor diagnostics and inadequate initial management may lead to suboptimal care in early disease, jeopardising a time-critical phase when certain interventions could provide the greatest benefit.47 In reality, OA has several phenotypes and is a continuum with variable signs and symptoms at each stage and in each joint, posing real challenges for clinicians. Sufficient disease knowledge, identification of validated biomarkers and proficient clinical skills to diagnose OA are precursors to the provision of time-appropriate management.
Inadequate implementation of clinical guidelines
The implementation of existing clinical guidelines for OA is inadequately reflected in management, which consequently remains substandard.48 The limited uptake of OA guidelines is an indictment of our practice, given the evidence and clinical guidelines available.49 The fact that there is strong agreement between guidelines from different sources and organisations suggests that it is not a lack of consistency that is obstructing their uptake. Recent clinical algorithms provide an alternative, stepwise approach to OA management by identifying commonalities across guidelines to enhance their clinical use.50 It is argued that potential barriers from the wider organisational context of healthcare services down to the decision-making of individual practitioners can prevent evidence translation into practice. The interactions within and between these complex and multiple levels may influence the adoption of new clinical practices in primary care.51 Implementing best practice is often reliant on the necessary infrastructure and resources being in place, which may not always be the case.
Persistent use of treatments when evidence is lacking
We underscore that some popularised interventions continue despite little or no evidence of efficacy.52 Arthroscopic surgery for the management of hip or knee OA is not recommended. It provides minimal, inconsequential benefits in those with pain or symptoms indicative of knee OA.53 A systematic review found that arthroscopic hip surgery improved short-term outcomes, but clinical gains are reduced in patients with OA compared with those without. Greater severity of OA and older age were associated with higher risk of, and quicker progression to, arthroplasty. In addition, the reported beneficial effects of surgery may be inflated due to trial bias.54
The use of some analgesics is ill-advised as the evidence does not support their continued widespread use or recommendation in guidelines. A recent study concluded opioids were not cost-effective and provided no additional treatment benefit, except in patients suitable for, but averse to, undergoing a total knee arthroplasty.55 Despite huge financial costs and potentially serious side effects, the use of opioids continues to increase. Between 2003 and 2009, the proportion of patients with OA prescribed opioids rose from 31% to 40%.56 A systematic review found the effectiveness of acetaminophen in either reducing pain or improving function was not meaningfully superior to that of placebo.57 This was confirmed in another review which concluded acetaminophen should not be recommended for use in the treatment of OA.58 In addition, this study assessed non-steroidal anti-inflammatory drugs, highlighting significant variation in efficacy and outcomes related (in part) to dose. A meta-analysis emphasised the significance of delivery methods (eg, oral, intra-articular) for different medications and subsequent variation in effectiveness.59 Therefore, the type, dose and the delivery method are important considerations when seeking to optimise pharmacological clinical outcomes.
There has been increased scrutiny recently over the use of some traditional tests and treatments unlikely to provide patient benefit—and potentially causing harm.60 One outcome of this interest is ‘Choosing Wisely’, an American Board of Internal Medicine initiative (which lists the American College of Rheumatology, American Medical Society for Sports Medicine and American Academy of Orthopaedic Surgeons among partner societies). This aims to facilitate better informed conversations between clinicians and patients regarding which treatment options comprise best practice. When used in combination with guidelines advocating effective treatment options, the culture of ‘Choosing Wisely’ offers potential opportunities for clinicians treating OA. Assessing the impact of ‘Choosing Wisely’ is challenging and multiple barometers of effectiveness (healthcare provider attitudes and awareness, overuse of low-value services, patient perceptions and outcomes) coexist.61 That said, the principles behind the movement appear to be gaining traction with relevant clinical specialities and occupations. For example, it has been recommended that physiotherapy training curricula are reviewed so students are educated on the concept of low-value healthcare.62 The ultimate effect of the campaign on OA may take time to emerge.
Inappropriate use of imaging
There is a poor correlation between symptoms and imaging in OA. The use of unnecessary imaging (both X-ray and MRI) or dependence on it to confirm what is already known following thorough examination is financially costly and typically clinically redundant. MRI is considered superior to radiograph in exploring early joint changes.63 MRI can evaluate more anatomical structures in the joint than X-ray, and this may provide future direction in early OA management. The recent development of semiquantitative MRI scoring methods for features of knee OA may improve disease-modifying interventions, since certain biomarkers appear to correspond to disease progression.64 Nevertheless, the clinical interpretation of MRI findings should be treated with caution. One or more abnormalities suggestive of OA are found in almost 90% of MRIs conducted on patients with a normal X-ray, and abnormalities are common regardless of whether pain is present or absent.65 Importantly, investigations must positively contribute to clinical decision-making; however, this is not always the case. Evidence shows that 48% of MRIs conducted on patients aged over 40 with knee pain did not influence treatment choices.66
Lack of diagnostic specificity
Management of OA should be joint-specific and consider the affected compartments within each joint to improve outcomes. The knee joint should not be considered as a single articulation where only the tibiofemoral joint (TFJ) is susceptible to OA. It is well established that isolated radiographic TFJ OA is rare in symptomatic patients. Patellofemoral joint (PFJ) OA has been underappreciated, but recent evidence suggests that it may be particularly relevant following ACL reconstruction.67 68 Of all patients who undergo such surgery, nearly half present with radiographic PFJ OA 5–10 years later.68 Targeted interventions specifically for the PFJ, rather than the knee in general, will enhance their effect in these patients, but this is reliant on adequate clinical assessment and knowledge of the treatment modalities that work best in these specific cases.
Bridging the evidence to practice gap
The Osteoarthritis Research Society International (OARSI) methodological recommendations for research69 aim to support production of high-quality information that contributes to enhanced guidelines and clinical care. Research priorities should be agreed with input from end users (clinicians, patients, patient advocacy groups and policymakers) to ensure questions relevant to patient care are identified and addressed. The benefits of involving patients and the public at the centre of healthcare improvements are widely acknowledged, but a shift towards partnership rather than consultancy is required.70
The trivialisation of OA as a condition and inadequate dissemination of best practice contribute to poor outcomes.71 72 It is recommended that to improve uptake, OA guidelines must be made applicable to real-world situations, formatted for ease of use and readily accessible.73 Perhaps it would be wise to consider a broader role for implementation science to help bridge the gap between evidence and practice.74 Implementing evidence involves multiple stakeholders acting within and around complex healthcare systems.75 Greater understanding of knowledge translation in OA would be beneficial for all involved.
There is a gap between evidence and practice, but there are other areas where gaps exist. It could be argued that the understanding and actions of influential groups and organisations, including the media and research funders, may not always be aligned with current evidence. Closing any gaps could facilitate improved patient outcomes. For example, healthcare professionals with a full understanding of the evidence working with the media to counteract unvetted or non-peer-reviewed public information could raise the profile of OA, improve early symptom recognition and aid patient treatment choices.
Implementing measures that prevent OA is difficult due to traditional models of healthcare; patients present at clinic once symptoms have developed rather than being accessible when interventions may be used to prevent onset. The challenge is to develop large-scale, long-term studies—including cohorts that reflect real-world patients—that provide evidence-based methods to prevent OA based on known risk factors in order to influence policymakers and shift this reactive philosophy. The population must then be sufficiently engaged in activities that will take time to comprehend and complete without necessarily seeing any tangible benefit in the short term. This will require significant education—which is likely to be a major solution—to the extent that the population takes greater responsibility for their own health and demands beneficial, preventative interventions for the disease.
Opportunities remain to improve OA clinical outcomes from early diagnosis through to late-stage disease. Enhanced appreciation of the different phenotypes associated with OA, and how these change over time, will provide the framework required for personalised clinical care.76 Such patient stratification will require the development and use of valuable dry and wet biomarkers to enrich the understanding of genotyping.
We need a better knowledge of how best to ensure long-term joint function and health, specifically how to identify, delay or prevent OA. The effect of PA on joint structure and the implications for structural progression of the disease remain largely undefined. Further research is needed to examine interventions employed postinjury and in early OA. It is possible that specific exercises (irrespective of whether the goal is muscular conditioning or weight loss), if sufficiently maintained, could play a vital role in preventing disease progression. Despite being one of the most efficacious treatment options, the effectiveness of exercise for OA can improve. Specific exercise prescription throughout different phases of OA management is necessary and guidelines should aim to make such distinctions. Research consistently reports that any exercise is beneficial in reducing pain, improve function and quality of life—regardless of age, gender, BMI or radiographic/symptom severity—but the most effective exercise dose (eg, frequency, duration, intensity) remains the subject of debate.10 15 77 A meta-analysis concluded that exercise programmes focusing on one specific aim (eg, improving strength or aerobic capacity) are more effective than programmes that combine multiple disciplines together, but neither resistance nor aerobic exercise is significantly superior to the other.77
Understanding the wider context of OA
Arthritis Research UK’s recently published ‘Musculoskeletal Conditions and Multimorbidity’ highlights the essential need to recognise musculoskeletal health as a health priority for people living with multimorbidity.78 In practice, there is a need for greater interdisciplinary working between clinical specialities that may currently provide individual streams of care. ‘Siloed’ packages (health, welfare, social or employment) can obstruct optimal outcomes for patients with complex inter-related needs.79 Better education of all healthcare professionals is important to enable early recognition of musculoskeletal conditions and ensure appropriate referral.
The prevalence and burden of OA, already a major societal concern, continue to increase apace. We have outlined some key initiatives that could help reverse the trend. Directing greater attention towards modifiable risk factors (obesity and joint injury) could provide huge public health and individual joint specific benefits. A focus on early intervention should reap benefits for both joint structure and symptoms. Optimal management needs to be implemented cognisant of patient preference, efficacy, evidence and safety. Presently there is widespread over-reliance on imaging and use of interventions when good evidence states that they are ineffective or harmful. Reducing evidence–practice gaps through appropriate implementation of research is critical. When we revisit this issue in the near future, we hope there has been progress towards addressing these concerns.
What is already known on this topic?
Osteoarthritis (OA) is a common condition for which there is no cure.
Due to the ageing population and increasing obesity, the prevalence of OA is increasing.
OA can contribute to inactivity with ageing—because of pain and reduced function—ultimately impairing quality of life.
Evidence reporting risk factors for lower limb OA consistently highlights joint injury and obesity as two of the most common attributable causes; action to prevent these in society has had limited effect.
Current guidelines, despite broad agreement on management principles, are poorly adhered to in clinical practice; as a result, the outcome of OA management is suboptimal.
What are the findings?
The focus of OA management should shift to prevention (addressing known risk factors) and early in the disease (when the treatment effects are greatest and the structural and symptomatic changes are potentially reversible).
The persistent and widespread use of interventions supported by little or no evidence should cease as these contribute to unsatisfactory outcomes for patients with OA.
Research is required to provide the best approach for managing chronic conditions—especially those sharing common risk factors (eg, obesity, physical inactivity) for disease onset and progression—to enhance treatment effectiveness and reduce the healthcare burden.
Contributors All authors were involved equally in the writing, amendment and approval.
Funding DJH is supported by an NHMRC Practitioner Fellowship.
Competing interests RDL and MEB are funded by Arthritis Research UK. DJH provides consulting advice on scientific advisory boards for Merck Serono, TissueGene and Flexion.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.