Dr. Hochberg has received consultant fees (less than $10,000 each) from Abbott Laboratories, Amgen, AstraZeneca Pharmaceutical Co., Bayer Health Care LLC, BioIberica SA, Bristol-Myers Squibb Company, CombinatoRx, Covidien, Eli Lilly and Company, Endo Pharmaceuticals, GlaxoSmithKline, Hoffman-La Roche, Inc., Merck & Co., Inc., Merck Serono International SA, NiCox SA, Novartis Pharma AG, Pfizer, Inc., Pozen Inc., Rand Corporation, Roche Laboratories Inc., Sanofi Aventis SA, Smith & Nephew, Stryker Biotech, UCB, Inc., and Wyeth Pharmaceuticals, Inc.
Disability due to osteoarthritis (OA) and increasing disease prevalence are colliding to create a major public health problem. By 2020, the number of people with OA will have doubled, due in large part to the exploding prevalence of obesity and the graying of the “baby boomer” generation (1). Despite growing concern, there remain few safe and effective interventions. Although practice patterns may vary, current clinical management for OA is often limited to the use of analgesic and/or antiinflammatory medication and cautious waiting (2) for the eventual referral for total joint replacement. Recent doubts about the safety of several commonly prescribed OA medications have served to highlight deficiencies in the traditional medical approach to management (3). With few conservative options offered by their doctors, increasing numbers of patients are turning to untested folk remedies and aggressively marketed dietary supplements with little substantive evidence to support their efficacy (4).
It is in this context that we appraise the current quality of clinical management of OA. It is critical to assess the consistency, or lack thereof, between evidence-based recommendations and actual clinical practice. A number of recent studies have highlighted that there is a great divergence from such recommendations despite numerous efforts to disseminate the many recommendations that exist. Concerns over the quality of health care delivery raise serious questions about the legitimacy of prior efforts not only to disseminate, but more importantly ensure implementation of, what is best practice. Some cautious steps to encourage implementation, including the development of quality indicators, have occurred, but have made little impact. Quality indicators are elements of health care for which evidence or consensus exists that they are indicative of the quality of health care (5, 6). A number of quality indicators have been developed in OA that thus far have not been widely used or systematic in considering the breadth of OA management. In this era of health care reform, it is critical that appropriate efforts toward improved delivery of quality care in this burdensome disease are made. What follows is a narrative review of how standard clinical practice diverges from evidence-based recommendations, some key challenges facing clinicians with regard to optimizing the quality of care delivery in OA, and steps to improve OA health care quality. This is written as a potential roadmap for those developing and enforcing policy decisions, as well as for clinicians who could be provided further encouragement to enact practice change.
The public health burden of OA
Recent estimates suggest that symptomatic knee OA occurs in 13% of persons ages ≥60 years (7–9). The risk of mobility disability (defined as needing help walking or climbing stairs) attributable to knee OA alone is greater than that due to any other medical condition in people ages ≥65 years (10, 11).
In addition to its societal impact, the burden of OA can be measured by its impact on quality of life. OA was estimated to be the eighth-leading cause of nonfatal burden in the world in 1990, accounting for 2.8% of total years lost due to disease, approximately the same percentage as schizophrenia and congenital anomalies (12). In the version 2 estimates for the Global Burden of Disease 2000 study, published in the World Health Report 2002 (13), OA is the fourth-leading cause of years lost due to disease at the global level.
The US spends more than $2.2 trillion nationally on health care each year, equal to 16% of the gross domestic product (GDP). Health care spending is projected to increase to $4 trillion by 2017, or almost 20% of the economy. This continued growth is not sustainable and has been responsible in part for recent calls to health care reform, which should include cost containment through optimizing effectiveness and quality.
According to the US Centers for Disease Control and Prevention, arthritis and other rheumatic conditions cost the US $128 billion in 2003, a 24% surge since 1997 and an amount equal to 1.2% of the GDP (14). The most prevalent and disabling form of arthritis is OA. Given the aging of the population, increasing incidence of OA, and large volume of surgical procedures performed for OA, such as total joint arthroplasty and arthroscopy, OA contributes a substantial burden to the cost of arthritis care. Management of OA and potential health care quality reforms are the focus of this review.
Plethora of recommendations
Evidence-based recommendations have the potential to improve the quality of health care by promoting interventions of proven benefit and discouraging unnecessary, ineffective, or harmful interventions (15). Numerous recommendations for OA management have been developed in recent years by a number of scientific societies and health care organizations (16–20). Confusion could potentially arise from this multitude of sometimes inconsistent guidelines, although some amount of replication is necessary to facilitate appropriate contextualization of recommendations for specific users, such as clinicians, countries, or health care organizations. Fortunately, however, there is general consistency, despite some differences in quality, in the numerous available OA recommendations (15, 21–23).
Clinical practice not reflecting evidence-based recommendations
Despite remarkable consistency between recommendations and in spite of some dissemination attempts, clinical practice does not reflect these recommendations (2, 23–25). Current clinical practice reflects a multitude of factors, including clinician and patient preference and health care system support. In the absence of a cure, current therapeutic modalities are primarily aimed at reducing pain and improving joint function using agents mainly targeted toward symptoms that do not facilitate improvement in joint structure or long-term disease amelioration. Frequently, clinicians do not adequately recommend conservative nonpharmacologic management modalities (24), leading to unnecessary ordering of imaging and inappropriate referral to orthopedic surgeons (2). This divergence from evidence-based care is also commonly seen with other diseases, where it is estimated that half of the treatment rendered by clinicians is inappropriate (26).
Conservative nonpharmacologic management.
The majority of persons with OA are either overweight or obese. There is good evidence for the efficacy of weight management (27), and this is advocated by most OA recommendations. However, in practice, weight management is not frequently implemented (24, 28, 29). Another pivotal and frequently ignored (24, 28, 29) aspect of conservative treatment of OA is exercise. Recommendations routinely advocate exercise (16–20, 30, 31); however, clinical practice does not reflect this recommendation (24, 28, 29). Of course, there are often difficulties in the implementation of such recommendations related to patient motivation and adherence and the inability of clinicians to effectively provide such interventions within their practices, and more broadly, within the current health care system.
Surgery should be resisted when symptoms can be managed by other treatment modalities. At present, there is no metric whereby the use of conservative management prior to surgery is monitored. The typical indications for surgery are debilitating pain and major limitation of functions such as walking, working, or sleeping (32–34). Although joint replacement has a large effect size (35–37), this is an invasive treatment that has attendant risks. Surgical funding and resources are rarely linked to optimizing outcomes and reducing risk. In addition, several studies have suggested that up to 30% of some surgical procedures are inappropriate (38, 39). Eliminating inappropriate surgery has the potential to contain rising costs.
The most common orthopedic surgery performed on the knee has no demonstrable efficacy (40, 41). Arthroscopic debridement and meniscal resection remain the most frequently performed procedures by orthopedic surgeons in the US (42, 43), with up to 1 million knee arthroscopies performed annually in the US. Recent recommendations (including those from Osteoarthritis Research Society International [OARSI] and the American Academy of Orthopedic Surgeons [AAOS] [19, 20]) recommend against the use of routine arthroscopy for knee OA management. To date, these recommendations have not been reflected in changes in clinical practice.
Given the fact that OA is primarily a clinical diagnosis made on the basis of history and physical examination, imaging should be reserved for instances in which the diagnosis is unclear and radiography can refute other diagnoses that could plausibly result in the patient's symptoms (44). Currently, there is an overuse of inappropriate imaging to make a diagnosis that can be made clinically. In light of the current lack of therapy that can modify the disease course and measurement imprecision, there is currently no rationale for obtaining serial radiographs if the clinical state remains unchanged. Magnetic resonance imaging (MRI) should only be used in infrequent circumstances to facilitate the diagnosis of other causes of joint pain that can be confused with OA (e.g., osteochondritis dissecans, avascular necrosis).
Eliminating unproven procedures and reducing needless costs is necessary if the nation is to improve the quality and lower the cost of care overall. The use of computed tomography, MRI, and positron emission tomography scans has grown to a $100 billion industry in the US, of which $14 billion has been shouldered by Medicare. Recent estimates suggest that of the 95 million high-tech scans done in the US annually, 20–50% are unnecessary because their results failed to help treat or diagnose the patient's ailment (45).
Authors reporting divergence of OA clinical practice from management guidelines have advocated improved disease management education during medical school, and have called for reminder systems to promote physician adherence (2, 23–25). These steps could potentially be helpful, but further systematic development, implementation, and audit of quality measures are more likely to lead to change in practice.
From guideline recommendations to implementation
The discrepancy in OA care between what is recommended in guidelines and what happens in clinical practice suggests that there are issues with dissemination, implementation, or both (46). Given the number of guidelines available for OA and the consistency of recommendations within them, and considering the time and resources required for guideline development, future efforts to guide the management of OA of the hip and/or knee are better directed toward implementing practices known to be effective, and facilitating research to answer important questions where there is little evidence (15). There are a number of methods that can be used to facilitate guideline implementation, reduce unexplained variation in practice patterns, and optimize care, including development of quality indicators. A number of groups have developed and tested quality indicators in OA (47). To date, methods used to develop the quality indicators have relied more upon expert consensus than available evidence. At best, small validation studies have been done, and the indicators have not typically addressed the breadth of management concerns in OA. As a consequence, very little system-wide change has occurred. Here we will briefly summarize such prior efforts and then present a vision for further development of quality indicators.
Arthritis Foundation (AF) quality indicator project.
The AF quality indicator project was developed to establish a set of measures that could be used to assess the quality of care for patients with OA, as well as rheumatoid arthritis (48). The AF quality indicator set for OA includes 14 indicators that were limited to aspects of physical examination, assessment of pain and function, nonpharmacologic modalities (including patient education, exercise, weight loss, and the use of assistive devices), the use of acetaminophen, surgical evaluation, and the role of radiographs (49).
Assessing Care of Vulnerable Elders (ACOVE).
The ACOVE project has addressed quality indicators for OA (50, 51). A total of 11 indicators were developed that were focused on the assessment of pain and function, the exclusion of superimposed joint infection in patients with an acute synovitis, the prescription of nonpharmacologic modalities and acetaminophen for patient management, and referral to orthopedic surgeons. Ganz and colleagues reported the results of an ancillary study of a sample of community-dwelling elders ages ≥75 years who reported a physician or nurse diagnosis of OA (52). The authors concluded that the quality of care for community-dwelling vulnerable elders with OA was suboptimal, with the lowest pass rates for indicators reflecting medication safety.
British primary care.
This study used 9 evidence-based indicators to measure the quality of recorded primary care for OA (53). Included indicators covered the provision of information; regular assessment of pain, function, and side effects; and the provision of treatment. A total of 320 patient records from 18 practices were examined. There was substantial variation in the recorded achievement of individual indicators (range 5–90%), and in addition, they found considerable scope for improvement in the recording of high-quality care. The lack of variation between practices suggested that system-level initiatives may be needed to achieve improvement. In a prior study, the investigators demonstrated that the introduction of financial incentives was associated with substantial apparent quality improvement for incentivized conditions. For nonincentivized conditions, the quality did not appear to improve (54).
Additional quality measure development initiatives (55) have considered OA as part of broader comorbidities, including the Physician Quality Reporting Initiative (56). The lack of appropriately developed and validated quality indicators has caused vigorous debate, as indicators may be enforced by payors despite poor evidence and a lack of improvement in patient care.
Toward improved health care quality in OA
There is worldwide interest in developing and implementing patient-centered models of care to support integration of evidence into practice and improve patient health outcomes for people with chronic conditions, including OA (57, 58). The complex factors driving these health service reforms are well summarized in a report from the Institute of Medicine (59). The Institute of Medicine report provides 10 rules for system redesign to meet the needs of those with chronic conditions, one of which relates to evidence-based decision making. Care should not vary illogically from clinician to clinician or from place to place.
OA management system redesign should consider all quality of care domains: that care is safe, effective, patient centered, timely, efficient, and equitable (60). The historical approach to the uptake of evidence into practice has been to focus on passive dissemination (diffusion) of information to individuals through social networks in which an assumption is made that the target individuals are aware, motivated, and capable of implementing necessary change. Support for this process has relied on the provision of resources in the form of standardized evidence-based clinical practice guidelines and provision of skills training in the use of evidence-based medicine methods. Educational methods have involved the use of passive continuous medical education methods.
Despite widespread support for OA guidelines, there are delays in uptake, particularly of nonpharmacologic recommendations, and variance in the application of recommendations by clinicians in different contexts (24, 25, 61–63). In addition, qualitative information suggests that the needs of patients are not being met with regard to the quantity and quality of information provided about OA and its treatment (64, 65).
Three main issues that influence the use of evidence in medical care are quality of evidence, barriers and facilitators to practice change, and effective dissemination and implementation strategies (66). Barriers to changing practice exist within the patient, professional, health care team, health care organization, and wider environment. Although lack of awareness of the existence of guidelines is often considered an important barrier to implementation, it is by no means the only factor limiting implementation (67). Pencharz et al evaluated 6 lower extremity OA guidelines, including the European League Against Rheumatism (EULAR) and American College of Rheumatology (ACR) guidelines (18), and found that almost none of them addressed implementation strategies and barriers to use of the guidelines (23). In general, very few guidelines are evaluated for effectiveness of dissemination strategies, and even fewer for the health impact of the guideline (68).
Future research should be directed toward overcoming both patient and provider barriers to the use of OA guidelines and their documentation, guideline implementation, and evaluation of outcomes. Although the evidence for many OA treatments is good, the complexity and high number of treatment recommendations available for OA may be a hindrance to the use of the guidelines. Many strategies have been evaluated to improve implementation of evidence into practice.
Grimshaw et al, in their systematic review of strategies, found that there was no “magic bullet” to promote implementation; however, generally effective strategies include financial interventions for clinicians (compensation hitched to quality metrics), reminder systems, interactive small group meetings, and computerized decision support or a combination of some of these interventions (66, 69–72). In one study, reminders had the largest average effect and were especially influential for prevention (70). Some examples of reminders include provider prompts (e-mail and reminder cards), computer-assisted treatment plans, and patient prompts (telephone and e-mail reminders). Interventions targeted at specific barriers and combined interventions appear to be more effective (66, 69). More education does not appear to be necessary, as clinicians are aware of the available treatments; however, all of the therapeutic modalities were not consistently used.
Guideline implementation, quality indicators, and consumer-driven health care models
Implementation of clinical recommendations is influenced by several factors, including the performance gap between theory and practice (66). Upon endorsement of recommendations by the sponsoring body, the process of implementation starts with dissemination of recommendations (71), encouragement of implementation of recommendations, and the monitoring and evaluation of their impact (73, 74). For too long we have relied on passive diffusion of information that has largely been ineffective.
Quality indicators (evidence-based performance measures) can be developed from evidence-based clinical recommendations. They are the next logical step in the clinical quality improvement cycle after guideline development (75). Quality indicators are usually categorized as structures, processes, and outcomes (5, 6). Structure and process indicators are based on recommendations for practice and intervention, whereas outcome indicators represent the results of care. Insight on adherence to clinical recommendations using indicators is important to identify barriers and provide information for further implementation of the guideline (76). In order to develop appropriate quality indicators for OA, a number of preparatory steps are required (Figure 1) that are described in further detail here.
Appraisal of existing guidelines.
Clinical practice guidelines in the management of OA have been elaborated to summarize evidence related to the management of this health problem and to facilitate uptake of evidence-based knowledge by clinicians. There has, however, been increased scrutiny of the quality of recommendations in recent years. This emphasis is in part related to the relatively recent work of the Appraisal of Guidelines Research and Evaluation (AGREE) collaboration, an “international collaboration of researchers and policy makers working together to improve the quality and effectiveness of clinical practice recommendations by establishing a shared framework for their development, reporting and assessment” (77). The AGREE instrument is composed of 23 items organized into 6 domains: scope/purpose, stakeholder involvement, rigor of development, clarity/presentation, applicability, and editorial independence. A review of the quality of knee OA recommendations using the AGREE instrument was published in 2002, concluding that the quality of the guidelines varied and could generally be improved (23). In 2007, an updated application of AGREE to knee OA guidelines concluded that the clinical management recommendations tended to be similar among guidelines, although the interventions addressed varied (22).
Several guidelines have been published or updated since then, including those from EULAR (17), OARSI (19, 21), ACR (78), and AAOS (20), following advancements in knowledge regarding the management of this condition, particularly as it relates to the use of nonsteroidal antiinflammatory drugs, opioid analgesics, and surgical intervention. In addition, prior appraisals have focused on the knee and not included the hip. Although imperfect, the AGREE instrument allows for the assessment of commonality and limitations of guidelines. Adaptation of evidence-based guidelines should be systematic and transparent to indicators for quality care, should involve stakeholders, and should report the key factors that influence decisions, including those in the original evidence-based guidelines, and the reasons for any modifications that are made (15).
GuideLine Implementability Appraisal (GLIA).
Guideline implementation involves “the concrete activities and interventions undertaken to turn policies into desired results.” Implementability refers to a set of characteristics that predict the ease of (and obstacles to) guideline implementation (79). Shiffman and colleagues developed and validated a tool to help in guideline implementation: “The GuideLine Implementability Appraisal” (79). The GLIA facilitates the selection of implementable recommendations and devises implementation strategies that address identified barriers. It is critical to consider the context where these would be rolled out (80).
Develop and appraise methods of implementation focusing on quality indicators and pay for performance (P4P).
In order to determine, monitor, and improve quality in health care, there is a need for quality indicators. Evidence-based recommendations are particularly suited to the development of indicators. The desirable attributes of performance measures were distilled by the Agency for Healthcare Research and Quality in 2005 (81).
Beginning in July 2007, the Centers for Medicare and Medicaid Services (CMS) initiated the voluntary Physician Quality Reporting Initiative (56). The objectives of the initiative are to help physicians obtain information to improve the quality of care and avoid unnecessary costs. It is also intended to change physician behavior, promote patient-centered care, and increase the cost-effectiveness of the delivery of health care services for payers, providers, and consumers. The Physician Quality Reporting Initiative measures were published in the Federal Register as “The Final Rule” on November 27, 2007 (56). The CMS offered a maximum 1.5% lump sum bonus payment for reporting at least 3 quality measures. At this level of reimbursement and with the program only being voluntary, it will be surprising if this has a major impact on health care quality. A recent survey showed that reimbursement is strongly tied to reported likelihood of adherence (82). Financial incentives appear to be stronger drivers of adherence to clinical practice recommendations than attitudes toward them or their general accessibility. A recent physician survey inquired whether 3 different bonus levels would increase the likelihood of adherence, with a clinical guideline for Medicare patients that required an additional 10 minutes of work per patient on the part of the physician or his/her staff. At the lowest bonus level (2% of Medicare reimbursement), 1 of 20 physicians would be somewhat or much more likely to adhere to recommendations. At 9%, 1 of 6 physicians would be more likely to adhere. At the highest bonus level (20% of Medicare reimbursement bonus), 1 of 2 reported that they would be more likely to comply.
One possible method of encouraging adoption of recommendations, indicators, and monitoring of this performance is through P4P (83, 84). P4P represents an approach to improvement of a clinical care system through financial incentives. These incentives reward clinicians for efficiency, quality, and safety in patient care. In particular, quality measures may be tied into patient satisfaction, processes of care (i.e., compliance with clinical recommendations), outcomes (such as lower mortality after surgery), and structural benchmarks such as technology investments. A P4P plan requires: 1) consensus agreement on minimum standards of care, 2) a method to consistently remind doctors or clinicians to adhere to these standards, 3) methods of gauging adherence to those standards, and 4) a way to document and communicate performance to monitoring agencies such as the CMS.
Economic incentives can impact overall physician compensation or reimbursement incentives for particular procedures. These incentives include capitation, salary, fee for service, and target payments. Regulatory bodies have previously demonstrated the capacity to affect adoption of clinical practice recommendations by clinicians (85). Demonstration projects have met with success in a nationwide P4P program involving orthopedic surgeons on hip and knee replacement (86). High surgical volume centers and teaching hospitals may be better prepared for these types of initiatives. It is critical that we do not adopt a system where reimbursement is a punitive system of withholding reimbursement if certain quality indicators are not met, where many current quality indicators are not evidence based, nor are they likely to produce a positive change in a patient's health state.
Quality measures can be used for improvement, accountability, and research. By improving the model of service delivery, clinicians would be supported in their implementation of pharmacologic and nonpharmacologic recommendations for the care of hip and knee OA, patients would be supported in the self-management of their condition, and integration of acute care and community care services would be improved.
Recent efforts by the CMS and private third-party payers to advance P4P and related quality improvement initiatives have given evidence-based practice-increased prominence. In part, this is because these initiatives have developed to the point where some reimbursement is already based on whether a physician conforms to evidence-based performance measures. It is important to recognize the limitations of current performance measures, as some of the quality indicators currently enforced are not entirely evidence based, but rather simple survey measures of current practice. Even if the P4P movement disappears, it seems likely that physicians will be reimbursed on the quality of the care they provide, and this may also involve measuring patient-based outcomes.
Accomplishing quality improvement will require fostering the growth of more organized systems of care and implementing fundamental payment reform. Policymakers would need to remove legal barriers to collaboration and offer incentives, such as larger payment updates or subsidies for implementing electronic health records, to providers who were willing to establish real or virtual accountable care systems (87). These volume-based payment systems could then be changed to incorporate partial capitation, bundled payments, or shared savings, thereby fostering accountability for overall costs and quality of care. There need to be financial rewards for collaboration, coordination, or conservative practice. These will be directed to ensuring that the barriers to implementation can be overcome, that appropriate investment and support of information technology innovations to advance clinical decision support occur, and that ultimately comparative data sharing to facilitate transparency occurs.
There are numerous evidence-based guidelines available to the practicing clinicians that have the potential to improve the quality of health care by promoting interventions of proven benefit and discouraging unnecessary, ineffective, or harmful interventions. Despite the presence of numerous consistent OA management guidelines and some dissemination attempts, clinical practice does not reflect these recommendations. Future efforts to guide the management of OA are better directed toward implementing practices known to be effective in a context-dependent manner to optimize health care quality. With the convergence of the increase in medical need, progress in information technology, and unsustainable health care costs, there is a favorable set of situational conditions for the funding, development, use, and publication of this new vision for the linkage of the patient-centered and provider-integrated model for OA management. In efforts to improve quality and reduce cost, quality measures will be integral. Quality measurement, provider accountability, and reporting are challenging to do in clinical practice, yet this will be expected of us going forward and ultimately will become a standard and important part of clinical practice. Clinicians and patients need to be involved in the development of quality measures lest this responsibility be assumed by payers in isolation of substantive clinical input.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Hunter had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Hunter.
Acquisition of data. Hunter.
Analysis and interpretation of data. Hunter, Neogi, Hochberg.