Effectiveness of a cognitive-behavioral group intervention for knee osteoarthritis pain: protocol of a randomized controlled trial

Osteoarthritis (OA) is the most common form of arthritis and a major contributor to functional disability [1]. It represents a major social and health problem in the elderly, imposing an increasingly heavy economic burden on social welfare and health care systems. This is due to the need for surgical and medical interventions and frequent absenteeism from work [2]. Although not all OA is symptomatic, the World Health Organization has estimated that OA is the cause of disability in at least 10% of the population over the age of 60 years [3].

Knee OA is the most common type of arthritis. The main symptom of knee OA is pain, which is generally related to joint use and relieved by rest. However, the association between radiological changes in knee OA and the severity of pain or the level of disability is not straightforward. As OA progresses, the pain may become more persistent and also manifest itself at rest. In addition to pain, loss of function and joint stiffness are typical symptoms of OA, which often lead to difficulties in performing daily activities.

The assessment and treatment of pain is vital to the management of OA [4]. It is known that chronic pain is associated with increased considerable psychological distress, such as anxiety and depression. One population study from 17 different countries found that depression and anxiety disorders occurred significantly more often in individuals with self-reported arthritis, with depression present in 5−10% of those with arthritis [5]. In another study conducted by Smith and Zautra, measures of anxiety and depression emerged as independent and significant predictors of current and future pain [6].

The cognitive-behavioral (CB) perspective presented by Turk et al. [7] is the most widely accepted model in the field of pain psychology. It has led to the identification of cognitive and other psychological factors that are associated with pain severity and disability. If one considers the constructs that have the strongest empirical support, factors like pain catastrophizing [8], fear-avoidance [9], self-efficacy and lack of perceived control [10, 11], and passive pain coping [12] have been claimed to be of importance. Factors, such as pain catastrophizing and pain-related fear, have been found to be strongly and consistently associated with pain severity and disability in patients with musculoskeletal pain [13] and knee OA [4]. Low self-efficacy and helplessness have also been identified as predictors of disability in OA patients [14, 15].

According to several reviews, psychological factors influence not only pain and disability, but also in particular the transition from the presence of acute to chronic pain [13, 16]. However, there have been relatively few attempts to prevent chronic disability in OA by adopting a CB approach. Calfas et al. conducted a randomized controlled trial (RCT) testing a 10-week group-based CB program on 40 OA patients [17]. They concluded that in the long term, physical and psychological functioning did not differ between the CB and the education control group. In 2011, Riddle et al. published the results of a quasi-experimental study testing an 8-session individual pain coping skills training (PCST) program in patients with elevated pain catastrophizing who were scheduled for knee arthroplasty [18]. They found that the PCST program resulted in significantly greater reductions in pain severity and catastrophizing, and greater improvements in function when compared with the usual care cohort. Somers et al. recently reported the results of an RCT studying the effects of a PCST program and a behavioral weight management intervention [19]. They concluded that the combination of these two treatments yielded significantly better outcomes in terms of pain, physical disability, stiffness, activity, arthritis self-efficacy, and weight self-efficacy than either of the intervention modalities alone or that evident in the control group.

There are several trials that have used behavioral interventions with similarities to CB principles as well as studies that have integrated CB principles with other forms of rehabilitation in OA patients. For example, Hurley et al. conducted an RCT assessing the effects of a combined exercise, self-management, and active coping strategies rehabilitation program [20]. They found over the long-term, that the rehabilitated participants enjoyed better physical function, lower community-based health care costs, medication costs, and total health and social care costs as well as concluding that there was a high probability (80–100%) that the program was cost effective. On the other hand, Keefe et al., adopted an RCT setting to investigate the separate and combined effects of spouse-assisted pain coping skills training and exercise training in patients having persistent osteoarthritic knee pain [21]. They concluded that the kind of intervention combining spouse-assisted coping skills training and exercise training could improve physical fitness, strength, pain coping, and self-efficacy in patients suffering from pain due to OA.

One CB-based approach in treating pain-related disability and chronicity is the 6-session group intervention model presented by Linton [22]. The model was originally developed for early identification and intervention in the prevention of musculoskeletal pain [23, 24]. The standardized 6-session program focused on coping, function and cognitions and the application of learning principles to allow the individual to utilize more adaptive methods of pain management and active coping [22]. The model has previously been tested in RCT setting in patients with back and neck pain [23‐25]. In these studies, the CB intervention led to less short- and long-term work absenteeism, fewer health care visits, decreased perceived risk and fear-avoidance beliefs, and a larger number of pain-free days.

The present study aims to explore the effectiveness and cost-effectiveness of the CB group intervention described by Linton [22], modified for patients with knee OA. As far as we are aware, there have been no previous studies using this approach in patients with knee OA. The working hypothesis is that patients with symptomatic knee OA will benefit from this kind of CB rehabilitation program. More specifically, we intend to examine the effect of the intervention in terms of self-reported physical function and pain, pain-related work abstinence, the number of pain-related health care visits, and health-related quality of life (HRQoL). We also aim to determine the effect of the intervention on several psychological variables such as depression, anxiety, sense of coherence, pain catastrophizing, kinesiophobia, self-efficacy, and life satisfaction. Finally, we plan to run a cost-utility analysis of the intervention based on quality-adjusted life years (QALY).

There are several reasons why this study will advance our understanding of effective interventions to improve conservative knee OA treatment. First, there have only been a few studies evaluating the effect of psychological interventions on knee OA pain. Thus, good quality RCTs are needed to determine their effectiveness. As present, very few studies have evaluated a CB intervention for knee pain treatment, as far as we are aware this will be the first study to tackle this particular CB intervention model in treating OA patients. Second, the proposed research will address the very current topic of non-pharmacological conservative treatment of knee OA, and the results may have a significant impact on current care recommendations. Since knee OA is the most common form of arthritis, representing a major economic burden on social welfare and healthcare systems for both society and individual patients, it is worthwhile to find novel cost-effective interventions promoting conservative treatment. Third, it is important to evaluate this kind of CB intervention since cognitive and psychological factors are known to play major roles in chronic pain, pain severity and disability. Moreover, evidence-based guidelines indicate that one should adopt a the combination of pharmacologic and nonpharmacologic modalities to achieve adequate management of knee OA [64‐66]. With respect to nonpharmacologic treatment modalities, it has been occasionally recommended that patients should participate in self-management programs with psychosocial interventions [64]. We hope that the results of the cost-effectiveness and cost-utility analysis of the proposed study will favour their inclusion into the current care recommendations in the future.

The study protocol has some limitations. First, due to volunteer bias, the results of psychological intervention studies may have limited generalizability [67]. Second, as the majority of the study patients will probably originate from primary care, the results of this study will mainly apply to that environment. Third, the CB protocol presented by Linton [22] indicates that patients participating in a CB-type intervention should be evaluated to ensure that they are at risk of developing a persistent pain problem. Linton himself has developed a questionnaire, the Örebro Musculoskeletal Pain Questionnaire (MPQ) [59], to be used for this purpose. In our study, however, we have decided not use this questionnaire or any other cognitive or psychological evaluation as part of the inclusion criteria. One reason for this decision is that the Örebro MPQ has only been validated in back pain patients, and there is a lack of data concerning the similarity of pain experiences in back and knee OA pain in general. Furthermore, our inclusion criteria already include several pain measures, assessing both intensity and duration of pain. We consider that by applying to these criteria, in all probability, the patients would be at risk of developing a persistent pain problem if they were not already suffering from it. However, the Örebro MPQ will be included in the baseline patient questionnaire for analysis purposes. With the availability of the final data of the study, it will be interesting to see whether those patients determined to be at risk of developing a persistent pain problem according to the Örebro MPQ have benefited more from this intervention.

Another limitation of the study might be the different amount of attention paid to the two groups. As the group of patients in the intervention will receive 12 hours more attention, a Hawthorne effect [68‐70] is probable. However, all patients will receive the same basic instructions about knee OA treatment according to Finnish CC guideline [26] when they are listening to a lecture delivered by a general practitioner prior to the randomization. This lecture will be held in groups of about 20 people and the whole research group will be present to answer questions. During the lecture the patient-version booklet of Finnish CC guideline on knee and hip OA [26] will be handed out to all participants. The participants in both groups also will continue to receive standard care (i.e. normal routine care offered by their own general practitioners including analgesics and physiotherapy).

BDI-21, BAI, 13-item SOC scale, TSK, PSEQ and LS questionnaires have been used in several clinical studies in Finland [35‐37, 39, 44‐46, 50, 52‐55]. However, at present BDI-21 and 13-item SOC scales have been validated in Finland [37, 46] and the validation process of TSK is almost complete. We have available translated versions of separate questionnaires that have been in use in clinical studies and clinical routine in Finland. As the psychometric questionnaires are all secondary outcome measures in our study we decided to include them even though there is a shortage of supporting evidence for the validity of some of their psychometric measurement properties in Finnish patients.

Obtaining and assessing KL grades are not simple tasks. Valid measures are dependent on knee flexed standing views [71, 72] and examiner experience [73, 74]. Although it has recognized weaknesses, the KL score has been the most widely used radiographic score to assess knee OA [73]. We decided to use KL grade ≥2, which has been extensively used as the inclusion criterion in knee OA studies [33]. Unfortunately, we will not be able to control the radiographic quality since the knee X-rays will have been taken in public primary care locations and in the outpatient clinics. However, we use the combined radiographic and clinical criteria that have been proposed for use when diagnosing knee OA. If one combines the clinical and radiographic factors, the sensitivity and specificity of knee OA diagnosis are 91% and 86%, respectively [75]. The observers also will be well trained and acquainted with the radiographic atlas of the KL scores before the start any assessments. Although reliability for the separate KL radiographic features may depend on the level of the investigator’s experience, the intra- and inter-observer reliability for the overall score is known to be relatively high [74].

No significant side effects or adverse events are expected.