Systematic Review of Educational Interventions for Rheumatoid Arthritis

Educational Approaches of Occupational Therapy in Rheumatology

The education process is defined in the Occupational Therapy Practice Framework: Domain and Process (3rd ed.; American Occupational Therapy Association [AOTA], 2014) as the “imparting of knowledge and information about occupation, health, well-being, and participation that enables the client to acquire helpful behaviors, habits, and routines that may or may not require application at the time of the intervention session” (p. S30). Occupational therapy interventions that are based on educational processes aim to increase patient participation and independence in chronic disease management. Because of their diverse applications and content, these interventions have the potential to produce positive outcomes for people with RA, regardless of disease duration and symptom presentation.

Two systematic reviews and one scoping review have included patient education within the scope of occupational therapy interventions for RA (Hammond, 2004a, 2004b; Steultjens et al., 2002). Steultjens et al. (2002) reported on seven mechanisms for occupational therapy intervention for RA. Among these seven mechanisms, limited evidence was found for the categories of “instruction on joint protection” and “advice/instruction in the use of assistive devices,” and no high-quality studies were found for the categories of “counseling” and “training of skills.” The authors concluded that this finding was in direct opposition with how frequently occupational therapy practitioners use these techniques in practice. Similarly, Hammond (2004b) concluded that evidence for occupational therapy–specific psychological interventions (e.g., coping and relaxation skills) was limited but found emerging potential for the use of cognitive–behavioral interventions targeting people who were motivated for change (Hammond, 2004a). Although these reviews provided a comprehensive assessment of the state of research at the time of their publication, the health care environment has changed significantly to include the wide use of new prescription drugs, such as tumor necrosis factor inhibitors, and the implementation of new diagnostic criteria released by the American College of Rheumatology in 2010 (Aletaha et al., 2010), both of which have made clinical remission a more attainable treatment target. A more updated review is needed in light of these advances to incorporate intervention trials conducted within the past decade.

Method

The data presented in this article are part of a larger study evaluating the full spectrum of occupational therapy interventions for RA (including remediation, adaptive and preventive therapies, and educational approaches) that aimed to address the following question: What evidence exists to support occupational therapy intervention approaches for RA?

A large volume of varied intervention approaches was identified, including didactic, psychosocial, remediative, adaptive, compensatory, and preventive interventions. Because the evidence base for these intervention approaches was so diverse, the systematic review was divided into two parts: educational interventions versus remediation and adaptation interventions (e.g., strengthening, range of motion, and provision of orthoses). In this article, we evaluate the evidentiary base for the former category.

Search Strategy

We searched two electronic databases, MEDLINE and CINAHL, on two dates (the first on November 7, 2013, and an updated search on June 18, 2015) using search terms and strategies listed in Supplemental Appendix 1 (available online at http://otjournal.net; navigate to this article, and click on “Supplemental”). In addition, the first author hand-searched the most frequently used peer-reviewed journals in U.S. occupational therapy academic programs for the keyword “rheumatoid arthritis” (NBCOT, 2013). These journals included the American Journal of Occupational Therapy; British Journal of Occupational Therapy; Australian Journal of Occupational Therapy; Journal of Hand Therapy; Journal of Occupational Science; and OTJR: Occupation, Participation and Health. Titles, abstracts, and full-text PDFs of all eligible articles were imported into a citation manager.

Articles were excluded in five steps, as shown in Figure 1. The first author and the second author (Pyatak) independently reviewed all articles and completed Steps 1 and 2 to eliminate studies that were outside the realm of occupational therapy or that did not target symptoms of RA. The first author independently completed Steps 3–5 to obtain the final articles to be included within the study. To ensure replicability and accuracy, a graduate research assistant independently completed the same steps for 50% of the articles. In case of disagreement, the full-text article was reviewed by the second author, who made the final determination regarding categorization.

OPEN IN VIEWER

Figure 1.

Exclusion steps.

Results

Fifty-seven textbooks were screened for inclusion. Of these, four textbooks were excluded as secondary because of their focus on assessments (n = 2) or use as reference guides (n = 2) rather than intervention, and 38 textbooks did not include the search terms “rheumatoid arthritis,” “arthritis,” or “juvenile arthritis” in their indexes. Descriptors of RA interventions from the remaining 15 textbooks were extracted and synthesized into 12 intervention approaches. These approaches were grouped according to the four categories from the Framework (see Table 1): Education Process, Preparatory Methods, Consultation Process, and Occupation-Based Intervention and Purposeful Activity. Four intervention strategies fell under the “Education Process” of occupational therapy intervention—psychosocial intervention, joint protection and energy conservation, disease education, and pain management—and were included in the current systematic review.

Discussion

This review provides an update of the evidence base for occupational therapy interventions, as previously described in Steultjens et al. (2002) and Hammond (2004a, 2004b) more than a decade ago. There was strong evidence found for the following outcomes: Pain management interventions improved joint pain, coping efficacy, and problem-focused coping; psychosocial education interventions improved affective pain; and combination interventions improved pain coping efficacy, pain catastrophizing, positive affect, disease activity (tender joints), and helplessness. Moderate evidence was found for joint protection and energy conservation interventions improving joint protection behaviors. There was limited or no evidence for other outcomes.

Research on RA interventions within the scope of occupational therapy has increased significantly: from 15 randomized controlled trials (RCTs) studying occupational therapy interventions for RA before 2002 (Steultjens et al., 2002) to 22 RCTs focused on the educational process of occupational therapy published in the past 13 yr. Although this increase demonstrates promising growth of the evidence base for occupational therapy intervention for people with RA, a majority of the constructs assessed continue to remain in Level III or Level IV of evidence. Because search terms were obtained from occupational therapy textbooks, this raises the question of whether there is a disconnect between the continuously changing base of evidence in rheumatology and what is being highlighted in occupational therapy textbooks and educational programs. Therefore, this article, combined with similar evidence-based reviews, may serve to guide areas of future research to support current guidelines of occupational therapy practice.

When interpreting the findings of this review, outcomes falling within Levels III and IV of evidence may indicate that few methodologically sound studies evaluated these constructs or that evidence was conflicting across studies. One explanation for these results stems from the inconsistency of target outcomes between trials. The interventions assessed in this review targeted a wide range of constructs, including disease activity, function and participation, and quality of life. Although interventions may have used similar mechanisms (i.e., cognitive–behavioral approach), their content was targeted toward specific symptoms or behavioral goals (i.e., increasing disease knowledge vs. decreasing fatigue). Therefore, disagreement between trials may not stem from the ineffectiveness of intervention mechanisms but rather may reflect differences in the primary outcomes targeted by the intervention content.

In research in which replication is a goal, it is important to establish a core set of target outcomes to determine effectiveness across trials. To address this issue, attendees at the Outcome Measures in Rheumatology (OMERACT) conference ratified a core set of outcomes in 1994 for use in rheumatology clinical trials. Despite an overall increase of researchers using this set of outcomes within rheumatology trials, their use remains low within nonpharmacological studies. Explanations for the lack of uptake of these core outcomes include researchers’ lack of awareness or consciously choosing against their use because of their poor fit with rehabilitation interventions (Kirkham, Boers, Tugwell, Clarke, & Williamson, 2013). The importance of developing and using a core set of outcomes continues to be discussed, with emphasis shifting to include patient perspectives on what is important to their well-being (e.g., Kirwan et al., 2005).

Another explanation for conflicting evidence is varying patient responsiveness because of the unpredictable course of RA and range of medical regimens prescribed to people with RA. Some people are greatly affected by symptoms that are absent in others or have a more severe presentation during symptom flare-ups. Therefore, people may not show significant improvements if provided with interventions incompatible with their current symptoms and needs. In addition, medical treatment of RA varies widely by type of medication (e.g., nonsteroidal anti-inflammatory drugs, oral corticosteroids, biologics) and dosage, which may affect a person’s response to nonpharmacological programs. Although properly conducted RCTs substantially reduce the potential for confounding by individual variation in symptoms and medication use, identifying people likely to respond to a given intervention is particularly challenging in RA research.

In future studies, researchers may benefit from analyzing the participant cohort for unique factors that contribute to intervention responsiveness. Tailoring interventions to address people’s current symptoms and interest and ability to adopt health-promoting behaviors has been established as a potential strategy to optimize the effectiveness of health behavior interventions while maintaining a patient-centered approach (Prochaska, DiClemente, & Norcross, 1992). By targeting interventions to people experiencing certain clusters of symptoms, researchers may have increased statistical power to discern effectiveness of interventions in specific populations. Current research is already moving in this direction. For example, post hoc analyses by Zautra et al. (2008) demonstrated that mindfulness education was significantly more beneficial for participants with a history of recurrent depression than for other groups. Although in the current review we analyzed the main effects of each intervention, such subgroup analyses have the potential to shed greater insight on which populations benefit most from a given intervention.

Limitations

Limitations are reported at both the level of the included studies and the level of the review. Three methodological issues made it difficult to draw inferences from the included RCTs. First, several studies noted difficulty both recruiting and retaining people because of lack of interest or motivation, hospitalization, changes in medication, distance from hospital or clinic, and severity of illness. These factors warrant concern for selection bias, particularly because of people with a higher burden of illness being less likely to participate. In practice, it is imperative to review not only the needs and interests of the patient population but also the feasibility and cost-effectiveness of providing educational services for a population at high risk for attrition. Second, 6 of the 22 included studies did not designate a primary outcome. This information is necessary to determine the overall hypothesis of the study and whether the study has sufficient statistical power to test this hypothesis.

Finally, many of the studies included in this review were relatively small, and thus an inability to find a statistically significant effect for the intervention may be more a result of lack of statistical power than a lack of intervention efficacy. Lack of statistical power is likely to be a particular problem when secondary analyses are involved. Nevertheless, several small studies showed that their interventions produced statistically significant results and thus support the potential of educational occupational therapy approaches for RA. Even when small studies produce statistically significant results, their effect estimates may be very unstable, and thus we have refrained from discussion of effect size estimates. In addition, small samples are frequently the result of convenience and thus may not be representative of the populations to which we wish to draw inferences.

With respect to the limitations of the systematic review, the methodology included all RCTs within the educational scope of occupational therapy; however, studies were not limited to interventions provided by occupational therapy practitioners. Although all of the interventions being evaluated were within the scope of occupational therapy practice, it is possible that unspecified aspects of the interventions departed from the occupational therapy scope of practice, thus limiting their potential relevance to the field. In addition, the methods to obtain intervention categories through textbooks, and subsequently search terms, may have limited the scope of the intervention categories. Although the selected textbooks were identified as the most used within occupational therapy programs (NBCOT, 2013), there may be nontextbook resources describing occupational therapy treatment approaches for RA that were not captured in this review.

Conclusion

Our review demonstrates that occupational therapy practitioners who use interventions with the educational approach assist people with RA to increase coping skills for symptoms such as pain and fatigue as well as increase positive affect. Given that rheumatology is continuously progressing with medical advancements, occupational therapists and other health professionals should continue to evaluate the medical landscape to develop best practices in RA intervention. In future research, investigators should incorporate methodological changes, such as using a core set of outcomes and targeting interventions toward specific symptoms, to demonstrate effectiveness across studies.

Implications for Occupational Therapy Practice

The findings of the study have the following implications for occupational therapy:

Supplemental Material