 |
|
|
 |
Outcome Measures for Arthritis Care Research: Recommendations from the CARE III Conference
CATHERINE L. BACKMAN ABSTRACT. The Outcome Measures workshop at CARE III addressed the topic of selecting appropriate and relevant outcome measures for research of nonpharmacological, nonsurgical interventions. Consumers, practitioners, and researchers contributed to small and large group discussions following a series of 7 overview presentations. The importance of measuring what matters to consumers using measures that match the purpose of research studies was emphasized. It was generally agreed that care researchers have access to effective measures of disease status, physical and functional status, and quality of life. Gaps exist in the repertoire of existing measures, including satisfactory ways to measure engagement in more complex roles, the area that has come to be known as participation. It was proposed that client-centered outcome measures addressed this gap, but only in part. Research priorities proposed that new or improved outcome measures be designed for participation in employment, parenting, social relationships, leisure, and school, and that more attention be given to process measures that capture the mediating factors associated with improved health outcomes. (J Rheumatol 2006;33:1908–11) Key Indexing Terms:
OUTCOME MEASURES
From the School of Rehabilitation Sciences, The University of British Columbia, and Arthritis Research Centre of Canada, Vancouver, British Columbia, Canada. C.L. Backman, PhD, OT(C), Associate Professor, School of Rehabilitation Sciences, University of British Columbia, Research Scientist, Arthritis Research Centre of Canada. Address reprint requests to C. Backman, School of Rehabilitation Sciences, The University of British Columbia, T325 – 2211 Wesbrook Mall, Vancouver, BC V6T 2B5. E-mail: backman@interchange.ubc.ca
The third international conference on arthritis care research, CARE III, was held in Toronto, Canada, May 12-14, 2005. For the purpose of deliberations at CARE III, care research was defined as methods to address nonpharmacological, nonsurgical care for persons with arthritis1. The objectives of the outcome measures session at CARE III were (1) to summarize current knowledge and emerging approaches to measuring outcomes of arthritis care research; (2) to identify relevant domains to be measured and gaps between these domains and existing measures; and (3) to recommend an agenda for advancing the state of the art in outcome measures relevant to care research. PROCEDURE In the months prior to the conference, one-third of CARE III delegates were assigned to the outcome measures electronic discussion group, which generated and ranked potential topics (Table 1). The top 6 determined the content of overview presentations, intended to address the first workshop objective, as well as stimulate discussion among delegates. Presentations were clustered into 2 panels: (1) measuring relevant outcomes and (2) important measurement issues. Each panel was followed by round-table discussions (see discussion questions below) and general observations shared in a final large-group discussion aimed at achieving the second and third objectives.
Discussion questions 1. What outcomes are of greatest interest to you? Why? 2. Are there good instruments for measuring those outcomes? 3. Is the arthritis research community adequately involving consumers in the development and/or selection of outcome measures? 4. Did the presentations and/or your round-table discussion stimulate ideas for future research priorities regarding outcome measures best suited to arthritis care research? 5. Is a core data set for care research desirable? If yes, should it be any different than the outcome measures used in drug trials? 6. What domains or areas should be included in an international core data set for care research? 7. Can the CARE initiative play a role to facilitate consensus and standardization in this area? 8. What was the most important message you heard this afternoon about outcome measures in arthritis care research? GENERAL ISSUES IN OUTCOME MEASURES FOR CARE RESEARCH Ten years prior to CARE III, Lynch and Caughey summarized benefits of the team approach to managing chronic arthritis2. In citing selected studies of team care, they noted that the outcome measures chosen included traditional clinical and laboratory measures of disease status, as well as measures of pain, function, life satisfaction, depression, and psychological adjustment. They proposed the addition of patient satisfaction measures to this set of outcome measures. What progress have we made in identifying relevant outcomes and appropriate measures of those outcomes, a decade later? In her talk, "Measuring What Matters — The Consumer Perspective," Anne Fouillard spoke about living with osteoarthritis, navigating through the healthcare system, and dealing with the impact of arthritis on herself — her physical, economic, and social well-being — and on her family3. She posited that current outcome measures tend to focus on clinical outcomes immediately relevant to clinicians and scientists. However, from the patient's perspective many important issues were inadequately addressed, including fatigue, intimacy, sexual activity, relationships with others, self-employment, coping, communication, and self-management skills. While some of these may be difficult to measure and may vary in importance over one's lifespan as well as across individuals, they are nevertheless important to consumers' quality of life and their access to appropriate healthcare. Consumer input needs to be integrated in research models during the earliest stages of planning studies, to ensure that research questions and outcome measures address what matters most to consumers. Given the pivotal role of general/family practitioners in helping consumers access health services, it was also suggested that they be more integrally involved in care research. It was also noted that many people with arthritis are not as effective in accessing healthcare as those consumers who are actively engaged as research partners, and future research priorities should address the unmet needs of this potentially underserved group of people. Ways to best capture outcomes of relevance to consumers have been explored in presentations at the 3 international CARE meetings and other venues such as the patient perspective discussions at OMERACT. A progress report from OMERACT 7 suggests that in trials of rheumatoid arthritis interventions increasing attention be given to measuring fatigue4, and recent work has attempted to better document the impact of fatigue from the patient's perspective5. Measuring what matters to consumers remains a priority for future arthritis care research. Another way of ensuring patient perspective is by using client-centered (also known as patient-centered) outcome measures. Such tools consider individual patients' preferences or goals for intervention, and measure whether these goals are achieved. Typical goals tend to involve participating in activities or fulfilling a life role of importance to the individual person. Several CARE III delegates emphasized the need for measures of individual experiences that go beyond the traditional, standardized approaches to measuring functional or health status. Existing tools inadequately assess participation (i.e., complex tasks and involvement in life roles), as defined in the International Classification of Functioning, Disability and Health (ICF)6. In discussion of the concept of participation, some delegates pointed to a need to clarify how participation differs conceptually from quality of life; more specifically, it was contended that participation requires outcome measures that are different from established quality of life tools. Other delegates expressed concern that applying participation measures for comparison across groups of patients in research studies or for generalization beyond the individual experience might not be a valid approach. It was also noted that Stamm and colleagues7 have embarked on important work linking concepts from the ICF to existing outcome measures, including client-centered measures that may address at least some aspects of activity (and activity limitations) and participation (and participation restrictions). Alongside debate about appropriate and relevant outcome measures to determine effects of nonpharmacological care, there was discussion about the need to use process measures. Many rehabilitation and psychosocial interventions involve building knowledge and skills and changing behavior. Our interventions are often delivered by an interdisciplinary team, or any one of a number of models of care. In order to draw conclusions about outcomes and attribute those outcomes to the intervention, it is necessary to evaluate the process of care. For example, if it is hypothesized that better health outcomes will be achieved when a consumer is well informed and empowered, then it is necessary to measure specific attributes such as communication and coping skills as part of the causal pathway to better outcomes. To this end, current work on a tool to measure the effectiveness of consumers in managing their healthcare was presented8. An effective consumer was defined as one who is effective at using the healthcare system, managing their illness, and interacting in the social environment. The impetus for this work arises from the growing interest in self-management programs that aim to empower patients to take an active role in their healthcare, ensuring that services are centered on the patient and address the patient's needs. The skills identified as important to being an effective healthcare consumer include information seeking, decision making, negotiation, and interacting within the social environment. The workshop next examined the current status of instrument development and psychometric principles. Areas that need greater attention were identified: item response theory, computer adaptive testing, and differential item function hold promise for improving approaches to measuring outcomes of interest. (Item response theory, as opposed to classical test theory, uses a mathematical model for test development that accounts for the difficulty of test items and their ability to differentiate individuals with different levels of the trait being tested.) Better understanding of the concept of responsiveness, or detecting a meaningful change in outcome measures, was also discussed. Given this background of classical and contemporary approaches to measuring outcomes, participants generated a list of recommendations for further consideration, of which highlights are presented here (Table 2).
CONCLUSION Delegates had high expectations for better outcome measures that would adequately capture aspects of life most important to people living with arthritis, and people in general. While not always achievable, high expectations can guide the way to improving quality of care and quality of life. Points raised during outcome measure sessions at CARE III set the stage for further consensus development to establish a core set of relevant outcomes to measure effects of nonpharmacological interventions. A model for the identification of core and supplementary domains is provided by others, such as the IMMPACT recommendations for chronic pain clinical trials10. ACKNOWLEDGMENT The author wishes to thank delegates, presenters, moderators, and facilitators of the Outcome Measures session at CARE III. Delegates to the outcome measures consensus discussions collectively contributed the ideas presented in this article. Participation in the discussions was stimulated by short overview presentations from Anne Fouillard, Nancy Santesso, Tanja Stamm, Ted Pincus, Ingemar Petersson, Erik Taal, and Dorcas Beaton. Christina Stenström and Alan Tennant moderated panel discussions; Helena Axler and Dorcas Beaton facilitated group discussions.
1. Li LC, Backman C, Bombardier C, et al. Focusing on care research: A challenge and an opportunity [editorial]. Arthritis Rheum 2004;51:874-6. [MEDLINE] 2. Lynch NM, Caughey DE. Team management of chronic arthritis [editorial]. J Rheumatol 1995;22:1005-6. [MEDLINE] 3. Fouillard A. Measuring what matters: The consumer perspective. In: Proceedings of CARE III, The Third International Conference on Care in Rheumatic Diseases, Toronto ON; May 12-14, 2005:59-61. 4. Kirwan JR, Hewlett SE, Heiberg T, et al. Incorporating the patient perspective into outcome assessment in rheumatoid arthritis — Progress at OMERACT 7. J Rheumatol 2005;32:2250-6. [MEDLINE] 5. Hewlett S, Cockshott Z, Byron M, et al. Patients' perceptions of fatigue in rheumatoid arthritis: Overwhelming, uncontrollable, ignored. Arthritis Rheum 2005;53:697-702. [MEDLINE] 6. World Health Organization. International classification of functioning, disability and health (ICF). Retrieved April 7, 2005 from http://www3.who.int/icf/intros/ICF-Eng-Intro.pdf. 7. Stamm TA, Cieza A, Machold KP, Smolen JS, Stucki G. Content comparison of occupation-based instruments in adult rheumatology and musculoskeletal rehabilitation based on the international classification of functioning, disability and health. Arthritis Rheum 2004;51:917-24. [MEDLINE] 8. Tugwell PS, Wilson AJ, Brooks PM, et al. Attributes and skills of an effective musculoskeletal consumer. J Rheumatol 2005; 32:2257-61. [MEDLINE] 9. Pincus T, Wolfe F. Patient questionnaires for clinical research and improved standard patient care: Is it better to have 80% of the information in 100% of patients or 100% of the information in 5% of patients? J Rheumatol 2005;32:575-7. [MEDLINE] 10. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005;113:9-19.[MEDLINE]
|
