Abstract
Objective. The treatment of rheumatoid arthritis (RA) should target patient-relevant outcomes, making patient perspective on remission essential. In 2010, patients, physicians, health professionals, and researchers at the Outcome Measures in Rheumatology (OMERACT) conference developed an ambitious research agenda to study the concept of remission. Qualitative research has since helped us understand the concept of remission from the patient perspective.
Methods. During OMERACT 12, the OMERACT working group on patient perspective on remission in RA elaborated on data generated to date and discussed the methodological challenges ahead. Challenges included (1) selection of domains, (2) choice of a patient remission definition or a single domain to add to the current remission definition, and (3) the importance of pain in defining remission from a patient perspective.
Results. Focus in the coming years will be on increasing our understanding by identifying the most important domains from the patient perspective regarding remission and investigating how these domains can be measured. Investigation into the Rheumatoid Arthritis Impact of Disease questionnaire, disease flare, as well as the concordance of domains from our ongoing remission survey is appropriate. More data and further discussions are needed to decide on the next steps.
Conclusion. Progress summarized over 4 years highlights the main methodological challenges discussed within the working group on patient perspective on remission in RA during OMERACT 12.
Developments in the treatment of rheumatoid arthritis (RA) have improved the control of joint inflammation and pain1,2, and a goal of minimal disease activity or even remission is now realistic3,4,5. Work on defining stronger criteria for remission6,7 reviewed outcome measures available from clinical trial data. These included only the 3 patient-reported outcomes (PRO) incorporated in the initial RA core outcome measurement set: patient global, pain, and physical function8,9. Data were not consistently available on other potentially important aspects of remission from the patient perspective, including fatigue, which had since been recognized as an important outcome to include in clinical trials10. At the Outcome Measures in Rheumatology meeting in 2010 (OMERACT 10), both participating patients and professionals endorsed the need to study the concept of remission in RA from the patient perspective, to identify new potential PRO to optimize targeted therapy and possibly improve the American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) definition of remission11.
Progress since OMERACT 10
Qualitative exploration
To understand the patient perspective on periods of remission, we undertook a qualitative study12. In 3 European countries (Austria, United Kingdom, and the Netherlands), focus group discussions were organized with patients with RA. From these discussions, 26 aspects of remission were identified and grouped into 3 major themes of patient-perceived remission (Figure 1). For elaboration on the different items and grouping into themes, we refer to this work12.
Main themes of rheumatoid arthritis remission. The 3 main themes (symptoms, impact, and normality) are underpinned by a theme of influential factors, with an overarching theme of assessment issues.
Rating for importance
To determine the relative importance of the many aspects of remission identified in our qualitative work, a large group of patients with RA is currently completing a ranking exercise. The goal of this exercise is to determine a short list of aspects related to remission that can then be further researched. The survey contains all 26 aspects of remission identified in focus group discussions. First, respondents are asked to rate each item as either “not important,” “important,” or “essential” for characterizing a period of remission. Second, respondents are then asked to determine their personal top 3 most important/essential items that characterize remission. A copy of the survey is available from the corresponding author.
Patients from the 3 aforementioned European countries are participating. In addition, all RA patient research partners (PRP) who took part in OMERACT 12 were invited to participate by e-mail. We aim to collect data from at least 100 patients.
Data Presented at OMERACT 12
During the Patient Perspective on Remission in RA special interest group (SIG) session at OMERACT 12, results of the focus group discussions12 as described above were presented. Additionally, preliminary results of the survey were presented. These data were limited to those from the first 49 respondents obtained prior to the conference, comprising patients with RA from Austria (n = 28), the Netherlands (n = 16), and OMERACT PRP (n = 5; Table 1).
Preliminary survey results: respondent demographics.
There are several ways to reduce the number of domains from 26 to a manageable number, while taking into account the importance of the different features of periods of remission. For presentation during our SIG, we approached the preliminary data in 2 ways:
We calculated the frequency that a particular domain was mentioned in the top 3, irrespective of its position (1, 2, or 3). Table 2 provides the 12 most-mentioned domains in the top 3 for 49 respondents. While pain was by far the most-mentioned feature, differences between other features were small. It was confirmed that a larger number of patients would be needed to identify other domains that play a key role in the patient perception of remission.
Of the 26 domains that were each rated as either “not important,” “important,” or “essential,” we removed domains that > 30% of patients identified as “not important.” The remaining domains were sorted by the percentage of patients that evaluated a particular domain as “essential.” This resulted in the removal of domains “Unpredictability of the disease;” “use of DMARD” (disease-modifying antirheumatic drugs); “use of pain-killers;” “mood;” and “the way other people see me.” Table 3 presents the 12 domains that are most frequently rated as “essential” to characterize a period of remission.
Top 3 features by frequency (%), irrespective of the first, second, or third position (preliminary data).
Frequency of features ranked as essential (%) for 49 respondents (preliminary data).
Approaching the data in these 2 ways resulted in agreement regarding the top 3 and essential features; however, there was discrepancy between the top 3 items and items rated “not important” by > 30%. Two items from the top 3 were also ranked as “not important” (use of DMARD, mood). Because these are preliminary findings based on limited data, final datasets will show whether the originally planned analysis method needs to be reevaluated.
Discussion at OMERACT 12
An estimated 60 to 80 OMERACT participants joined our SIG, including more than 6 PRP. About 15% of the participants actively contributed to the discussion.
Following the presentation of results as summarized above, 3 points for discussion were prepared:
Survey
How do we select the most important domains from a list of items rated as important and essential, in combination with a top 3 list?
Outcome
Do we need to work toward a patient remission definition, or should we determine the added value of including one of the PRO in the current ACR/EULAR remission definition?
Pain
What is the value of pain when included in the ACR/EULAR remission definition in addition to, or in place of, the patient global assessment of disease activity?
Survey: No specific methodological preference was evident for selecting the most important domains for patient-perceived remission from the survey. Several methodologies were discussed, including use of the Rheumatoid Arthritis Impact of Disease (RAID) methodology for ranking domains and distributing points13; use of principal component analysis to group domains, assuming that some of the items are not independent of each other and might represent the same underlying construct; use of computer assisted methodology; and use of a Venn diagram.
The issue was raised that, by asking patients to rank domains for importance, some will rank everything as either important or essential. It was suggested that the discrepancy between the methods of “top 3” and “ranking of essential items” might be solved by checking for and removing survey responses where all items are ranked as either important or essential.
Outcome: Preliminary data of 49 respondents as presented during the SIG were insufficient to make concrete plans about the next steps, i.e., whether to revise the ACR/EULAR criteria or draft a patient-reported core set of remission/absence of effect scale. More data are needed before deciding whether changing the criteria is necessary: Altering the criteria is not a trivial decision, with implications for researchers and funding bodies.
It was discussed in the context of the OMERACT filter 2.015 that it may be useful for the ACR/EULAR remission criteria to be retained as a pathophysiological definition, in addition to separate effect or functional remission criteria that might allow for identification of patients needing treatment beyond current targeted therapy models.
Overall, there seemed to be a preference for a separate patient remission definition. However, worries were also expressed: If not incorporated into the ACR/EULAR remission definition, the patient perspective of remission might not be used; as well, policy-maker viewpoints need to be taken into account before deciding the outcome of this project.
It was suggested that the anticipated patient sample (n = 100) for rating domains for importance is not large enough to enable the data to be divided according to disease and contextual factors. Several participants offered help in collecting more data, from different countries in and outside Europe, e.g., France and Australia. The centers in Amsterdam, Bristol, and Vienna agreed to expand their sample from the original 30 patients per country to at least 50 (excluding previous focus group participants and OMERACT patient delegates). It was discussed that we need to ensure inclusion of patients with both early and late disease, as well as erosive and nonerosive disease.
Pain: Although currently the highest ranked domain, pain is already reflected in the core set, albeit not in the remission criteria. The idea of taking this measure forward in additional research was discussed.
A discussion on the value of the patient global assessment of disease activity versus a measure of pain followed. Despite literature that questions the validity of the patient global assessment of disease activity in the remission criteria16,17, it was deemed an appropriate outcome because it would measure several important aspects of the effect of the disease on a patient (pain, fatigue, and physical functioning), rather than one aspect such as pain17. Stressing the importance of having multiple aspects of effect in the remission definition, patients expressed their preference for the patient global assessment above a measure of pain as a single component of the remission criteria.
We would need to know to what extent pain and other PRO are reflected in the patient global assessment.
General Discussion
During and after the SIG, similarities with the RAID and the disease flare group were further highlighted: As members of the remission group, we are studying the effect of RA disease activity from the perspective of its absence; the disease flare group is studying the effect of RA disease activity from the perspective of an acute and sustained increase; and the RAID is an instrument to assess the effect of RA, including disease activity. Rather than devise new tools to detect (the effect of) flare or remission, it was suggested to measure the additional identified domains not found in the RAID alongside the RAID; or to use a low cutoff RAID score to reflect remission in case the most frequently mentioned domains from the survey strongly resemble the 7 domains from the RAID.
Table 4 summarizes our research agenda.
Research agenda for the patient perspective on remission in rheumatoid arthritis.
Conclusion
The SIG on patient perspective on remission presented at OMERACT 12 elaborated on the data that have been generated to date and discussed the methodological challenges ahead. Focus in the coming years will be on increasing our understanding by identifying the most important domains from the patient perspective regarding remission and investigating how these domains can be measured. Investigation into the RAID and disease flare and their concordance with domains from our ongoing survey is appropriate. More data and further discussions are needed to decide whether this will lead to revision of the ACR/EULAR remission criteria or to establishing a patient core set of remission measures/an absence of effect scale.
Footnotes
Supported by a research grant from the European League Against Rheumatism.
