Abstract
Objective. To assess which clinical features are most important for patients, parents, and clinicians in the course of juvenile idiopathic arthritis (JIA).
Methods. Forty-nine people participated in 6 audience-specific focus group discussions and 112 reciprocal interviews in 3 Canadian cities. Participants included youth with JIA, experienced English- and French-speaking parents, novice parents (< 6 mos since diagnosis), pediatric rheumatologists, and allied health professionals. Participants discussed the importance of 34 JIA clinical features extracted from medical literature. Transcripts and interview reports underwent qualitative analysis to establish relative priorities for each group.
Results. Most study participants considered medication requirements, medication side effects, pain, participant-defined quality of life, and active joints as high priority clinical features of JIA. Active joint count was the only American College of Rheumatology core variable accorded high or medium priority by all groups. Rheumatologists and allied health professionals considered physician global assessment as high priority, but it had very low priority for patients and parents. The parent global assessment was considered high priority by clinicians, medium to high by parents, and low by patients. Child Health Assessment Questionnaire scores were considered low priority by patients and parents, and moderate or high by clinicians. The number of joints with limited motion was given low to very low priority by all groups. Parents gave high priority to arthritis flares.
Conclusion. If our findings are confirmed, medication requirements, medication side effects, pain, participant-defined quality of life, and active joint counts should figure prominently in describing the course of JIA.
- JUVENILE IDIOPATHIC ARTHRITIS
- PAIN
- PHYSICIAN GLOBAL ASSESSMENT
- TREATMENT
- PARENT GLOBAL ASSESSMENT
- CHILD HEALTH ASSESSMENT QUESTIONNAIRE
Juvenile idiopathic arthritis (JIA) encompasses conditions that share chronic arthritis of unknown cause as their defining characteristic1,2. The course of the disease and eventual outcomes vary among and within each of the International League of Associations for Rheumatology JIA categories3,4,5,6. Despite the advent of effective treatments and trends toward more favorable outcomes, our ability to predict the course of JIA remains limited7,8,9,10,11.
Currently, the course, outcomes, and response to treatment in JIA are described by focusing on 6 core variables endorsed by the American College of Rheumatology (ACR; Table 1). These variables were chosen based on the expert opinion of researchers and pediatric rheumatologists, and on the variable’s statistical performance12. The current definition of JIA improvement is based on these variables12, and recent definitions of inactive disease and remission use a subset of them13. The opinions of patients, parents, and allied health professionals about what is important in the course of JIA are less well known and were not formally elicited in selecting these 6 variables or the definitions of improvement, inactive disease, and remission.
In our study, first, we used qualitative research methods to assess which clinical features are most important for patients, parents, and clinicians in the course of JIA. Second, we also elicited their opinions about terms used to name JIA disease course and desirable attributes of clinical tools to predict JIA course.
MATERIALS AND METHODS
We conducted 6 study sessions in 3 Canadian cities between March and November 2012. Each session consisted of a focus group discussion and reciprocal interviews (participants interviewed each other). Participants were asked to address 3 questions: (1) what clinical features should be used to define JIA disease course (the experience of JIA and its treatment over time)? (2) what words should be used to describe the expected course of JIA to newly diagnosed children and their parents? (3) If we had a tool that would allow us to give better predictions of how JIA would turn out, what features of this tool would you think would be important?
We chose a qualitative research approach because this methodology allows for the gathering of complex, in-depth perspectives that are otherwise lost in formal surveys and hypothesis-driven quantitative research14,15,16. Our study was approved by the University of British Columbia’s Behavioural Research Ethics Board.
Subject recruitment
We recruited youth with JIA (YJIA) and parents of children with JIA during regular rheumatology clinic visits. We sought study participants with different types of JIA, varied disease severity, and different disease duration. Pediatric rheumatologists (PR) and allied health professionals (AHP) were recruited by contacting their professional associations and all academic pediatric rheumatology centers in Canada. To be included, they had to have at least 30% of their current practice focused on counseling and managing patients with JIA.
Study sessions
Sessions were professionally facilitated and audience-specific. Each consisted of a focus group discussion and reciprocal interviews. This format allowed researchers to document interactions, and similarities and differences in opinion within a group. It also allowed participants to hear and consider arguments from other participants in formulating their own opinions. Participants were provided with a list of definitions for 34 JIA clinical features often recorded in cohort studies and clinical trials. These 34 items were extracted from published literature by the authors as important items in monitoring cohorts of children with JIA. For patients and parents, disease course was defined as “the experience over time related to symptoms, signs and consequences of the disease and its treatment in a person”. In each session, the facilitator asked participants to review all the definitions provided, add any additional items considered important, and discuss what items were most important to them. Most groups were able to reach consensus on a number of high priority items, but did not rank all items.
Reciprocal interviews, modeled after Redelmeier, et al17, were used to focus discussions at a personal level, facilitate peer interactions, and allow correlation of perceived severity of JIA course with data extracted from medical records (data not shown). Clinicians were asked to do reciprocal role-play interviews, taking the role of the parent of a child with JIA. This was particularly helpful in assessing terms that should be used in discussing disease course with families.
Focus group and reciprocal interview probes used in our study are included in the supplementary data (available online at jrheum.org). To address our secondary aims, participants also received lists of terms used to name JIA disease course and of attributes of prediction tools (supplementary data available online at jrheum.org)18, and 2 examples of prediction tools19,20. A prediction tool was defined for patients and parents as “a way of using a combination of features about a person known today to provide an educated guess of what will happen years later”.
Analysis
Transcripts of focus group audio recordings, detailed facilitator’s session notes, participant-generated lists of high priority items, and reciprocal interview answers underwent detailed analysis as follows21. First, major themes discussed in each session (emerging themes) were identified by OGR and discussed with JG. This provided interpretative context before focusing on the prioritization of items. Second, analytical tables were created for each session. These tables reported the supporting and/or counter-arguments about the relative priority of each item discussed by participants. Third, drawing on these tables, relative high, medium, low, or very low priority were assigned to each item by 2 investigators (JG, OGR). In general, high-priority items were those explicitly chosen by participants as top priority; medium priority were those that were initially identified by participants as important, but after discussion were thought less important than those in the top priority list; low priority were those that were not substantively addressed in the discussion or did not draw enough qualitative supporting arguments in our analysis; and very low priority were those that were not discussed to any extent during the session. Fourth, session summaries were discussed with all investigators. These summaries contained descriptions of emerging themes, illustrative quotes, and items listed in order of priority with the supporting rationale for their assigned priority. Finally, after all sessions had been individually examined, investigators met to discuss overall findings and key messages of our study.
RESULTS
A total of 49 people participated in 1 of the 6 focus groups and in 2 or 3 of the 112 reciprocal interviews carried out for our project. Characteristics of study participants are listed in Table 2.
Clinical features
Table 3 shows the relative priority of all clinical features discussed, according to each group. All groups agreed that the medications required for controlling the disease and the side effects of these medications were high priority features of the JIA disease course.
The number of swollen or active joints was seen as high priority by most groups, except YJIA who saw it as medium priority. Pain and quality of life were also seen as high priority by most groups, except by AHP who accorded these features medium priority. It should be noted that AHP focused on clinical features they felt predicted later outcomes. Interestingly, while quality of life assessed by the simple Quality of My Life scale was seen as high priority, the total Juvenile Arthritis Quality of Life Questionnaire (JAQQ) score was not.
With the exception of the number of active joints, ACR core variables were not given high priority across groups. The parent global assessment was given high priority by PR and AHP, medium to high priority by parents, and low priority by YJIA. Both PR and AHP considered the physician global assessment had high priority, but this was hardly discussed or given any priority by patients and parents, although they said the concept of disease activity was important. Functional ability, defined as Child Health Assessment Questionnaire (CHAQ) scores, was given low priority by patients and parents, medium priority by PR, and high priority by AHP. The number of joints with limited motion was given low to very low priority by all groups. Laboratory markers of inflammation were not included in the list of items for discussion.
As requested, study participants added other clinical features they deemed important. These included uveitis by experienced parents, PR, and AHP, and nonadherence and compliance by AHP. Experienced parents were particularly interested in identification of disease flares and flare triggers.
Most groups struggled with the overlap among clinical features. It was argued that some were different ways of saying the same thing, and some were parts or components of others. For example: the Quality of My Life scale and the JAQQ seemed to aim at the same idea, and some YJIA stated that quality of life included things such as needing help, symptom difficulties, and missing school. Because of this overlap, participants sometimes argued that a particular feature should not go into the top priority list because it was a component of, or equivalent to, another feature already in the list.
Secondary aims
When discussing terms used to describe JIA course to families, all groups agreed that positive, easy-to-understand terms were preferable. Yet, this was tempered by parents and AHP stating that terms needed to be honest and direct, especially for children with more severe forms of JIA, so that the seriousness of the disease was not misjudged, potentially compromising adherence to treatment. “Active JIA”, “Inactive JIA”, and “Remission” were terms accorded medium to high priority by all groups. “Controlled”, “Mild”, “Moderate”, and “Severe JIA” were seen as helpful terms by YJIA, PR, and AHP, but parents were concerned they lacked specificity and felt that the term “Severe JIA” was too pessimistic.
“Ease of use”, defined as “using the tool requires little time and training,” was the only tool attribute accorded high priority by all groups. “Inclusion of Guidelines” and “Consistency” were seen as important by most groups.
Developing a JIA course prediction tool was seen as a worthwhile undertaking, but participants differed in their opinions about what the tool should look like and its potential uses. Experienced parents, PR, and AHP expressed skepticism about the feasibility of developing an accurate prediction tool of JIA course at diagnosis. Reasons for this skepticism included intrinsic unpredictability of JIA, insufficiency of current knowledge, and management decisions that will change the disease course. Although prediction tools provided as examples took the form of risk calculators (i.e., Centor and Framingham scores)19,20, more creative formats were envisioned during the discussions. Experienced parents argued that the tool should be a checklist they could use to detect JIA flares and flare triggers; AHP visualized an educational tool about the course of JIA that could be used in counseling families; and PR suggested a road map showing decision points as forks in the road, and not just the destination or final outcome.
Additional details of the prioritization of terms and tool attributes are included in the supplementary data (available online at jrheum.org).
Emerging themes
Analysis of transcripts revealed several emerging themes important in contextualizing the discussions that took place and the resulting prioritization of items (Table 4). YJIA were concerned about making their views heard, and thus gave lower priority to standard questionnaires and scores that did not allow much space for them to express themselves. Similarly, parents also voiced dissatisfaction with numerical indicators of outcome.
Experienced parents still recalled how shattering it had been when receiving a JIA diagnosis and felt that dealing with JIA was a time-consuming roller coaster. This likely influenced their interest in knowing what precipitated flares and how to prevent them. For novice parents, the recent diagnosis was a shocking experience and prioritizing clinical features was a very difficult task. They were just beginning to come to terms with the nature of the disease and what it would mean for their child’s future. Experienced and novice parents highlighted their role in monitoring their children’s disease, and this may be why they selected the parent global assessment as an important feature.
PR discussed in depth how to use indices and scores to reflect the course of JIA. AHP emphasized their role in educating families about JIA and some stated it was important to say that JIA is unknown and unpredictable, to avoid offering false hope.
DISCUSSION
Our study describes the opinions of groups of Canadian patients, parents, and clinicians about what matters most in the course of JIA. Their priorities often differed from traditional priorities reported in the literature, and there were important similarities and differences among groups.
The most obvious similarity among groups was that medication requirements and medication side effects were considered very important in describing the course of JIA. This suggests that a qualitative change in how we describe disease course may be needed, because traditional descriptions have focused on features of the disease itself, with relatively little attention to required treatments and side effects. The risk of side effects needs to be balanced with medication-induced improvements. Burnett, et al22 found that parents of children with JIA favored effective treatments that reduced pain and improved functioning despite their risk of side effects.
Most patients, parents, and clinicians in our study agreed on the importance of the number of active joints, pain, and participant-defined quality of life. This finding is consistent with previous studies23,24,25,26,27,28,29,30,31 and, if confirmed in other populations, would argue for the inclusion of pain and quality of life as core features in describing the course of JIA32.
Study groups differed in the significance accorded to the ACR core variables. While PR and AHP considered active joint count, parent global, physician global, and CHAQ to have medium to high priority, patients and parents only agreed on the importance of the active joint count. Despite its low rating by patients and parents, in our view, the physician global assessment should not be discarded from research studies and clinical trials as our study and others show it is important for clinicians33,34. Erythrocyte sedimentation rate was not included in the list of features provided to participants, and thus we are unable to comment on its priority.
It could be argued that the question we asked study participants (what clinical features should be used to define the course of JIA?) has a different purpose than the ACR core variables (to provide standard outcome measures to assess biologic response and treatment efficacy)12. We agree that standard measures are essential for research in JIA. Yet, if our findings of low prioritization of ACR core variables by patients and parents were confirmed, it would mean that treatments for JIA can be deemed beneficial because they improved measures of little relevance to patients and parents. From their point of view, it would seem illogical that improvements in pain and quality of life are not considered when confirming that arthritis treatments work. In our opinion, it is appropriate to ask that criteria used to define improvement and demonstrate treatment efficacy in JIA include the issues that matter most to patients and their parents.
It is possible that the discrepancies between the ACR core variables and the priorities reported in our paper are not as profound as our findings might suggest. For example, it could be argued that the parent/patient assessment of overall well-being included in the ACR core set is a good proxy for quality of life. Although this may be a reasonable assumption, the question used in most studies to assess overall well-being (Considering all the ways that arthritis affects your child, rate how your child is doing.) was included in our list (parent global assessment), and was seen as different from directly asking the child how their health has affected their quality of life. In general, assessing the conceptual equivalence between ACR core variables and the priorities reported in our present study would require (1) demonstration that patients, parents, and clinicians agree with the proposed equivalence of concepts; and (2) empirical demonstration of close correlation of the respective measures in patients over time.
One other interesting difference among groups was that “Remission on Medications” and “Clinical Remission” had low priority for parents, while they were high priority for patients and clinicians. We believe this is partly explained by the high priority given by parents to disease flares, the contrary to remission. A similar explanation could be proposed for the differences in the priority of “Active Disease” and “Inactive Disease”; while parents considered “Active Disease” had moderate priority, “Inactive Disease” was barely discussed. It should be pointed out that Consolaro, et al have shown that parents and physicians have somewhat differing views about what constitutes inactive disease33.
Study strengths and limitations
Our study provides a summary of the opinions of groups of Canadian patients, parents, and clinicians about the relative priorities of JIA clinical features, but does not provide estimates of the proportions of patients, parents, or clinicians endorsing each opinion. The latter would require a followup survey with a larger sample. We included participants from different JIA categories, degree of severity and length of experience with the disease, and clinicians with diverse background and location of practice. Participants from Canadian cultural or ethnic minorities were included, but their views were not analyzed separately. The opinions of people in other countries may differ. Additional surveys or focus groups are needed to confirm whether our findings hold across populations.
Caution is warranted in interpreting differences in priorities among study groups for at least 3 reasons: (1) it is possible that an item received little discussion by a group (and, thus, was assigned a low or very low priority) because participants felt the item overlapped with or was embedded into another important item, and not because the item itself was unimportant; (2) it is possible that more in-depth knowledge of an item would convince a group of participants to change their mind about its priority (for example, if participants had more intimate knowledge of the components of the JAQQ and how carefully the components were selected, they may accept the JAQQ as a better reflection of quality of life than a simple direct question); and (3) the differentiation of high priority items was discussed in-depth during the sessions, but the distinctions between moderate, low, and very low priority were made by researchers analyzing transcripts and interview responses; thus, the difference between 1 group giving a moderate priority to 1 item and another group giving it a low priority may not be substantial.
One of the strengths of our qualitative design is that it allows for unexpected results to emerge. We found that conventional terms used to describe clinical features and disease course were often used by patients and parents with a different meaning. Sometimes there was clear disagreement between participants’ understanding of a concept and its intended use in scientific reports. For instance, when patients said that the effect of arthritis on their quality of life was very important, they did not mean that the JAQQ score was very useful. Patients clearly preferred to use their own words (qualitative) instead of standardized scores and scales (quantitative) developed for rigorous measurement. These discordances reflect the differing ways in which qualitative and quantitative studies detect and express clinical data, and how they may supplement each other14,15,16.
With only 1 study session per group, we cannot assure thematic saturation21, or that an additional study session would result in the same emerging themes. However, our study was not designed to reach thematic saturation, but to allow thoughtful prioritization of clinical features.
Implications for practice
In our opinion, the points of view of patients and parents should be systematically considered in designing JIA care plans, outcome measures for research and clinical care, and in developing prediction tools. If our findings of low prioritization of ACR core variables by patients and parents are confirmed in other settings, it would mean that definitions of improvement, inactive disease, and remission derived from these variables have little relevance for them. To make them more relevant, core variables would need to include measures of pain and quality of life, and perhaps exclude less relevant measures, such as the number of joints with limited movement or the CHAQ score. Simple, positive terminology should be used in discussing JIA disease course with families, and innovative ways of describing the course of JIA in verbal narratives, in addition to numbers, may help better engage patients and parents in meaningful decision-making. This could be achieved with the use of clinical vignettes, analogous to the ones used in case-based learning in medical schools.
Patients, parents, and clinicians participating in our study identified medication requirements, medication side effects, pain, participant-defined quality of life, and active joint counts as top priorities in the course of JIA. Some ACR core variables were accorded low priority by patients and parents. If our findings are confirmed in other populations, this would suggest that current definitions of JIA improvement, inactive disease, and remission have little relevance for patients and parents, and that their relevance could be increased by explicitly including changes in the child’s pain and quality of life into those definitions.
ONLINE SUPPLEMENT
Supplementary data for this article are available online at jrheum.org.
Acknowledgment
We thank Angelyne Sarmiento, Kevin Huang, Emma Hitchcock, Stephanie Wong, Michele Gibbon, Solen Roth, Lynn Skotnitsky, and Isabelle Eaton for their assistance with participant recruitment, study session logistics, focus group facilitation, data entry, and translation. We would also like to express our gratitude to the patients, parents, and clinicians who generously took the time to participate in this project.
Footnotes
-
Supported by a research grant from the Canadian Rheumatology Association under the Canadian Initiative for Outcomes in Rheumatology Care/Initiative Canadienne pour des resultats en soins rhumatologiques (CIORA/ICORA) program.
- Accepted for publication July 15, 2014.