NTNU Norwegian University of Science and Technology Faculty of Medicine and Health Sciences Department of Clinical and Molecular Medicine
Isabelle Pignatel Jenssen and Sofie Mikalsen Arneng
The Patient's Voice
Association between self-reported physical
health and other patient- and physician-reported outcome measures in Juvenile Idiopathic Arthritis (JIA) based on data from a Norwegian JIA cohort study (The NorJIA study)
Graduate thesis in Programme of Professional Study, Medicine Supervisor: Marite Rygg
Co-supervisor: Pål Richard Romundstad January 2022
Gr aduate thesis
Isabelle Pignatel Jenssen and Sofie Mikalsen Arneng
The Patient's Voice
Association between self-reported physical health and other patient- and physician-reported outcome measures in Juvenile Idiopathic Arthritis (JIA) based on data from a Norwegian JIA cohort study (The NorJIA study)
Graduate thesis in Programme of Professional Study, Medicine Supervisor: Marite Rygg
Co-supervisor: Pål Richard Romundstad January 2022
Norwegian University of Science and Technology Faculty of Medicine and Health Sciences
Department of Clinical and Molecular Medicine
v
Sammendrag
Bakgrunn
Juvenil idiopatisk artritt (JIA) er en kronisk sykdom som påvirker barns liv på en rekke områder. De siste årene har det vært et økende fokus på pasientrapporterte utfallsmål (PROMs) både for bruk i forskning og i klinikken. The Child Health Assessment
Questionnaire (CHAQ) og CHQ Physical Summary Score (CHQ PhS) er to eksempler på PROMs som begge vurderer den fysiske tilstanden til barnet. Vårt mål med denne studien var å evaluere assosiasjonen mellom pasientrapportert fysisk helse målt med det
sykdomsspesifikke verktøyet CHAQ, og andre pasient- og legerapporterte utfallsmål.
Videre ville vi studere pasientrapportert fysisk helse ved hjelp av CHQ PhS, som er et generisk verktøy, i samme studie sammenliknet med en matchet kontrollgruppe.
Metode
Vi har brukt data samlet inn i NorJIA studien, en prospektiv multisenter
observasjonsstudie som også inkluder en alders- og kjønnsmatchet kontrollgruppe. Vi har sett på kliniske sykdomskarakteristika, CHAQ- og CHQ-skår, selvrapportert smerte og generell allmenntilstand, morgenstivhet, legerapportert sykdomsaktivitet,
sykdomsstatus og behandling. For å vurdere sammenheng mellom pasientrapportert fysisk funksjonshemming med pasient- og legevurderte utfallsmål har vi brukt kategorisk tabellanalyse for ujusterte analyser, og logistisk regresjonsanalyse justert for alder og kjønn for å vurdere justerte odds ratio (OR) med 95% konfidensintervall (CI).
Resultater
I alt 221 av 228 deltakere med JIA fra NorJIA-studien fylte ut CHAQ, og ble derfor inkludert i hovedstudien vår. Videre fylte 218 av disse, samt 207 av de 224 kontrollene i NorJIA studien ut CHQ, og ble derfor inkludert i vår CHQ delstudie. I tillegg fylte 218 av disse, samt 207 kontroller, ut CHQ og ble derfor inkludert i vår CHQ delstudie. Av de 218 deltakerne med JIA rapporterte 57.9% at de hadde fysisk funksjonshemming (CHAQ >
0). Videre rapporterte 24.3% en fysisk helse under befolkningsnormalen (CHQ PhS <
40), og 8.7% rapporterte en psykososial helse under normalen (CHQ PsS < 40). Dette var høyere sammenliknet med kontrollgruppen, hvor 0.5% skåret under normalen for fysisk helse og 1.9% skåret under normalen for psykososial helse. Videre fant vi en sterk assosiasjon mellom selvrapportert fysisk funksjonshemming og andre utfallsmål, både pasient- og lege-rapporterte. De mest fremtredende funnene var assosiasjonen mellom fysisk funksjonshemming og aktiv sykdom (justert OR 32.8, 95% CI 9.7-110.8) og fysisk funksjonshemming og selvrapportert fysisk helse (CHQ PhS) (justert OR 21.2, 95% CI 6.2-71.1).
Konklusjon
Resultatene våre viste at andelen barn som rapporterer funksjonshemming fremdeles er høy blant barn med JIA i Norge, og at det er en sterk assosiasjon mellom
pasientrapportert fysisk funksjonshemming og andre pasientrapporterte utfallsmål. Det er også en klar assosiasjon mellom pasienters egenvurdering av fysisk
funksjonshemming og legens vurdering av pasientens sykdomsaktivitet.
vi
Abstract
Background
Juvenile idiopathic arthritis (JIA) is a chronic disease that affects several aspects of children's lives. Recently, there has been a growing focus on patient-reported outcome measures (PROMs) for use in both research and clinical settings. The Child Health Assessment Questionnaire (CHAQ) and the Physical Summary Score of the Child Health questionnaire (CHQ PhS) are two examples of such PROMs that assess the physical function of the child. Our objective was to evaluate the association between patient- reported physical disability assessed with the disease specific CHAQ with other patient- and physician-reported physical outcome measures. Furthermore, we wanted to study patient-reported physical health, assessed with the generic CHQ PhS, in the same study cohort compared to a matched control cohort.
Methods
We used data collected from the NorJIA study, a prospective, multicentre, observational study. We also included data from the age- and sex-matched control cohort of the NorJIA study. The measures used in this thesis were clinical characteristics, CHAQ and CHQ scores, self-reported pain and general well-being, morning stiffness, physician-reported general assessment of disease activity, disease status, and treatment. To evaluate the association of patient-reported physical disability with other patient- and physician- reported outcome measures, we used categorical table analysis for crude analysis and logistic regression analyses adjusted for the age and sex to estimate the adjusted odds ratio (OR) with 95% confidence interval (CI).
Results
In this study, 221 out of the 228 participants with JIA filled in the CHAQ, and were therefore included in our main study. Furthermore, 218 of the same participants, and 207 of the 224 controls in the NorJIA study had filled in the CHQ, and were therefore included in our CHQ substudy. Of the 218 participants with JIA, 57.9% reported physical disability of some level (CHAQ > 0). Compared to the control group, where 0.5% scored below the norm (CHQ PhS < 40) for physical health and 1.9% scored below the norm for psychosocial health (CHQ Psychosocial Summary Score (PsS) < 40), the corresponding numbers for the JIA group were 24.3% and 8.7%, respectively. Furthermore, we found an association between self-reported physical disability and all the other patient-and physician-reported measures investigated. The most pronounced results were the strong associations between physical disability and active disease evaluated by the physician (adjusted OR 32.8, 95% CI 9.7-110.8) and between physical disability and physical health scored by CHQ PhS (adjusted OR 21.2, 95% CI 6.2-71.1).
Conclusion
Our results showed that the proportion of children reporting physical disability remains high in JIA, and that there is a strong correlation between patient-reported physical disability and other patient-reported outcome measures. There is also a clear correlation between the patients’ evaluation of their own physical disability and the physicians’
evaluation of the patient’s disease activity.
vii
Collaborators
Supervisor: Marite Rygg, MD, PhD, Department of Clinical and Molecular Medicine (IKOM), NTNU, and Department of Paediatrics, St. Olavs Hospital, University Hospital of Trondheim.
Co-supervisor: Pål Richard Romundstad, PhD, Department of Public Health and Nursing (ISM), NTNU.
Collaborator: Anette Lundestad, MD, Department of Clinical and Molecular Medicine (IKOM), NTNU, and Department of Paediatrics, St. Olavs Hospital, University Hospital of Trondheim.
Acknowledgements
First and foremost, we want to express our gratitude towards the children and parents who participated in the NorJIA-study and provided the valuable information that made this thesis possible. We also want to thank the research group of NorJIA for giving us permission to use their data for this study.
Furthermore, we want to thank Marite Rygg for her excellent guidance and enthusiasm throughout the whole process of writing this thesis, and for teaching us so much about both JIA and research in general. We thank her for all her time and advice. We also want to thank Pål Richard Romundstad and Anette Lundestad for being available to answer our questions whenever we needed it. A final thank you to Berit for her Excel competence and attention to detail.
viii
ix
Table of contents
Abbreviations ... x
1 Introduction ...12
1.1 Outcome in JIA ...12
1.2 Patient-reported outcome measures (PROMs) ...13
1.3 Treatment and follow-up of JIA ...14
1.4 Objectives ...14
2 Materials and methods ...15
2.1 Study design ...15
2.2 Data collection ...15
2.3 Measures ...16
2.3.1 Clinical characteristics ...16
2.3.2 Patient-reported outcome measures ...16
2.3.3 Physician-reported outcome measures ...16
2.3.4 Composite disease activity measures ...17
2.3.5 Treatment ...17
2.4 Statistical analysis ...17
3 Ethics ...18
4 Results ...19
4.1 Sample characteristics ...19
4.2 Demographics ...20
4.3 Disease characteristics in the JIA group ...21
4.4 Physical and psychosocial health – CHQ ...23
4.5 Self-reported physical disability according to disease characteristics ...24
4.6 Disease characteristics according to sex, age, and disease categories ...25
4.7 Association between self-reported physical disability (CHAQ) and other disease outcomes ...26
5 Discussion ...29
5.1 Main findings ...29
5.2 Strengths ...29
5.3 Limitations ...29
5.4 Comparison with other studies ...30
5.5 Clinical implications ...31
6 Conclusion ...33
7 References ...34
x
Abbreviations
ACR American College of Rheumatology ANA Anti-nuclear antibody
bDMARDs Biologic DMARDs BMI Body mass index
CHAQ Child Health Assessment Questionnaire CHQ Child Health Questionnaire
CHQ-PF50 Child Health Questionnaire Parent form CHQ-PhS Physical Summary score
CI Confidence interval CRP C-reactive protein
DMARDs Disease-modifying antirheumatic drugs ESR Erythrocyte sedimentation rate
GDPR General Data Protection Regulation HAQ Health Assessment Questionnaire HLA-B27 Human leukocyte antigen B27
IKOM Department of Clinical and Molecular Medicine
ILAR International League of Associations for Rheumatology IOTF International Obesity Task Force
IQR Interquartile range
ISM Department of Public Health and Nursing JIA Juvenile idiopathic arthritis
NorJIA Norwegian multicentre study on imaging, oral health and quality of life in children with JIA
NRS Numeric rating scale
NSAIDs Nonsteroidal anti-inflammatory drugs
OR Odds ratio
PatGA Patient’s global assessment of overall well-being PhysGA Physician’s global assessment of disease activity
PRINTO Pediatric Rheumatology International Trials Organisation PROM Patient-reported outcome measure
PsS Psychosocial summary score
REC Regional Committees for Medical Research Ethics RF Rheumatoid factor
SD Standard deviation sDMARDs Synthetic DMARDs
TkMidt Center for Oral Health Services and Research, Mid-Norway
TkNN Public Dental Health Service Competence Centre of Northern Norway TkV Oral Health Centre of Expertise in Western Norway
VAS Visual analogue scale
xi
12
There are several conditions that may cause chronic arthritis in children. The most common are grouped under the umbrella term Juvenile Idiopathic Arthritis (JIA) [1].
According to the International League of Associations for Rheumatology (ILAR), JIA describes arthritides with an onset before the age of 16, and where the aetiology is unknown and other known conditions have been excluded [2]. ILAR defines arthritis as
“Swelling within a joint, or limitation in the range of joint movement with joint pain or tenderness, which persists for at least 6 weeks, is observed by a physician, and is not due to primarily mechanical disorders or to other identifiable causes” [2]. A systematic literature review found that the incidence rate of JIA in children below 16 years of age in Europe is 8.9 per 100 000 with a large range between studies and different geographical origin [3]. In the Nordic countries the incidence is higher, with an incidence of 23 per 100 000 children in Northern and Central Norway [4].
JIA comprises a range of conditions, and the JIA patient population is a heterogenous group [5]. Therefore, ILAR has provided 7 categories of JIA, based on the patients’
clinical and laboratory characteristics during the first 6 months after disease onset [2].
The causes of JIA seem to be related to both genetic predisposition and environmental factors, but generally there is limited knowledge on the aetiology [6, 7]. The
pathophysiology of JIA includes persistent joint inflammation, and the disease process involves immune dysregulation and a range of proinflammatory cytokines. Symptoms of JIA vary from patient to patient, but generally include joint pain, swelling and limited range of joint motion, morning stiffness, reduced function, and fatigue [1]. Some patients also may experience extraarticular symptoms such as uveitis and skin manifestations.
1.1 Outcome in JIA
Children living with JIA face an unpredictable disease course with periods of
exacerbations and remissions. A follow-up study conducted in the Nordic countries found that only 33% of patients were considered as being in clinical remission off medication 18 years after disease onset [8]. With the development of new and more effective drugs in JIA during the later years, the ultimate treatment goal is no longer only to reduce suffering, but to achieve sustained inactive disease (remission) [9].
But what is inactive disease? The complex set of symptoms in JIA and numerous potential consequences for the patient, cannot be measured with one simple measure, but rather must be considered in a multidimensional way [10]. Since children with JIA are a heterogenous group, one outcome measure cannot cover the range of the patients’
symptoms. Consequently, some outcome measures are based on clinical findings scored by the physician, some focus on patients’ and parents’ reports and some combine these two approaches into various composite scores to obtain a quantitative measure of the patient’s general condition in a multidimensional way [10]. These composite measures may be used in both research and clinical practice [11].
To be able to precisely measure and describe disease activity and inactivity, Wallace et al. defined preliminary criteria for inactive disease in 2004 [12], and in 2011 a revised
1 Introduction
13
set of criteria defining inactive disease were endorsed by The American College of Rheumatology (ACR), including the following criteria [13]:
• No active arthritis
• No fever, rash, serositis, splenomegaly or general lymphadenopathy due to JIA
• No active uveitis (< 1 cell in field size 1 mm by a 1 mm slit beam)
• Normal erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) attributable to JIA
• The physician’s global assessment of disease activity (PhysGA) Visual Analogue Score (VAS) [14, 15] indicates no activity
• Patient-reported morning stiffness ≤ 15 minutes
Wallace et al. have also established two types of clinical remission criteria [12]:
• Clinical remission on medication: Maintained inactive disease on medication for a minimum of 6 continuous months
• Clinical remission off medication: Maintained inactive disease without medication for a minimum of 12 continuous months
As previously mentioned, JIA is a complex, unpredictable condition which differs highly from patient to patient, and it has been shown that the importance of the various clinical features of JIA may be assessed and weighted differently by the physician, parents, and patient, respectively [16]. Thus, in addition to the physician’s evaluation of disease status and disease activity, there is a need to include the patient’s voice into the assessment of the disease.
1.2 Patient-reported outcome measures (PROMs)
Over the past decade, patient-reported outcome measures (PROMs) have been given increasing attention and significance in the clinical evaluation and follow-up of JIA, as well as in JIA research [17]. There are several different PROMs used in JIA, each focusing on different outcomes, such as patient-reported pain, patient’s assessment of general well-being, patient-reported quality of life, and patient’s assessment of physical health related to their disease. PROMs are thought to help improve the patient’s treatment by providing the physician with important insight into both the patient’s perception of their condition as well as what the patient considers to be the most pressing issues [17]. This insight may then be used to tailor the individual patient’s care and treatment. Challenges may arise when comparing PROMs from several patients, as each patient regards their disease in a different and subjective way (intra-individual comparisons). When used as repeated measurements reported by a patient, however, PROMs may give valuable information about the consequences and course of the disease (repeated intra-individual measurements). Thus, PROMS are an important instrument in the work of enhancing the patient’s quality of care [11, 17, 18].
The Childhood Health Assessment Questionnaire (CHAQ) is a tool validated for use in JIA.
This is a disease-specific tool covering the child’s functional ability in their everyday life [19, 20]. In 1997, the Pediatric Rheumatology International Trials Organisation (PRINTO) defined a core set of outcome measures to be used in clinical trials. The CHAQ was chosen as the principal disease-specific tool to be used in JIA cross-culturally [20]. It is considered to be both valid and reliable, whilst also being quick and easy to use [19].
Nevertheless, CHAQ has been shown to have some limitations, both in research and in clinical settings. It has been emphasized that CHAQ may miss information on certain
14
aspects of the disease [21]. Furthermore, when parents fill in the questionnaire on behalf of their children below 9 years, they might not capture the true perspective of the child [22]. There has also been shown discordance between the parents' and physicians' evaluation of the physical disability of the child [23].
Whereas the CHAQ is a disease-specific tool for assessment of physical health and functional disability, the Child Health Questionnaire (CHQ) is a multidimensional generic tool used for measuring the physical and psychosocial status of the child not related to any specific disease [24]. The most frequently used version of the CHQ is the parent- administered version - CHQ-PF50 (hereafter referred to as CHQ). Like CHAQ, it has been selected by PRINTO as a tool to be used for JIA cross-culturally [20]. The CHQ is a
generic instrument for physical and psychosocial health and is commonly used in JIA, and the scoring may also be compared to control groups from the normal population due to the generic nature of the instrument. Its purpose is to capture the physical, emotional and social components of the patient’s health status, which in other words can be
described as their health-related quality of life [24]. Unfortunately, and opposite the free use of CHAQ, the use of CHQ require a project-specific limited-use license.
Other examples of PROMs are patient-reported pain and patient-reported global
assessment of well-being (PatGA), both measured on a visual analogue scale (VAS pain), and patient-reported morning stiffness.
Despite the increased focus, limited research has been done on the associations between CHAQ, CHQ and other PROMs, and especially on the associations of CHAQ with physician- reported outcome measures and composite disease activity measures.
1.3 Treatment and follow-up of JIA
The majority of children with JIA need a treatment plan consisting of a combination of pharmacological, physical, psychosocial and in some very few cases surgical approaches [1]. During the last 20 years, we have seen major pharmacological advancements within rheumatology, making it possible to have clinical remission as the main goal of the
patient’s treatment plan. The most central medications in this development are intra- articular glucocorticoids and disease-modifying antirheumatic drugs (DMARDs), including synthetic (sDMARDs) such as methotrexate, and biologic drugs (bDMARDs). In addition to these drugs, nonsteroidal anti-inflammatory drugs (NSAIDs) have been, and are still, part of the initial treatment of JIA and to relieve pain in shorter periods of disease flares [9]. Furthermore, physical and occupational therapy are also an important part of helping the child in their everyday life, both for minimizing disability and maintaining function.
Counselling and providing the family with knowledge about their child’s disease is also of major importance [1].
1.4 Objectives
The aim of this thesis was to evaluate the association between patient-reported physical health, assessed with the disease-specific CHAQ, with patient- and physician-reported disease activity and outcome measures in a Norwegian JIA cohort. Furthermore, we wanted to study patient-reported physical health, assessed with the generic Physical Summary Score of The Child Health questionnaire (CHQ PhS), in the same study cohort compared to a matched control cohort.
15
2.1 Study design
The NorJIA study is a prospective, multicentre, observational study, https://norjia.com/.
Cases were recruited from out-patient clinics at the Department of Paediatrics at St.
Olavs Hospital in Trondheim, Haukeland University Hospital in Bergen, and University Hospital of North Norway in Tromsø. The participants with JIA, who had random disease durations, were examined at baseline and then after 24 months. The baseline clinical examinations were set in Bergen and Tromsø between 2015 and 2018, and in Trondheim between 2016 and 2018. Inclusion criteria were children between the age of 4 and 16 years meeting the ILAR criteria for JIA [2], with parents’ or legal guardians’ (hereafter only referred to as parents) informed consent. There were no exclusion criteria.
The matched control cohort consisted of children without JIA, matched 1:1 for age and sex. They were recruited from dental offices related to The Oral Health Centre of Expertise in Western Norway (TkV), The Center for Oral Health Services and Research, Mid-Norway (TkMidt), and the Public Dental Health Service Competence Centre of Northern Norway (TkNN). The children were visiting for a free dentist appointment according to the Norwegian public dental service policy and were thereby recruited randomly. The study design of this thesis is a cross-sectional study using data from the baseline examination. For this particular study, the inclusion criteria were participation in the baseline visit of the NorJIA study with a completed CHAQ. In sub-analyses,
participants with or without JIA (controls) with completed CHQ were included.
2.2 Data collection
The NorJIA study ran over 5 years and included extensive clinical, laboratory, radiological and oral examinations at baseline and after 2 years’ follow-up. It also included several PROMs, including CHAQ and CHQ. Relevant data from the baseline study visit includes:
Sex, age, parental education level, ethnicity, anthropometric data (weight and height), date of disease onset as judged by the paediatric rheumatologist, blood test results, JIA category according to the ILAR classification criteria [2], joint status evaluated by the physician at the study visit (including number of active joints), disease status at the study visit according to the Wallace criteria [12, 13], and medication (past and present).
Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) were both assessed at the clinical examination, while other blood tests such as Human leukocyte antigen B27 (HLA-B27), rheumatoid factor (RF), and anti-nuclear antibody (ANA) were all measured around the time of the diagnosis. ANA and RF were measured twice at least 3 months apart; for ANA an indirect immunofluorescence assay on HEp-2 cells was used.
In addition to the clinical examination, either the patient (if the patient was nine years or older), or the parent (if the child was younger than nine years), filled out the CHAQ form.
In addition, pain as well as overall well-being, were reported by the participants (patient or parents, as above) on visual analogue scales (VAS). They also reported duration of morning stiffness. Parents (regardless of the patient’s age) also filled out the CHQ
questionnaire. Parents from participants of the sex- and age-matched control cohort also filled out the CHQ.
2 Materials and methods
16
2.3 Measures
2.3.1 Clinical characteristics
Parent education was reported by the parents of both children with JIA and controls, and subsequently divided into 4 levels: Primary and middle school (7 to 10 years of
education), high school (11 to 13 or 14 years), less than 5 years university education, and more than 5 years of university education. The last two levels are hereafter collectively referred to as Parents with higher education.
Body mass index (BMI) was calculated using the following formula: Weight
(kilograms)/[height (metres)]2 and subsequently adjusted for age and sex according to the International Obesity Task Force (IOTF) cut-off values to provide an Iso-BMI
stratification that allows for comparison with adult BMI groups: Underweight (BMI <
18.5), normal (BMI ≥ 18.5), overweight (≥ 25), obesity (≥ 30) [25, 26].
2.3.2 Patient-reported outcome measures
CHAQ: Patient-reported physical health was assessed with Child Health Assessment Questionnaire (CHAQ). CHAQ evaluates the patient’s physical functioning in their everyday life during the last week, with 30 items divided into 8 domains: Dressing,
arising, eating, walking, hygiene, reach, grip, and activities. The need of assistance in the form of aids, devices and help from another person for physical functioning are also registered. Based on the patients’ answers, a global CHAQ score is calculated for each patient. This score ranges from 0 (no or minimal physical disability) to 3 (very severe physical disability) [19, 20].
VAS pain: Patient/parent-reported disease-related pain during the last week, was measured on a 21-numbered circle VAS, where 0 equals “no pain at all” and 10 equals
“very severe pain”. The VAS pain score is extensively used both in clinical follow-up and in research, both in the form of a continuous 10 cm scale and a 21-numbered scale [15].
PatGA: The patient’s global assessment of overall well-being (PatGA) is a tool where the patient/parent scores the influence of the illness on the patient’s life during the last week on a 21-numbered circle VAS from 0-10, where 0 represents “no influence at all” and 10
“severely influenced”. The PatGA is extensively used both in clinical follow-up of JIA and in research [10, 15].
Morning stiffness: Morning stiffness during the last week was reported by the patients (≥ 9 years old) or parents (for children < 9 years old) in minutes.
CHQ: The Child Health Questionnaire (CHQ) includes 50 not disease-specific questions about topics such as general health, physical functioning, physical pain, mental health, as well as the impact on both the patient, parents, and family in general, and reported for the last 4 weeks [20, 24]. The answers are compiled into a summarized score with a scale from 1 to 100, where a higher score indicates a higher level of functioning and well- being comparable to norm scores (50 +/- 10) from the general U.S. population. The responses are used to calculate a Physical Summary Score (PhS) and a Psychosocial Summary Score (PsS), also with norm scores (50 +/- 10).
2.3.3 Physician-reported outcome measures
PhysGA: The physician’s global assessment of disease activity (PhysGA) was measured on a 21-numbered circle VAS from 0-10 where 0 equals “no activity” and 10 equals “high activity” [15].
17
2.3.4 Composite disease activity measures
Disease status: Inactive disease and remission was defined according to Wallace et al.
[12] with modifications endorsed by ACR [13].
2.3.5 Treatment
The participants’ past and current use of medication, including type of medication, was registered during the study visit. This study focuses on the patients’ past and current use of DMARDs (sDMARDs or bDMARDs). sDMARDs included methotrexate,
hydroxychlorochine, cyclosporine, and mycophenolate mofetil. bDMARDs included etanercept, infliximab, adalimumab, tocilizumab, abatacept, certolizumab, golimumab, and rituximab. Information about the use of NSAIDs, systemic steroids and intraarticular steroid injections was registered at the study visit, but not used in the analyses in this study.
2.4 Statistical analysis
To describe the clinical characteristics and disease activity of the cohort, either mean and standard deviation (SD) or median and 1st to 3rd interquartile ranges (IQR) have been used for continuous variables. Absolute frequencies and percentages have been used for categorical variables.
To evaluate the association of patient-reported physical disability with other patient- and physician-reported outcome measures, we have used categorical table analysis for crude analysis and logistic regression analyses adjusted for age and sex chosen by a priory knowledge, to estimate the adjusted odds ratio (OR) with 95% confidence interval (CI).
The CHAQ scores have been dichotomized into 0 (no disability) or > 0 (disability) and used as the main outcome variable. The other variables have been categorized as follows; VAS pain (0 = no pain, > 0 = pain), PatGA (0 = no disease influence on well- being, > 0 = disease influence on well-being), morning stiffness (< 15 minutes = no significant morning stiffness, ≥ 15 minutes = significant morning stiffness), PhysGA (0 = no disease activity, > 0 = disease activity), and past and present medication (no
DMARDs, DMARDs (synthetic and/or biologic)). We categorized disease status according to Wallace et al. [12, 13] (Rem. off med. = Remission off medication for ≥ 12 months.
Inactive disease = inactive disease on medication < six months or off medication < 12 months, or remission on medication (inactive disease on medication for more than six months). Active disease = flare or continuous active disease). JIA categories, were dichotomized to oligoarticular persistent JIA or not.
To study patient-reported physical health in a Norwegian JIA cohort compared to the matched control cohort, the Physical Summary Score of the Child Health Questionnaire (CHQ PhS) in the JIA cohort was compared to the CHQ PhS in the control cohort. The physical and psycosocial CHQ Summary Scores in the JIA cohort were compared to the CHAQ results. CHQ scores were dichotomized into CHQ ≥ 40 (normal) and < 40
(impaired physical or psychosocial health).
Percent point differences with 95% confidence intervals have been used to assess
categorical data, and the T-test to compare medians for ordinal data. Statistical analyses were carried out using STATA version 16, software (STATA Corp., College Station, Texas, USA).
18
The NorJIA study is approved by The Ethics Committee, Helse Vest (REC no 2012/542).
The study is registered on ClinicalTrials.gov with Identifier: NCT03904459
https://ClinicalTrials.gov/show/NCT03904459. The NorJIA research group has granted access to the data material for this study. Personal data is collected, managed, and stored according to the General Data Protection Regulation (GDPR). Parents of
participants <16 years have given their written informed consent. Younger children have been informed of the study by their parents and the researchers.
3 Ethics
19
4.1 Sample characteristics
Of the 360 children with JIA that were invited to participate in the NorJIA study, 228 accepted, yielding a response rate of 63% (Figure 1). The response rate among
individuals with JIA was 56.1% in Bergen, 68.4% in Trondheim and 67.1% in Tromsø. In five of the 228 participants with JIA included in the NorJIA study, the CHAQ forms were lacking, another two had only answered part of the questionnaire and lacked answers to six and 19 questions, respectively. These seven participants were excluded from the final study population of this study. Three of these were female, four were male. Five of the remaining participants were missing responses to one of the questions, and one of the participants was missing two responses. These were included in the final study
population. This leaves 221 participants with JIA in the final study population.
The response rate among the controls was 224/294 invited controls (76%). Three of the participants with JIA and 17 of the controls did not complete the CHQ. The final inclusion of the CHQ sub-study was therefore 218 participants with JIA and 207 controls.
Figure 1: Flow chart of children and adolescents with juvenile idiopathic arthritis (JIA) and controls in the study
4 Results
20
4.2 Demographics
In the study population, 59.3% of participants with JIA were female (Table 1). Median age was 12.7 (IQR 9.4-14.7) years, 11.8 for girls and 13.5 for boys. Median age in the control group was 12.6 (IQR 9.6-14.9) years. The proportion of parents with higher education was 16.5 percent points (95% CI 7.2 - 25.7) higher in the control group than in the JIA group. The majority, 70.1% in the JIA group and 75.1% in the control group, were classified with normal weight. The proportions of both underweight and obese participants in the JIA group tended to be slightly higher compared to the control group, 2.6 percent points (95% CI -1.6 - 6.9) and 2.2 percent points (95% CI -1.4 - 5.8), respectively. Of the 132 patients with JIA who declined to participate in the study, 58.3%
were female, approximately the same as in the JIA cohort (59.3%). Compared to the mean age in the JIA study population (12.0 years), the mean age among those who declined tended to be lower (10.5 years).
Table 1: Clinical characteristics in the study population
JIA Controls
Na Values Na Values
Females, n (%) 221 131 (59.3) 224 134 (59.8)
Parent with higher educationb, n (%)
Mother 211 136 (64.5) 211 155 (73.5)
Father 208 84 (40.4) 209 121 (57.9)
Caucasian ethnicityc, n (%) 220 214 (97.2) 221 205 (92.8) Iso-BMId, n (%)
Underweight 221 15 (6.8) 217 9 (4.1)
Normal weight 221 155 (70.1) 217 163 (75.1)
Overweight 221 40 (18.1) 217 39 (18.0)
Obesity ≥ 30 221 11 (5.0) 217 6 (2.8)
Age at examination, median years (IQR) 221 12.7 (9.4-14.7) 224 12.6 (9.6-14.9) N =Number, JIA = Juvenile Idiopathic Arthritis, Iso-BMI = Body Mass Index adjusted for age and sex, IQR = Interquartile range,
a N assessed for each variable, excluding missing/unknown values. Total JIA group, N = 221. Total control group, N = 224.
b Parent education was divided into 4 levels: Primary and middle school (7 to 10 years of education), high school (11 to 13 or 14 years), university education less than 5 years, and
university education more than 5 years; the last two levels were grouped together as parents with higher education.
c Ethnicity was divided into two groups: Caucasian and Non-Caucasian according to self-report.
d Calculated using the formula: Weight (kilograms)/[height (metres)]2 and subsequently adjusted for age and sex according to The International Obesity Task Force (IOTF) cut-off values to allow for comparison with adult BMI as defined by Cole TJ, et al. [25, 26].
21
4.3 Disease characteristics in the JIA group
Median age of disease onset in the JIA group was 6.1 (IQR 2.3-10.4) years, and median disease duration was 4.6 (IQR 2.6-8.2) years (Table 2). ANA was positive in 28.1%, and 44.8% had oligoarticular JIA (persistent or extended). At the study visit, 48/221 (21.7%) children were found to have one or more active joints, and 144/221 (65.2%) children received ongoing treatment with DMARDs. Some level of pain was reported by 139/221 (62.9%) of the children, with scores above zero on the VAS pain scale. The physician assessed 75/221 (33.9%) children to have active disease (PhysGA VAS > 0).
22
Table 2: Disease characteristics in the juvenile idiopathic arthritis (JIA) cohort
Na Values
Age at JIA onset, median years (IQR) 221 6.1 (2.3-10.4)
JIA duration, median years (IQR) 221 4.6 (2.6-8.2)
HLA-B27 positiveb, n (%) 221 62 (28.1)
RF positiveb, n (%) 221 6 (2.7)
ANA positiveb, n (%) 199 56 (28.1)
Uveitisc, n (%) 220 28 (12.7)
JIA category, n (%)d
Oligoarticular persistent 221 76 (34.4)
Oligoarticular extended 221 23 (10.4)
Polyarticular RF negative 221 51 (23.1)
Polyarticular RF positive 221 4 (1.8)
Enthesitis-related arthritis 221 22 (10.0)
Psoriasis arthritis 221 9 (4.1)
Systemic 221 7 (3.2)
Undifferentiated 221 29 (13.1)
ESR ≥ 20e, n (%) 218 8 (3.7)
CRP ≥ 5e, n (%) 217 12 (5.5)
Children with active joints > 0e, n (%) 221 48 (21.7) Disease statusf, n (%)
Remission off medication 221 29 (13.1)
Inactive 221 106 (48.0)
Active 221 86 (39.0)
Ongoingg DMARDsh, n (%) 221 144 (65.2)
VAS paine > 0, n (%) 221 139 (62.9)
PhysGA > 0, n (%) 221 75 (33.9)
CHAQ score > 0, n (%) 221 128 (57.9)
N = Number, JIA = Juvenile idiopathic arthritis, IQR = Inter-quartile range, HLA-B27 = Human leukocyte antigen B27, RF = Rheumatoid factor, assessed twice at least 3 months apart, ANA = Anti-nuclear antibody, assessed twice at least 3 months apart using HEp-2 cells, ESR = Erythrocyte sedimentation rate, CRP = C-reactive protein, DMARDs =disease modifying anti-rheumatic drugs, VAS pain = Visual analogue scale for self-reported pain (0 = No pain, 10 = maximal pain), PhysGA
= Physician's global assessment of disease activity (0 = No activity, 10 = high activity), CHAQ = Childhood Health Assessment Questionnaire (0 = no disability, 3 = maximal disability)
a N assessed for each variable, excluding missing/unknown values. Total participants, N = 221.
b Measured around the time the child was diagnosed with JIA.
c Treated for uveitis at any time during the disease course.
d Categories defined according to the ILAR classification criteria [2].
e Assessed as part of the clinical examination at the study visit.
f Disease status according to Wallace et al. [12, 13]: Remission off medication = inactive disease off medication for ≥ 12 months. Inactive disease = inactive disease on medication < six months or off medication < 12 months, or remission on medication (inactive disease on medication for more than six months). Active disease = flare or continuous active disease.
g Ongoing refers to ongoing medication for arthritis or uveitis at the study visit.
h DMARDs include both synthetic DMARDs (methotrexate, hydroxychloroquine, cyclosporine, mycophenolate mofetil) and/or biologic DMARDs (etanercept, infliximab, adalimumab, tocilizumab, abatacept, certolizumab, golimumab, rituximab).
23
4.4 Physical and psychosocial health – CHQ
Mean Physical Summary Score (CHQ PhS) was 10.3 (95% CI 8.7 - 11.9) percent points lower among children with JIA (45.7) than among the controls (56.0) (Table 3). Mean Psychosocial Summary Score (CHQ PsS) were more similar among the two groups with a mean percent point difference of 2.5 (95% CI 1.1 - 3.9). In the JIA group, 24.3%
reported a CHQ PhS score below the cut-off for the norm score (< 40), indicating impaired physical health, compared to 0.5% in the control group, percent point difference 23.8 (95% CI 18.1 - 29.6). The percent point difference for the Psycosocial Summary Score was 6.7 (95% CI 2.6 - 11.0).
Overall, no major difference was found between the CHQ scores of the girls and boys in the JIA group. For the CHQ PsS scores, 7.0% of the girls and 11.2% of the boys scored below 40, percent point difference 4.3 (95% CI -3.6 - 12.2). Likewise, no major
differences were seen according to age.
Table 3: Self-reported physical and psychosocial health assessed by the Childhood Health Questionnaire (CHQ) among participants with juvenile idiopathic arthritis (JIA) and controls
JIA Controls
Sex Age Sex Age
All Fem. Male < 9 years
≥ 9
years All Fem. Male < 9 years
≥ 9 years CHQ PhS
N a 218 129 89 45 173 207 123 84 42 165
Mean (SD)
45.7 (11.0)
45.1 (11.1)
46.5 (10.8)
47.8 (10.1)
45.1 (11.2)
56.0 (4.1)
56.0 (3.7)
55.9 (4.5)
57.2 (3.0)
55.7 (4.2) n <40
(%) 53
(24.3) 31
(24.0) 22
(24.7) 9
(20.0) 44
(25.4) 1 (0.5)
1 (0.8) 0
(0.0) 0
(0.0) 1 (0.6) CHQ PsS
N a 218 129 89 45 173 207 123 84 42 165
Mean (SD)
52.9 (8.1)
53.2 (7.7)
52.4 (8.6)
53.2 (7.6)
52.8 (8.2)
55.4 (6.4)
55.6 (6.2)
55.1 (6.8)
55.0 (4.7)
55.5 (6.8) n <40
(%) 19
(8.7) 9
(7.0) 10
(11.2) 3
(6.7) 16
(9.2) 4
(1.9) 2 (1.6) 2
(2.4) 0
(0.0) 4 (2.4) JIA = Juvenile idiopathic arthritis, Fem. = Female, CHQ = Child Health Questionnaire, N = Number, PhS = Physical Summary Score, SD = Standard deviation, PsS = Psychosocial Summary Score,
a N assessed for each respective variable, excluding missing/unknown values. Total participants, N
= 218. Total controls, N = 207.
24
4.5 Self-reported physical disability according to disease characteristics
More than half (57.9%) of the children with JIA reported disease-related physical disability (CHAQ > 0) (Table 2). The disability was mostly reported with low disability scores between 0 and 1 on a scale from 0-3. None of the participants scored more than two (Figure 2A).
No clear pattern was seen when examining self-reported physical disability (CHAQ) according to neither disease duration nor age at study visit (results not shown).
A
Figure 2: Frequency of Childhood Health Assessment Questionnaire (CHAQ) scores among the juvenile idiopathic arthritis (JIA) cohort of the NorJIA study population (A),the female participants with JIA (B), and the male participants with JIA (C)
B C
25
4.6 Disease characteristics according to sex, age, and disease categories
Some differences were found when looking at disease characteristics according to sex and age (Table 4). The proportion of girls with JIA reporting physical disability (CHAQ
> 0) (64.1%) was 15.2 percent points (95% CI 2.0 - 28.4) higher than the proportion reporting this among boys (48.9%) (Figure 2B and C and Table 4). A similar tendency could be seen for the VAS pain scores, where a 8.6 percent point (95% CI -4.3 - 21.7) higher proportion of girls (66.4%) than boys (57.8%) reported pain (VAS pain > 0) (Table 4). However, sex differences were less evident for physician-reported global assessment of disease activity (PhysGA) and the number of children not in remission off medication. Among participants with JIA below 9 years, whose parents filled out the CHAQ forms, physical disability was reported higher and pain levels were reported lower compared to participants 9 years or older, who filled out the CHAQ forms themselves.
No clear pattern was seen between the different JIA categories and self-reported physical disability, pain, physician-reported global disease activity, and remission status.
Table 4: Physical disability and other disease characteristics according to sex, age, and disease categories in the juvenile idiopathic arthritis (JIA) group
Na
CHAQ score
> 0
VAS pain
> 0
PhysGA
> 0 Not in remissionb Sex, n (%)
Female 131 84 (64.1) 87 (66.4) 47 (35.9) 118 (90.1)
Male 90 44 (48.9) 52 (57.8) 28 (31.1) 74 (82.2)
Age group, n (%)
< 9 years 46 29 (63.0) 22 (47.8) 12 (26.1) 42 (91.3)
≥ 9 years 175 99 (56.6) 117 (66.9) 63 (36.0) 150 (85.7) JIA categoryc
Oligoarticular persistent 76 36 (47.4) 46 (60.5) 23 (30.3) 56 (73.7) Oligoarticular extended 23 16 (69.6) 14 (60.9) 8 (34.8) 22 (95.7) Polyarticular RF negative 51 33 (64.7) 34 (66.7) 20 (39.2) 50 (98.0) Polyarticular RF positive 4 4 (100.0) 4 (100.0) 1 (25.0) 4 (100) Enthesitis-related arthritis 22 12 (54.6) 13 (59.1) 10 (45.5) 20 (90.9) Psoriasis arthritis 9 7 (77.8) 7 (77.8) 2 (22.2) 7 (77.8) Undifferentiated 29 18 (62.1) 18 (62.1) 11 (37.9) 28 (96.6)
Systemic type 7 2 (28.6) 3 (42.9) 0 (0.0) 5 (71.4)
N = Number, CHAQ = Childhood Health Assessment Questionnaire (0 = no disability, 3 = maximal disability), VAS pain = Visual analogue scale for self-reported pain (0 = No pain, 10 = maximal pain), PhysGA = Physician's global assessment of disease activity (0 = No activity, 10 = high activity), RF = Rheumatoid factor
a N assessed for each variable, excluding missing/unknown values. Total participants, N = 221.
b Not in remission = Patients who do not fit the criteria for “Remission off medication” as defined by Wallace et al [12, 13].
c Categories defined according to the ILAR classification criteria. [2]
26
4.7 Association between self-reported physical disability (CHAQ) and other disease outcomes
Self-reported disease-related physical disability (CHAQ scores) was compared to other disease outcome measures, including other patient-reported measures, physician- reported measures, and medication (Table 5). Of children reporting some degree of disease-related pain, 78% also reported physical disability (CHAQ > 0). On the other hand, among those reporting no pain, only 23.2% reported physical disability. Among children with JIA reporting physical health below normal (CHQ PhS score below 40), 94.3% also reported physical disability. When the paediatric rheumatologist reported ongoing disease activity (PhysGA >0), 82.7% of the children reported physical disability.
However, even when the physician reported no disease activity (PhysGA = 0), 45.2% of the children still reported some physical disability.
Regression analysis showed a clear association between physical disability (CHAQ) and the other outcome measures of this study, including both patient- and physician-reported measures. The association was strongest between physical disability and the other
patient-reported outcome measures. The odds of having a CHAQ score above zero (physical disability), when reporting some disease-related pain was high (adjusted OR 12.8, 95% CI 6.5-25.2) compared to those reporting no pain. Similarly, children
reporting that their lives were somewhat negatively affected by their disease (PatGA > 0) had higher odds of reporting physical disability (adjusted OR 12.3, 95% CI 5.8-26.0) compared to those reporting no disease-related influence on well-being. We found a strong association between patient-reported impaired physical health using the generic instrument (CHQ PhS < 40) and physical disability using the disease-specific instrument (CHAQ > 0). Participants reporting physical health below the norm (CHQ PhS < 40) had much higher odds of also reporting physical disability (CHAQ > 0), compared to those with a physical health within the normal range (adjusted OR 21.1, 95% CI 6.2-71.1).
Similarly, those with a mental health below the norm (CHQ PsS < 40) also had higher odds of reporting physical disability compared to those with normal mental health (adjusted OR 8.1, 95% CI 1.8-36.6).
Compared to the patient-reported outcome measures, the association was not equally high for the physician-reported measures. Participants with JIA assessed by the physicians to have some degree of disease activity had higher odds of self-reported physical disability (adjusted OR 5.9, 95% CI 2.9-11.7). An association could also be seen between physical disability and disease status. Compared to participants in remission off medication, the odds of self-reported physical disability were higher among participants with inactive disease (adjusted OR 5.4, 95% CI 1.7-16.7), and much higher among participants with active disease (adjusted OR 32.8, 95% CI 9.7-110.8). Some association was also seen between both treatment with DMARDs and with disease categories other than persistent oligoarticular JIA and physical disability (CHAQ > 0).
27
Table 5: Self-reported physical disability assessed by the Childhood Health Assessment Questionnaire (CHAQ) in the juvenile idiopathic arthritis (JIA) cohort according to disease characteristics at the study visit
Disease
characteristics at
study visit N a
CHAQ score Poor Physical health
= 0 > 0 CHAQ score > 0
n (%) n (%)
OR (95% CI) crude
OR (95% CI) adjustedb VAS painc
= 0 82 63 (76.8) 19 (23.2) 1.0 (ref.) 1.0 (ref.)
> 0 139 30 (21.6) 109 (78.4) 12.0 (6.3-23.1) 12.8 (6.5-25.2) PatGA VASd
= 0 61 50 (82.0) 11 (18.0) 1.0 (ref.) 1.0 (ref.)
> 0 160 43 (26.9) 117 (73.1) 12.4 (5.9-25.9) 12.3 (5.8-26.0) Morning stiffness
<15 minutes 166 85 (51.2) 81 (48.8) 1.0 (ref.) 1.0 (ref.)
≥ 15 minutes 49 7 (14.3) 42 (85.7) 6.3 (2.7-14.8) 6.6 (2.8-15.8) CHQ PhS
≥ 40 165 89 (53.9) 76 (46.1) 1.0 (ref.) 1.0 (ref.)
< 40 53 3 (5.7) 50 (94.3) 19.5 (5.9-65.1) 21.1 (6.2-71.1) CHQ PsS
≥ 40 199 90 (45.2) 109 (54.8) 1.0 (ref.) 1.0 (ref.)
< 40 19 2 (10.5) 17 (89.5) 7.0 (1.6-31.2) 8.1 (1.8-36.6) PhysGAe
= 0 146 80 (54.8) 66 (45.2) 1.0 (ref.) 1.0 (ref.)
> 0 75 13 (17.3) 62 (82.7) 5.8 (2.9-11.4) 5.9 (2.9-11.7) Disease statusf
Rem. off med. 29 25 (86.2) 4 (13.8) 1.0 (ref.) 1.0 (ref.) Inactive 106 55 (51.9) 51 (48.1) 5.8 (1.9-17.8) 5.4 (1.7-16.7) Active 86 13 (15.1) 73 (84.9) 35.1 (10.5-117.6) 32.8 (9.7-110.8) Medication ongoingg
No DMARDsh
ongoing 77 43 (55.8) 34 (44.2) 1.0 (ref) 1.0 (ref)
DMARDs ongoing 144 50 (34.7) 94 (65.3) 2.4 (1.4-4.2) 2.3 (1.3-4.0) Medication everg
No DMARDs ever 55 30 (54.5) 25 (45.5) 1.0 (ref) 1.0 (ref) DMARDs ever 166 63 (38.0) 103 (62.0) 2.0 (1.1-3.6) 1.8 (1.0-3.4) JIA category
Oligoarticular
persistent 76 40 (52.6) 36 (47.4) 1.0 (ref) 1.0 (ref)
Other categoriesj 145 53 (36.6) 92 (63.5) 1.9 (1.1-3.4) 1.9 (1.1-3.4) N = Number, CHAQ = Childhood Health Assessment Questionnaire, OR = Odds ratio, CI =
Confidence interval, VAS = Visual analogue scale, PatGA = Patient Global Assessment of well- being, CHQ = Child Health Questionnaire, PhS = Physical Summary Score, PsS = Psychosocial Summary Score, PhysGA = Physician's global assessment of disease activity, Rem off med. = Remission off medication, DMARDs = disease modifying anti-rheumatic drugs, JIA = Juvenile idiopathic arthritis
a N assessed for each respective variable, excluding missing/unknown values. Total participants, N
= 221. Total controls, N = 224.
b Adjusted for age, sex, and Iso-BMI.
c Self-reported disease-related pain measured on a 21-numbered circle VAS pain (0 = no pain, 10
= maximum pain).
d Assessed by the parent/patient on a 21 numbered VAS (0 = very well, 10 = very poor).
28
e Assessed by the physician on a 21 numbered VAS (0 = inactive, 10 = maximal activity).
f Disease status according to Wallace et al. [12, 13]: Rem. off med. = Remission off medication for
≥ 12 months. Inactive disease = inactive disease on medication < 6 months or off medication < 12 months, or remission on medication (inactive disease on medication for more than 6 months).
Active disease = flare or continuous active disease.
g Ongoing refers to ongoing medication for arthritis or uveitis at the study visit. Ever refers to medication ever used for arthritis or uveitis during the disease course, including ongoing medication. Registered medication was divided into No DMARDs and DMARDs.
h DMARDs includes both synthetic (methotrexate, hydroxychlorochin, cyclosporine, mycophenolate mofetil) and/or biologic DMARDs (etanercept, infliximab, adalimumab, tocilizumab, abatacept, certolizumab, golimumab, rituximab).
i Categories defined according to the ILAR classification criteria [2].
j All participants not categorized as “Oligoarticular persistent”.
29
5.1 Main findings
A majority (57.9%) of children with JIA reported physical disability, and children with JIA reported significantly poorer physical health (CHQ PhS) and also a minor reduction in psychosocial health (CHQ PsS) compared to sex- and age-matched controls. A very strong association was found between patient-reported physical disability (CHAQ > 0) in JIA and other patient-reported outcome measures (PROMs). The odds of reporting physical disability when reporting disease-related pain (VAS > 0) was high. The same association could be observed between physical disability and disease-related influence on well-being (PatGA). There was also a strong association between physical disability scored with the disease-specific instrument CHAQ and poor general health scored by the generic instrument CHQ; participants reporting CHAQ > 0 had higher odds of reporting a psychosocial health below the norm and much higher odds of reporting a physical health below the norm. Although somewhat lower than the association with the other PROMs, there was also a clear association between self-reported physical health and physician- reported disease activity, as well as disease status, especially between reporting physical disability and having active disease.
5.2 Strengths
Our study is based on the NorJIA study, which included a large number of participants with JIA in all age groups from four to 16 years, in addition to a matched control cohort.
The study also included children diagnosed with all the JIA categories, indicating that the study had a good representation of children with JIA in Norway. Another strength is that the participants were recruited from three different locations, representing three different parts of the country. Validated instruments were used for both patient- and physician- reported outcomes, and clinical examinations were performed by experienced
paediatricians following a standardized study protocol. Examining this within the
Norwegian JIA population may be considered especially relevant, since Norway has one of the highest JIA incidence rates in the world. In addition, the organisation of our public health system with free clinical consultations for children, and a stable and relative homogenous population, gives us opportunities to perform high-quality population-based cohort studies.
5.3 Limitations
There are however some limitations. Even with a relatively high response rate of 63%, possible selection bias must be considered. The mean age was 1.5 years lower among those who declined participation than in the study population. Possible explanations for this may be that parents of the youngest children with JIA were more hesitant of letting their child participate in a study consisting of two days of clinical examinations including an MRI examination which might be difficult for the youngest children. However, no clear association could be seen between age at clinical examination and self-reported physical disability (CHAQ > 0). Furthermore, we cannot exclude a selection bias towards
recruiting more children with active disease. One can imagine that children in remission
5 Discussion
30
are less often seen at the hospital and thus not asked to participate in the study. They might also be more hesitant of going through the hospital stay required for the study examinations. Another possible limitation is that some of the PROMs, like CHQ, are filled out by the parents and others by the children themselves. This makes comparing the scores from the different PROMs challenging. In addition, the CHAQ is supposed to be filled out by the parents if the child is under the age of nine and by the children themselves if they are 9 years or older. However, we cannot rule out that the parents may have affected the scoring of the children over 9 years old if they were sitting together when fulfilling the questionnaires. This is a well-known issue with the CHAQ;
parents experience the disease of their child from a different perspective than the child itself [22]. Likewise, parents’ objectivity on behalf of their children compared to
physician’s judgement, have been questioned [23]. Furthermore, many children scored zero on CHAQ, and we cannot exclude that some of these children felt quite well, but not without any disability, and were grouped together with children with full ability, due to the floor effect of the instrument [27]. Finally, the low number of children in some JIA categories was too small to allow meaningful comparison.
5.4 Comparison with other studies
The importance of patient-reported outocomes is highly recognised in JIA research and also for use in clinical settings [11, 17, 18, 28]. The importance of PROMs is also being recognised in other fields than JIA, and as a general concept to prioritize in the future of health care [29, 30]. In their realist synthesis, Greenhalgh et al. emphasized that PROMs stimulate patients to reflect more about their disease and how it affects their lives [29].
The PROMs indicates for the patients that the physician value their thoughts and opinions, and provide patients with a framework for bringing up matters with their
physician. Several specialists have studied and developed PROMS for use in research and clinical settings, but to our knowledge there are few studies on the association between the different PROMs and between PROMs and physician-reported outcome measures.
Compared to our results with a majority of the children scoring > 0 on the CHAQ (58%), other follow-up studies have generally found lower numbers. The closest was a Canadian study from 2002, which found that 55% scored CHAQ >0 [31]. In long-term follow-up, the percentages with disability have been lower. A Danish long-term follow-up study from 2000 carried out by Zak and Pedersen, and a Swedish longitudinal 17-year prospective study by Bertilsson et al. both found 46% with the adult version of CHAQ, the Health Assessment Questionnaire (HAQ) > 0 [32, 33]. In 2003, Flatø et al. found HAQ > 0 in 36% in a Norwegian study after a median disease duration of 14.9 years [34]. In a Nordic study from 2011 conducted by Nordal et al., 32% of the participants had CHAQ/HAQ at eight-year follow-up study [35]. Glerup et al.’s 18-year follow-up study of the same cohort revealed a similar proportion: 28% [8].
Comparison between studies is always hampered by different study design and different study populations. Overall, most of the other studies we have compared our findings with had longer JIA durations [8, 31-36]. This is the most apparent explanation for our
physical disability numbers being higher than in the other studies. The participants in our study were younger and some of the children in our study had recently been diagnosed with JIA. Our median JIA duration was 4.6 years. In comparison, Nordal’s eight-year follow-up had median JIA duration of 8.1 years, and Glerup’s 17.5. In the Danish study, the mean duration was 26.4 years [32]. During the first years of disease activity, more patients will have active disease or flares until adequate medication is found.
31
Furthermore, patients with a shorter disease duration could be more prone to focus on the symptoms, as they are not so familiar with their disease. A Norwegian qualitative interview study from 2009 indicated that participants gradually adjusted psychosocially, finding coping mechanisms and adapted their perspectives to tackle the disease burden that JIA entails [37]. The findings of Zak and Pedersen also support JIA duration as an explanation for the differences in physical disability result [32]. Zak and Pedersen found that there was a strong correlation between HAQ score and disease duration. Minden on the other hand found no statistically significant correlation between HAQ score and disease duration. Another apparent difference between our study and the other studies was the proportion of patients in remission off medication. Whereas in our study, only 13% of the participants were in remission off medication, the number ranged from 33%
to 63% in the other studies, supporting the impression that our study group had more active disease. The participants in many of these studies were older than 18 years, hence the use of HAQ and not CHAQ [37].
In our study, we found that the proportion of participants reporting physical disability (CHAQ > 0) was different in the different JIA categories. The same could be seen in Nordal’s and Glerup’s studies [8, 35].
Nordal’s CHQ scores were more similar to our findings. The proportion of participants with CHQ physical < 40 was 24% versus 19% in our study. The equivalent proportion for CHQ psychosocial < 40 were 8,7 in both our and Nordal’s participants [35].
Furthermore, our results showed that pain level had a strong association with physical disability. In their follow up study from 2002, Minden et al. found that pain score in young adults with JIA had a statistically significant correlation with HAQ score.
We also found an association between global well-being reported by the patient (PatGA) and physical disability which can be compared to the result found by Minden et al in 2002 [36]. Rather than using the PatGa form, they measured the overall well-being of
participants with JIA on a numeric rating scale (NRS) from 1-10. They found statistically significant correlation between overall wellbeing and HAQ score.
Our results also showed a strong association between the physician-reported global assessment (PhysGa) of disease activity and patient-reported physical disability (CHAQ).
Other studies have also found association between assessment by physicians and self- reported physical functioning: Both Zak and Pedersen in 2000 and Minden et al. in 2002 found that Steinbrocker functional classes correlated with HAQ [32, 36]. Steinbrocker functional class is given by the physician evaluating the capacity of the patient to perform acquired activities for a daily life [36].
The long-term follow-up studies of Minden et al., Zak and Pedersen, and Bertilson et al., all found correlations with disease activity and increased HAQ score [32, 33, 36]. This strengthens our results demonstrating that active disease had the strongest association to physical disability of all the variables we studied (adjusted OR 32.8, 95%, CI 9.7- 110.8). It also strengthens the impression that disease activity, and perhaps not so much permanent damage, accounts for a large part of the patient’s evaluation of his/hers disease-related physical disabiity.
5.5 Clinical implications
The results of this study clearly support the notion that patient-reported physical health data are an important supplement to the physician’s evaluation of the patient, providing a more complete picture of the effects the disease have on the patient [11, 18]. Despite
32
being a subjective measure, there is an evident association between the CHAQ and other instruments, including both other patient-reported tools, and more objective clinical activity measures. Bringing forth the patient’s viewpoints in a written, systematic manner provides unique and relevant information beyond what can be observed and measured by the physician. Recently, increasing attention has been given to the patient’s perspective.
The use of PROMs is a direct way to respond to this endeavour and thereby improving patient care [11]. When aiming to achieve individualized therapy the traditional general clinical instruments and descriptions are insufficient. More details about the patient’s view are needed: Does the disease impact the patient’s function? If so, what does the patient view as their main challenges? What are their main concerns? PROMS such as CHAQ are a validated, systematic way to collect such information, and are therefore highly relevant in both a research setting, and as part of the patient’s clinical follow-up.
Some studies, performed both in JIA research and in other fields, have shown that patients prefer being presented digital versions of PROMs [38-40]. This may imply that digital PROMs ought to be part of physicians’ relationships with their patients in the future.This development might have been facilitated during the covid-19 pandemic where telemedicine and video consultations have necessitated digital communication with the patients.
33
Our results show that patient-reported physical disability in JIA remains high and much higher than in children without JIA, and that there was a strong correlation between patient-reported physical disability and other patient-reported health data. There was also a strong association between patients
