Article Text

Download PDFPDF

Extended report
The patient perspective on absence of disease activity in rheumatoid arthritis: a survey to identify key domains of patient-perceived remission
  1. Lilian H D van Tuyl1,
  2. Martina Sadlonova2,
  3. Sarah Hewlett3,
  4. Bev Davis4,
  5. Caroline Flurey3,
  6. Niti Goel5,
  7. Laure Gossec6,
  8. Cecilie Heegaard Brahe7,
  9. Catherine L Hill8,
  10. Wijnanda Hoogland9,
  11. John Kirwan10,
  12. Merete L Hetland7,
  13. Dirkjan van Schaardenburg1,
  14. Josef S Smolen2,
  15. Tanja Stamm2,
  16. Marieke Voshaar9,
  17. George A Wells11,
  18. Maarten Boers1,12
  1. 1Amsterdam Rheumatology & Immunology Center, Amsterdam, The Netherlands
  2. 2Medical University of Vienna, Vienna, Austria
  3. 3University of the West of England, Bristol, UK
  4. 4Patient Partner, Bristol, UK
  5. 5Quintiles and Duke University School of Medicine, Durham, North Carolina, USA
  6. 6Department of Rheumatology, Sorbonne Universités, UPMC Univ Paris 06; AP-HP, Pitié Salpêtrière Hospital, Paris, France
  7. 7Copenhagen Center for Arthritis Research and DANBIO, Center for Rheumatology and Spine Diseases, Rigshospitalet, Glostrup, Denmark
  8. 8University of Adelaide, Adelaide, South Australia, Australia
  9. 9Patient Partner, Amsterdam, The Netherlands
  10. 10University of Bristol, Bristol, UK
  11. 11University of Ottawa, Ottawa, Canada
  12. 12Department of Epidemiology & Biostatistics, VUmc, Amsterdam, The Netherlands
  1. Correspondence to Dr Lilian H D van Tuyl, Department of Rheumatology, VU University Medical Center, 3A56, Amsterdam 1007 MB, The Netherlands; L.vantuyl{at}


Background Guidelines suggest treatment in rheumatoid arthritis (RA) to target remission, in close consultation with the patient. Our recent qualitative study of the patients' perspective on remission in RA identified 26 domains. The current study aimed to identify a short list of the most important aspects to inform future research.

Methods Patients with RA from the Netherlands, the UK, Austria, Denmark, France and the USA completed a survey that contained all domains identified in our qualitative study. They rated domains for importance (‘not important’, ‘important’ or ‘essential’ to characterise a period of remission) and if important or essential, whether this domain needs to be ‘less’, ‘almost gone’ or ‘gone’ to reflect remission. Respondents were also asked to determine their personal top 3 most important/essential domains. Frequency of specific domains in the top 3 was calculated, and domains were sorted on the percentage of patients that evaluated a particular domain as ‘essential’.

Results Of 274 respondents, 75% were female, mean (SD) age 57(13) years, disease duration 12(9) years. The top 3 were as follows: pain (67%), fatigue (33%) and independence (19%); domains most frequently rated as ‘essential’ were as follows: pain (60%), being mobile (52%), physical function (51%), being independent (47%) and fatigue (41%). Pain needed to be less (13%), almost gone (42%) or gone (45%) to reflect remission. Similar patterns were seen for fatigue, independence, mobility and physical functioning.

Conclusion Patients identified pain, fatigue and independence as the most important domains of RA disease activity that need to be improved to reflect remission.

  • Rheumatoid Arthritis
  • Patient perspective
  • Outcomes research

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Patients with rheumatoid arthritis (RA) are at risk of severe bone and cartilage damage in affected joints, causing chronic pain, fatigue and other extra-articular manifestations with a significant impact on daily life. The degree of disease activity and response to treatment are traditionally determined by evaluation of the RA core set or indices derived thereof.1 ,2 The core set contains the patient-reported outcomes (PROs) physical function, pain and global assessment of disease activity. Although not consistently associated with joint damage and differential response in disease stages, these PROs have been found to be at least as relevant as more ‘objective’ physical and biochemical measures in assessing baseline disease status, improvement during interventions or prediction of long-term outcome.3 ,4 Moreover, the relevance of some of them, especially those evaluating physical function, is revealed by observations that increasing joint damage causes increasing irreversibility of functional impairment, even if clinical activity has subsided into remission.5 ,6

It is becoming increasingly clear both inside and outside rheumatology that patients are crucial partners in obtaining relevant information, adding unique skills, values and experiences to research.7 Patients have identified domains such as fatigue and sleep quality to be important and thus core areas for measurement. Subsequent research has shown measurement of fatigue, one of the most important problems identified by patients with RA, to be highly reliable, sensitive to change and an independent determinant of disease activity.8–10 As a consequence, the scientific community now recognises fatigue as a core PRO to be measured in all RA clinical trials.11 ,12 Other products of close cooperation between patients and professionals include the recent development of patient-derived scores to capture the impact of RA and psoriatic arthritis on daily life.13 ,14

In the last decade, the development of new drugs for the treatment of RA has made a state of minimal disease activity and even remission an attainable goal in most patients.15–18 Because treatments are increasingly targeted at achieving remission, a good definition of remission is vital. In 2011, the three leading international rheumatology organisations, that is, the American College of Rheumatology (ACR), the European League Against Rheumatism (EULAR) and the Outcome Measures in Rheumatology group (OMERACT), led the initiative which redefined remission in RA.19–21 To this end, all important prognostic factors and outcome measures available in clinical trial data were evaluated for their potential use in defining remission. However, this included only the three core set PROs patient global, pain and physical function, as data on other potential important aspects of remission from the patient perspective were not available.

In response, both patients and professionals identified the need to study the concept of remission from a patient perspective, to evaluate whether additional domains (and PROs) could optimise targeted therapy.22 Therefore, we recently undertook a qualitative study to understand the patient perspective on remission in RA.23 Three major themes of patient-perceived remission emerged:

1. Symptoms such as pain, stiffness and fatigue would either be absent or be reduced in intensity.

2. The impact of the disease on daily life would diminish as shown by increased independence, the ability to do valued activities, improved mood and the ability to cope.

3. Remission would lead to a return to normality, including the ability to work, enjoy one's family role and be seen as normal by other people. Patients felt that the concept of remission was influenced by ageing, side effects of medication, comorbidities, accrued damage to joints and disease duration. This qualitative research identified many domains of interest to patients, but did not indicate the importance of one domain over another.

The aim of the current descriptive study was to determine the importance of specific symptoms, aspects of disease impact and normality in defining remission in RA from the patient perspective through a survey, to complete the information necessary for optimal clinical management.



Patients >18 years of age with a confirmed diagnosis of RA (2010 criteria)24 receiving usual care in one of five centres (VU University Medical Center/Reade in Amsterdam, the Netherlands; Medical University of Vienna, Austria; Bristol Royal Infirmary in Bristol, UK; Center for Rheumatology and Spine Diseases, Rigshospitalet, Glostrup, Denmark; Université Pierre et Marie Curie and Hopital Pitie-Salpetriere Paris in Paris) were invited to participate in this study. Medical ethical committees in these centres approved the study protocol where applicable and patients gave their informed consent before participation. In addition, a fixed sample of 50 patients with RA was recruited through a known community ( with pre-existing consent to contact for research purposes in the USA.25

Eligible patients had to speak, read and write the local language sufficiently to understand the study and complete the survey (physician's judgement where applicable).

In addition, all patient representatives who attended OMERACT 12 in 2014 were invited to participate by email,26 as well as all the patients who participated in one of the nine focus group discussions prior to this study.23

Data collection process

In Bristol, Amsterdam, Vienna and Copenhagen, the surveys and reply envelopes were distributed in the clinic; in Paris and the USA, they were distributed by email.

As the word ‘remission’ is a common word in the English language and might imply certain presumptions, this term was not used during the recruitment and data collection phase. Instead, remission was formulated as ‘disease activity as good as gone’. Where available in routine practice, a recent 28-joint count, physician global assessment and acute phase reactant were collected from the hospital files, within a period of 3 months before or after completion of the survey.

The survey

The goal of the survey was to determine a short list of the most important items that reflect remission according to patients with RA. The survey (see online supplementary file I) contained all 26 domains of remission that were identified in previous focus group discussions in Bristol, Vienna and Amsterdam,23 formulated as items which patients were asked to rate for importance. In addition, patients were asked to add any missing aspects of remission in free text fields. During the qualitative study, patients had indicated that demographic and disease-specific aspects were important to interpret the data, and so information on age, gender, disease duration, comorbidities and accrued joint damage was collected, all in a self-reported manner using the Routine Assessment of Patient Index Data 3 (RAPID3), with remission cut-off defined as 3 or less on a scale of 0 to 30.27 Where possible, the Clinical Disease Activity Index (CDAI) was calculated, with remission cut-off ≤2.8.28

German, French, Danish and Dutch versions of the survey were prepared by translation and back translation by the research team to verify the terminology. The language used was carefully written, based on the focus group terminology, and was reviewed by patient research partners (WH, MV and BD) to ensure that the instructions were clear to patients and that each item was understandable in terms of the RA symptoms and experience.

To reduce any order effect on decision making, two versions of the survey were distributed in the clinics (but not for the emailed assessments) with the domains and the items within them randomly ordered.


Patient characteristics were summarised as mean (SD) or median (IQR) where applicable. Fulfilment of RAPID3 remission, ACR/EULAR remission and CDAI remission was compared with the patients' self-reported judgement of remission (‘disease as good as gone’: no/yes) to determine concordance between the clinical definition of remission and the patient's judgement of remission, quantified using the κ measure for agreement (with 0.75 as excellent, 0.40–0.75 as fair to good and below 0.40 as poor); 2 by 2 tables and χ2 tests were used.

To determine the importance of domains, first, frequency of a particular domain mentioned in the top 3 was calculated. Second, domains that >30% of patients identified as ‘not important’ were removed. The remaining domains were sorted on the percentage of patients that evaluated a particular domain as ‘essential’.

To evaluate robustness of the results, data were stratified by the influential factors as identified by patients in the qualitative study, including self-reported age (above or below 50 years), gender (male/female), disease duration (more or less than 2 years), comorbidity (no/yes) and accrued joint damage (no/yes), to see if these factors influence the patient perspective on remission. In addition, data were stratified by country and location of filling out the survey (clinic visit or by email). χ2 tests were used to determine statistical significance (if p<0.05) where relevant.


A total of 274 patients completed the questionnaire: 54 from the Netherlands, 33 from the UK, 51 from Austria, 43 from Denmark, 43 from France and 50 from the USA. Response rate in the Netherlands and France was 59% and 42%, respectively.

The population was typical for RA (table 1), with 75% females, mean (SD) age of 57 (13) years, disease duration of 12 (9) years with 10% disease duration <2 years and self-reported erosive disease of 41%. Remission as reported by the patient was present in 38% and according to the RAPID3 in 30%. Concordance was reasonably good: of the patients in self-reported remission, 61% were in RAPID3 remission; of the patients in RAPID3 remission, 79% were in self-reported remission (observed agreement, 79%; κ 0.54).

Table 1

Patient characteristics by site, all self-reported

In the subgroup of patients with available clinical data (n=119), 42% were both in self-reported and in RAPID3 remission, and 24% in ACR/EULAR remission. Here concordance was good only in one direction: of patients in ACR/EULAR remission 86% were in self-reported remission (observed agreement, 76%; κ 0.48) and 97% were in RAPID3 remission (observed agreement, 81%; κ 0.59); but only 50% of patients in self-reported remission and 56% of patients in RAPID3 remission were in ACR/EULAR remission (table 2). Of the patients with available CDAI (n=47), CDAI remission was present in 21%. Of these, all patients were in RAPID3 remission, 80% was in self-perceived remission (observed agreement, 70%; κ 0.35) and 70% in ACR/EULAR remission (observed agreement, 81%; κ 0.48).

Table 2

Concordance between two patient-derived remission definitions and the ACR/EULAR remission definition (count (percentage))

Patients that considered themselves in remission (n=103) had a mean patient global assessment of disease activity of 1.5 (1.5) and disease duration of 10 (7) years) in contrast to 4.8(2.5) and 13(10) years for those not in self-perceived remission (n=171).

Most important domains

The most often-mentioned domains in the top 3 were as follows: pain (67%), fatigue (33%) and independence (19%) (table 3).

Table 3

Most important domains by site

Only one domain, ‘The way other people see me’, was regarded by more than 30% of patients (59%) as ‘not important’ and was removed from further analyses. The percentage of patients that choose a certain domain in their top 3 is shown in online supplementary file II. Domains that were most frequently rated as ‘essential’ to characterise a period of remission were highly similar: pain (60%), being mobile (52%), physical function (51%), being independent (47%) and fatigue (41%) (table 4).

Table 4

Percentage of patients that rate a domain as essential to be reduced or absent/improved to define remission

Pain needed to be less (13%), almost gone (42%) or gone (45%) to reflect remission. Similar patterns were seen for fatigue (23%, 40%, 37%). Independence needed to be better (16%), almost normal (31%) or normal (53%), with similar patterns for mobility (16%, 35%, 49%) and physical functioning (14%, 29%, 57%).



Importance of pain and fatigue were similar in patients under and over 50 years of age; however, independence was reported more frequently in the top 3 by patients over 50 (24%) as compared with patients under 50 years of age (12%) (p=0.03).

Disease duration

Importance of pain and fatigue were similar in patients with early (≤2 years) versus long-standing disease; numerically, independence was reported more frequently in the top 3 by patients with long-standing disease (21%) as compared with patients with early disease (12%) (p=0.40).


There were slight differences in the choice of a top 3 in different countries (table 3), although pain was the number 1 domain in all countries. Pain was mentioned in the top 3 by 65%, 69%, 52%, 74%, 92% and 79% in the UK, Austria, the Netherlands, France, the USA and Denmark, respectively (χ2 p=0.001). Likewise, fatigue was among the top 3 (48%, 16%, 30%, 51%, 40%, 29%, 35% (χ2 p=0.005)) and to a lesser extent independence (39%, 20%, 27%, 21%, 6%, 17% and 21% (χ2 p=0.01)). Bristol, Amsterdam and Paris had the same top 3 as that of the total group, but in Vienna, fatigue was replaced by stiffness and in both the USA and Copenhagen, independence was replaced by swelling.

Domains that were most frequently rated as ‘essential’ to characterise a period of remission were the same in the total group and in Bristol, while in Vienna fatigue was replaced by mental power, in Amsterdam fatigue was replaced by activities of daily living, in Paris fatigue was replaced by family role, in the USA physical function was replaced by mobility and in Copenhagen fatigue and independence were replaced by work and activities of daily living.

Version of the survey

The majority completed V.1 of the survey (79%). Patients that completed V.1 of the survey (appendix 1) reported pain in their top 3 more frequently (77%) compared with patients that completed V.2 of the survey (54%) (p=0.02). A similar pattern, although not significantly different, was seen for independence, reported by 19% in V.1 and 29% in V.2. However, of the patients that completed V.2 of the survey, the most often-mentioned domains in the top 3 were still pain (55%), fatigue (33%) and independence (29%).


There were no differences in choice of a top 3 between male and female patients; between patients that reported joints with versus without strongly reduced mobility, deformities or joint replacement surgery; between patients that reported to have been diagnosed with other diseases; or between patients that completed the survey during their clinic visit, compared with patients that completed the survey electronically (at home).

Table 4 gives an overview of the numerical differences of domains rated as essential for all stratifications.


This survey study identified the three most important domains of patient perceived remission, based on preceding qualitative research on the patient perspective on remission in RA: the absence or reduction of pain and fatigue and the improvement or maintenance of independence.

Pain is the most predominant PRO assessed in rheumatic diseases, present in all core sets and frequently inquired after in clinical practice.29 Fatigue has been acknowledged as an essential PRO in recent years and is recommended to be reported in RA clinical trials;11 however, independence is not a common PRO in rheumatology. We are not the first to report on independence as an important domain for patients with RA; a recent Czech study30 reported a significant difference between the healthy population and patients with RA in level of independence. In addition, qualitative work performed in Bristol generated the Rheumatoid Arthritis Patient Priorities for Pharmacological Interventions core set to complement the existing professional core sets.31 This patient-driven study also identified independence as one of eight priority outcomes of pharmacological interventions in RA. In a Swedish qualitative study evaluating the patient perspective on benefits of RA treatment, independence was identified as a theme covering domains as management of daily activities, care for oneself and for one's family, being able to work and enjoy leisure time.32

In contrast, in the context of our remission work, independence emerged as a separate domain, grouped in the theme ‘decreased impact of RA’. When reviewing our qualitative work, the domain independence seems mainly related to physical functioning, that is, ‘the ability to do things you have to do and not have to ask others to do things for you’. However, influence of other domains and factors like ability to work, performing one's family role and comorbidities is highly likely and warrants further study. Similar overlap might exist between other domains while this could have been avoided by grouping of domains into larger themes. Indeed, if we would have grouped the domains socialise, family role, work and leisure into one domain ‘participation’, the ratings for this grouped domain would have exceeded that of independence in two out of six sites. However, we have chosen to present all 26 domains in our survey, using patient quotes from focus groups, so that patients could optimally relate to the domains. In the case of independence, this domain was chosen more often as a top 3 domain than related domains such as physical functioning, mobility, work or activities of daily living. However, these related domains scored high in the rating of essential domains, with mobility ranked second, physical function third, independence fourth and fatigue fifth. As our objective was to identify the three most important domains for patients with RA to define a state of remission, we have chosen to rely on the report of top 3 domains, knowing that the ranking of essential domains pointed in a highly similar direction and that independence covers aspects of function/mobility.

Some differences exist between different patient groups when stratified for age, gender, disease duration and the presence of comorbidities and damage; for example, performing one's family role was more important to younger patients compared with older patients. These differences were already predicted by patients during the qualitative phase and are confirmed by this survey study.

In addition, some differences exist in the procedure of filling out the survey; we sent out two versions of the survey with a different (random) order of the domains, to reduce any order effect on decision making. Although there was a difference in frequencies of domains reported in the top 3 between patients that completed V.1 and those that completed V.2 of the survey, the three most often reported domains were still pain, fatigue and independence. It seemed that the higher the domain was listed in the survey, the more likely that patients would include the domain in their top 3, even though the top 3 was a separate section at the end of the survey. However, as the top 3 remained unchanged across the different versions, this order effect had no influence on the patients' choice for the three most important domains.

It can be questioned whether other domains besides the 26 identified from focus groups in Bristol, Amsterdam and Vienna would have emerged when groups were organised in Paris, Copenhagen and the USA; the survey allowed patients to add additional domains. Seven patients used this opportunity, but this did not result in new domains, suggesting that the survey was comprehensive. Yet, this does not indicate generalisability of our results to countries beside those studied here.

The majority of patients indicated that they needed their symptoms to be ‘gone’ instead of ‘less’ or ‘almost absent’ to reflect remission. Yet, patients that indicated themselves as currently in remission had a mean patient global assessment of 1.5. It remains a point of discussion how this relates to the ACR/EULAR remission definition, which requires that the patient global assessment of disease activity can be no >1. Interpreting these data, one could argue that a score of <2 on a visual analogue scale from 0 to 10 equals ‘gone’ from a patient perspective. Alternatively, to adhere to the rule of 1, one could allow rounding to one for scores that go up to 1.5. Interesting is the contrast between patients' rating of fatigue as a top 3 priority, yet needing it to be less (23%), almost gone (40%) or gone (37%) to reflect remission. This is different for the other four important domains, in which the majority feels a certain domain needs to be gone, rather than almost gone or less. Perhaps fatigue is a domain that seriously impacts the lives of people with RA, but is somehow more manageable or acceptable than for example pain.

Strengths of this study include the geographical spread of the patient sample as well as the dual analysis of domains, either within the top 3 of importance or rated most frequently as essential. The robustness of the results across countries and response modes enhances the reliability of the results.

A weakness of this study is that the response rate was not registered at every site, which might indicate that a selection of highly motivated patients was studied. When looking at the demographic characteristics, a heterogeneous pattern is observed, with a wide range in patients’ age, disease duration and remission experience. The size of our study population does not allow for extensive subgroup analysis. Yet the uniformity in a choice for a top 3 of patient important domains strengthens the importance of these three domains and suggests that these are—to a certain level—independent of other patient characteristics.

Another weakness of this study is the absence of extensive clinical data collection, which prohibited us from linking our results with other frequently used disease activity measures like the Clinical and Simplified Disease Activity Indices (except in a few patients). However, this work served primarily to identify the common top 3 of patient domains that need to be studied further in comparison with clinical measures.

With this in mind, a validation study has been initiated, studying measurement instruments for the three most important domains of patient perceived remission identified in this study in relation to clinical outcome. It is anticipated that the results of the validation study will inform whether any of these three domains add important information to the ACR/EULAR remission criteria.

In summary, this survey study identified the three most important domains of patient perceived remission, based on preceding qualitative research on the patient perspective on remission in RA. Follow-up research has been initiated to identify valid measurement instruments for these domains and quantify the contribution to the ACR/EULAR remission criteria.


The authors thank the OMERACT ‘Patient perspective on remission in rheumatoid arthritis’—working group for their participation in this work.



  • Handling editor Tore K Kvien

  • Twitter Follow Caroline Flurey at @C_Flurey

  • Funding This work was funded by a grant from EULAR.

  • Competing interests None.

  • Ethics approval Medical Ethical Committees of all involved study sites: VU University Medical Center/Reade in Amsterdam, the Netherlands; Medical University of Vienna, Austria; Bristol Royal Infirmary in Bristol, United Kingdom (UK); Center for Rheumatology and Spine Diseases, Rigshospitalet, Glostrup, Denmark; Université Pierre et Marie Curie and Hopital Pitie-Salpetriere Paris in Paris.

  • Provenance and peer review Not commissioned; externally peer reviewed.