Article Text

Five-year treat-to-target outcomes after methotrexate induction therapy with or without other csDMARDs and temporary glucocorticoids for rheumatoid arthritis in the CareRA trial
  1. Veerle Stouten1,
  2. René Westhovens1,2,
  3. Sofia Pazmino1,
  4. Diederik De Cock1,
  5. Kristien Van der Elst2,
  6. Johan Joly2,
  7. Delphine Bertrand1,
  8. Patrick Verschueren1,2
  1. 1 Skeletal Biology and Engineering Research Centre, Department of Development and Regeneration, KU Leuven, Leuven, Belgium
  2. 2 Rheumatology, University Hospitals Leuven, Leuven, Belgium
  1. Correspondence to Patrick Verschueren, Skeletal Biology and Engineering Research Centre, KU Leuven, Leuven, Belgium; Patrick.verschueren{at}


Objectives To compare outcomes of different treatment schedules from the care in early rheumatoid arthritis (CareRA) trial over 5 years.

Methods Patients with RA completing the 2-year CareRA randomised controlled trial were eligible for the 3-year observational CareRA-plus study. 5-year outcomes after randomisation to initial methotrexate (MTX) monotherapy with glucocorticoid bridging (COBRA-Slim) were compared with MTX step-up without glucocorticoids or conventional synthetic disease-modifying antirheumatic drug (DMARD) combinations with glucocorticoid bridging, per prognostic patient group. Disease activity (Disease Activity Score based on 28 joints calculated with C reactive protein (DAS28-CRP)) and functionality (Health Assessment Questionnaire (HAQ)) were compared between treatment arms using longitudinal models; safety and drug use were detailed.

Results Of 322 eligible patients, 252 (78%) entered CareRA-plus, of which 203 (81%) completed the study. Treatments for high-risk patients resulted in comparable DAS28-CRP (p=0.539) and HAQ scores over 5 years (p=0.374). Low-risk patients starting COBRA-Slim had lower DAS28-CRP (p<0.001) and HAQ scores (p=0.041) than those starting only on MTX. At study completion, 114/203 (56%) patients never had their original DMARD therapy intensified, with comparable rates between all treatments. Safety was comparable between treatments in high-risk patients. In low-risk patients, there were 18 adverse events in 10 COBRA-Slim and 36 in 17 patients treated with initial MTX monotherapy (p=0.048). Over 5 years, 22% of patients initiated biologics, 25% took glucocorticoids for >3 months and 17% for >6 months outside the bridging period.

Conclusions All intensive treatments with glucocorticoids bridging demonstrated excellent 5 year outcomes. Initiating COBRA-Slim was comparably effective as more complex treatments for high-risk patients with early RA and more effective than initial MTX monotherapy for low-risk patients with limited need for biologics and chronic glucocorticoid use.

  • glucocorticoids
  • arthritis
  • rheumatoid
  • biological therapy
  • methotrexate
  • outcome assessment
  • health care

Data availability statement

Data are available upon reasonable request. The authors commit to making the relevant anonymised patient level data available for a specified purpose approved by the institution and the principal investigator of the CareRA plus study and with a signed data access agreement.

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.

Key messages

What is already known about this subject?

  • The COBRA-Slim strategy, initiating methotrexate with glucocorticoid bridging and applying treat-to-target, is a clinically and health-economically very effective approach for patients with early rheumatoid arthritis (RA) up to 2 years of follow-up.

What does this study add?

  • In patients with RA with poor-prognostic factors, starting with COBRA-Slim led to similar and sustained effectiveness profiles over 5 years as starting with conventional synthetic disease-modifying antirheumatic drug combinations and glucocorticoid bridging.

  • In patients with RA without a poor prognosis, starting a COBRA-Slim scheme led to better effectiveness over 5 years than a conservative step-up from methotrexate without glucocorticoids.

  • Only about one in six patients ever used glucocorticoids chronically for >6 months and about one in five patients initiated biologicals over 5 years.

How might this impact on clinical practice or future developments?

  • The COBRA-Slim scheme can serve as an effective initial treatment option for all types of patients with early RA, avoiding chronic glucocorticoid use in a large majority and reserving more intensive treatment combinations for insufficient responders only.


It is recommended to treat patients with early rheumatoid arthritis (RA) immediately, intensively and to a predefined target to rapidly control disease activity, and to avoid joint damage and functional decline.1 2 Methotrexate (MTX) forms the core of initial RA therapy. The care in early RA (CareRA) study was designed to investigate whether MTX should be combined with an additional conventional synthetic (cs) disease-modifying antirheumatic drug (DMARD) and/or with glucocorticoid bridging to induce a rapid, stable clinical response in patients with early RA. We demonstrated that MTX monotherapy with glucocorticoid bridging (COBRA Slim) in a treat-to-target setting had a better effectiveness in patients with poor prognostic factors over 2 years, with similar efficacy but a better safety profile, compared with csDMARD combinations and glucocorticoid bridging.3–5 Moreover, COBRA Slim showed benefit over a tight-step-up with MTX monotherapy in patients with RA without poor prognostic markers.5 6 The COBRA Slim regimen also proved to be more cost effective and was endorsed in the updated European League Against Rheumatism (EULAR) recommendations of 2019 to treat RA.7 8

As EULAR recommendations emphasise the importance of sustained remission or at least low disease activity, long-term evaluation of treatment effectiveness is necessary. The 11-year follow-up of the original COBRA trial showed reassuring long-term efficacy and safety of early intensive csDMARD combination therapy, even without a strict treat-to-target approach.9 More recently, the 10-year follow-up of the 'BehandelStrategieën in Reumatoïde Arthritis' (BeSt) trial, incorporating the tight control principle, confirmed the importance of early intensive combination therapy and demonstrated that drug-free remission and normalised mortality rates have become realistic outcomes.10 Despite all evidence above, current guidelines are still debated, especially the early use of glucocorticoids.11 Therefore, our objective was to study the long-term effectiveness of the initial treatment schemes used in CareRA within the 3-year observational CareRA-plus follow-up study. We compared patients according to their original treatment arms in terms of sustained disease control, use of csDMARDs, glucocorticoids and biologic (b)DMARDs, as well as safety over 5 years.


Study design

CareRA-plus was a 3-year observational follow-up study of the investigator-initiated, multicentre, pragmatic, 2-year CareRA RCT. In CareRA, we included 379 patients with early RA (<1 year), naïve to and without contraindications for csDMARDs or glucocorticoids. Participants completing CareRA were eligible for inclusion in CareRA-plus, which was conducted in 10 Belgian rheumatology centres (1 academic centre, 6 general hospitals and 3 private practices).

Initial and subsequent treatments in CareRA and CareRA-plus

Before randomisation in CareRA, patients were stratified into a high-risk or low-risk group based on the presence of classical prognostic factors, including RF/anticyclic citrullinated protein (ACPA) positivity, high baseline disease activity (Disease Activity Score based on 28 joints calculated with C reactive protein (DAS28-CRP)>3.2) and X-ray erosions. Patients in the high-risk group were randomised to one of three remission induction schemes: COBRA Classic (initial combination of MTX and sulfasalazine), COBRA Slim (MTX monotherapy) or COBRA Avant-Garde (initial combination of MTX and leflunomide). All COBRA schemes included an initial step-down scheme of oral prednisone, started at a high or moderate dose, and tapered weekly over 6 or 7 weeks to a low maintenance dose which was discontinued at week 28. The schemes combining two csDMARDs were tapered to csDMARD monotherapy at week 40 in case patients achieved low disease activity (DAS28-CRP≤3.2). Patients in the low-risk group were randomised to one of two schemes: the same COBRA Slim scheme as in high-risk or Tight Step Up (MTX monotherapy without oral glucocorticoids). During follow-up, the treat-to-target principle was applied. When low disease activity was not reached, treatment was adjusted according to two predefined steps, from week 8 onwards during the initial study year. During the second year, treatment was at the discretion of the rheumatologist. The protocol was described in previous publications.4 5 In CareRA-plus, further application of the treat-to-target principle was recommended but was left to the shared decision of rheumatologists and patients.

Assessments and outcomes in CareRA-plus

During CareRA-plus, participants were assessed every 6 months for 3 years. Disease activity (DAS28-CRP and Simplified Disease Activity Index (SDAI)), clinical parameters and functionality measured by the Health Assessment Questionnaire (HAQ) were registered. All (serious) adverse events ((S)AEs) considered to be relevant according to the investigators, and all DMARD and glucocorticoid use was recorded. Demographic variables, including comorbidities were registered at baseline of CareRA.

We assessed DMARD changes from baseline CareRA and over 5 years, resulting in three possible trajectories: Patients adding or switching csDMARDs, patients initiating a bDMARD and patients who never had an intensification. In the latter trajectory, patients stayed on csDMARD monotherapy from week 40 onwards in COBRA Classic and COBRA Avant-Garde, and from baseline in COBRA Slim till year 5 or alternatively, discontinued all DMARD therapy. We assessed glucocorticoid use by the cumulative dose of all systemic glucocorticoids and by chronic use (>3 or>6 months) of oral glucocorticoids outside of the initial prednisone step-down periods.

Yearly radiographs of hands and feet were read chronologically from baseline CareRA till year 5 using the Sharp van der Heijde (SvdH) score by one blinded reader.12 This reader was trained by an experienced reader who scored previously all radiographs of the original CareRA trial in the same manner. This training was validated by calculating an intraclass correlation coefficient, using radiograph scores from baseline till year 2 of both readers, indicating a good inter-reader reliability (ICC=0.83 (95% CI 0.81 to 0.85)). Radiographic progression was analysed by the change in the total SvdH score from baseline CareRA till year 5 as well as the evolution of the SvdH scores over time, and visualised by a cumulative probability plot in completers.

Statistical analysis

Each analysis compared the outcomes between the originally allocated treatment arms of the CareRA trial. Potential differences in outcomes, were examined by χ2, analysis of variance or Kruskal-Wallis, independent t-test or Mann-Whitney U test, when appropriate.

Proportions of patients in low disease activity or in remission according to DAS28-CRP or SDAI were calculated based on observed data and on an ‘intention-to-treat’ analysis including all patients consenting to participate to CareRA-plus. For the latter, missing data of components of the disease activity indices were imputed with multiple imputation by chained equations, resulting in 100 datasets. Each dataset was analysed separately and results were pooled using Rubin’s rules.7 13

The changes in DAS28-CRP, SDAI and HAQ were analysed over 5 years using linear mixed models (LMM). Remission and low disease activity rates over 5 years were analysed by generalised linear mixed models (GLMM). These mixed models incorporated a random intercept and a random slope for time with an unstructured covariance matrix. This method accounts for the repeated observations within a patient and allows the estimation of a different regression line for each patient with a different baseline value and rate of change over time. SvdH scores over time were compared using a generalised estimating equations (GEE) analysis with a negative binomial working distribution to address skewness of these data. For each model, treatment and time were used as determinants and it was tested whether there was an interaction between treatment and time. The numbers of AEs occurring during CareRA-plus were compared using Poisson regression. Significance level was set at 0.05. Analyses were carried out using SPSS V.26 and R V.4.0.1.



Of 322 patients who completed the 2-year CareRA study, 252 (78%) reconsented to be enrolled in CareRA-plus. We compared patients according to their originally allocated treatment in the high-risk group: COBRA Slim (n=75) versus COBRA Classic (n=69) or COBRA Avant-Garde (n=59) and in the low-risk group: COBRA Slim (n=23) vs tight-step-up (n=26). In both risk groups, treatment arms were well balanced in terms of demographic and clinical characteristics, registered at baseline CareRA (table 1). Patients entering CareRA-plus had similar demographics and clinical characteristics at the final 2-year visit of the CareRA trial compared with patients not entering the follow-up study. CareRA-plus patients were enriched for being ACPA positive, compared with non-participants, but ACPA status did not differ between treatment groups (online supplemental file 1). In total, 203 (81%) participants in CareRA-plus completed the extra 3-year follow-up, with similar frequencies or reasons for discontinuation between original treatment arms (figure 1). The 49 patients not completing the study were followed up during a median of 19 months (IQR 13–26) in CareRA-plus and 29 (59%) of them were in remission based on their last registered DAS28-CRP.

Figure 1

Flow chart of participants during the 3-year observational care in early rheumatoid arthritis (CareRA)-plus study. TSU, tight step up.

Table 1

Demographic and clinical characteristics of patients enrolled in care in early rheumatoid arthritis (CareRA)-plus per original treatment arm, as recorded at baseline CareRA

Disease activity over time

Disease activity improved rapidly during the first 16 weeks in CareRA and remained stable over the following 5 years among patients of the high-risk group (figure 2). There were no differences in DAS28-CRP or SDAI scores over time between treatment arms (LMM: respectively p=0.539 and p=0.431 for overall comparison; online supplemental file 2A). In the low-risk group, disease activity (DAS28-CRP) over 5 years was lower in patients starting COBRA Slim compared with those initiating tight-step-up (LMM: β=−0.46; CI (−0.63 to −0.29); p<0.001). Accordingly, SDAI scores over 5-year follow-up were also lower with the COBRA Slim strategy (LMM: β=−2.46; CI (-3.87 to −1.04); p=0.001; online supplemental file 2B).

Figure 2

Disease activity and physical functioning during 5 years of follow-up. Data are shown as observed, based on patients still in follow-up at each time point. Error bars indicate the 95% CIs. DAS28-CRP, Disease Activity Score based on 28 joints calculated with C reactive protein; SDAI, Simplified Disease Activity Index; HAQ, Health Assessment Questionnaire.

Remission and low disease activity states

Based on available data of CareRA-plus participants who completed the 5-year follow-up since treatment initiation, overall, 89% had low disease activity (DAS28-CRP≤3.2), and 74% were in remission (DAS28-CRP<2.6) at year 5. Low disease activity measured by SDAI ≤11 was achieved by 89% of all patients and remission (SDAI ≤3.3) by 40% of patients. The proportion with a DAS28-CRP<2.6 at year 5 in high-risk patients was 72%, 77% and 64% for the Classic, Slim and Avant-Garde group respectively (p=0.403). In the low-risk population, 83% of patients in the Slim and 82% in the tight-step-up arm had a DAS28-CRP<2.6 at year 5 (p=0.945). Remission rates at year 5 based on an intention-to-treat analysis with imputation of missing data were comparable (online supplemental file 3). Remission and low disease activity rates are shown over time in figure 3/online supplemental file 4. Occurrence of remission over time assessed by DAS28-CRP or SDAI was similar between treatments in the high-risk group (GLMM: respectively, p=0.798 and p=0.224 for overall comparison;online supplemental file 2A). In the low-risk group, patients on COBRA Slim had higher odds of achieving remission over time, compared with patients started on tight-step-up (GLMM: OR=2.62 CI (1.43 to 4.81); p=0.002 for DAS28-CRP remission, OR=3.27 CI (1.35 to 7.91); p=0.009 for SDAI remission) (online supplemental file 2B).

Figure 3

Remission and low disease activity rates during 5 years of follow-up. Data are shown as observed, based on patients still in follow-up at each time point. DAS28-CRP, Disease Activity Score based on 28 joints calculated with C reactive protein; LDA, low disease activity; remission DAS28-CRP<2.6; LDA DAS28-CRP≤3.2; SDAI, Simplified Disease Activity Index; Remission SDAI≤3.3; LDA SDAI≤11.


In the high-risk group, mean HAQ scores over 5 years were comparable between treatment arms (LMM: p=0.374 for overall comparison; online supplemental file 2A). Among patients of the low-risk group, those treated initially with COBRA Slim had lower HAQ scores and thus better functionality over 5 years (LMM: β=0.21 CI (−0.41 to −0.01); p=0.041; online supplemental file 2B).

Radiographic progression

After 5 years, radiographic progression, measured as increase in SvdH score, in patients completing the study was limited and comparable between treatment arms in the high-risk population. More specifically, three patients in Classic, three in Slim high-risk and one in Avant-garde had an increase in SvdH score >5. There were 11 patients in Classic, 9 in Slim and 5 in Avant-Garde who had an increase in SvdH score >0.5 (p=0.399). In the low-risk group, there were no patients with a change in SvdH >5, and there was 1 Slim patient with a change >0.5 (p=0.283). A cumulative probability plot is shown in online supplemental file 5. Longitudinal analyses demonstrated that the mean change in SvdH score over 5 years was similar between treatment arms in the high-risk and the low-risk group (GEE: p=0.524 and p=0.928 for overall comparison, respectively; online supplemental file 2).

Treatment intensifications

At the year 5 visit, 71%, 61% and 50% of high-risk patients were on csDMARD monotherapy (mostly MTX) in the Classic, Slim and Avant-Garde arm respectively. In the low-risk group, 65% in the COBRA Slim and 62% in the tight-step-up arm were taking a single csDMARD. All treatment schemes showed similar trajectories of changes in csDMARD or bDMARD use over the 4 years following the protocolised first year of the trial (figure 4). Of all patients completing the study, 56% never had their DMARD therapy intensified for 5 years. More specifically, 64%, 58% and 48% of high-risk patients in the Classic, Slim and Avant-Garde arm reached year 5 without DMARD intensifications. In the low-risk group, 50% of Slim low risk and 52% of tight-step-up patients never had an intensification in their DMARD therapy. During the 5 study years, biologics were initiated in 22% of all participants consenting to CareRA plus: 23% of Classic, 23% of Slim high-risk, 25% of Avant-Garde, 17% of Slim low-risk, and 15% of tight-step-up patients. At the year 5 visit, only 9% of all participants were using oral glucocorticoid therapy for >3 months at a median dose of 5.0 mg prednisone equivalent. The cumulative glucocorticoid dose taken by patients was comparable between treatment arms in both risk groups from year 2 onwards (figure 5). Throughout the entire 5-year study, chronic glucocorticoid use of >3 or >6 months outside of the initial prednisone schemes was limited to, respectively, 25% or 17% of all patients (online supplemental file 6).

Figure 4

Medication profiles taken by participants during 5 years of follow-up in each treatment arm. No intensifications=participants who did not have to intensify their DMARD treatment; added or switched csDMARD=participants who added or switched a csDMARD; initiated biologic=participants who initiated biological DMARD(s); percentages are calculated on patients still in follow-up at each time point; Nb, number; csDMARD, conventional synthetic disease-modifying antirheumatic drug.

Figure 5

Mean cumulative glucocorticoid dose per treatment arm and per year of follow-up. All systemic glucocorticoids taken, including the initial step-down prednisone schemes, and all other oral, intramuscular or intra-articular glucocorticoids were considered. The mean cumulative dose of prednisone equivalent in each year is depicted, calculated on patients still in follow-up; y=year.


In high-risk patients, the total numbers of AEs throughout the 3-year follow-up in CareRA-plus were 70 in 36 Classic, 95 in 48 Slim and 80 in 36 Avant-Garde patients (p=0.182). In the low-risk group, there were 18 AEs in 10 Slim and 36 in 17 tight-step-up patients (β=−0.571 CI (−1.136 to −0.005); p=0.048) (table 2 and online supplemental file 2).

Table 2

Safety analysis from year 2 till year 5 within the observational care in early rheumatoid arthritis (CareRA)-plus follow-up study


Initial intensive treatment with csDMARDs and bridging glucocorticoids in CareRA, followed by treatment adaptation to a target of low disease activity, resulted in sustained control of disease activity over 5 years in a large majority of patients with RA with markers of poor prognosis. Additionally, these treatment schedules led to a sustained improvement in functionality and very limited progression in joint damage over 5 years. During the 3-year follow-up in CareRA-plus with further targeted treatment, safety data were reassuring and comparable between the schemes. The simplified COBRA-Slim scheme, starting MTX monotherapy with glucocorticoid bridging and, if needed, adapting treatment to a realistic disease activity target, can lead to comparable sustained treatment responses on the long term as more complex combination schemes and is an effective initial treatment option.

Patients without markers of poor prognosis who started COBRA-Slim had better disease control and functionality over 5 years than patients starting MTX without glucocorticoids and no more safety issues on the long-term. Suppression of joint damage progression over 5 years was comparable with both treatment schemes. Therefore, our results confirm the benefit of combining initial csDMARD treatment with bridging glucocorticoids also in patients without presumed markers of poor prognosis.14 15

Remarkably, 56% of all patients never had their DMARD therapy intensified during the first 5 years of treatment, and many patients were on DMARD monotherapy at year 5, without differences between treatment schemes. Hence, the simplified COBRA-Slim scheme did not lead to a higher need for DMARD intensifications on the long-term compared with the initial combinational regimens. Additionally, contrary to the common perception, the chronic use of glucocorticoids was limited, with the vast majority of patients being able to stop glucocorticoids after the induction phase. Moreover, chronic glucocorticoid use was comparable in low-risk patients who did or did not initiate a bridging scheme of prednisone. Further, overall bDMARD use was limited, with 22% of participants having ever taken a bDMARD over 5 years. These results indicate not only a sustained long-term effectiveness of the studied treatment schemes including glucocorticoids but also confirm our earlier findings on the feasibility of these regimens in clinical practice.16

The results of the additional 3-year follow-up in CareRA-plus confirm and extend on the conclusions after 2 years in the CareRA trial. Moreover, these results support findings regarding the sustained effectiveness of COBRA schemes from the original COBRA, the BeSt and the COBRA-Light trial.9 10 17 18 Results of the BeSt trial showed that initial combination therapy of MTX, SSZ and prednisone resulted in sustained clinical improvement over 10 years. Similarly as in CareRA, most patients in BeSt were able to taper their combinational DMARD treatment to monotherapy and glucocorticoid use was very limited by the end of the follow-up.10 Also, the COBRA-light trial demonstrated that MTX with prednisolone bridging had similar efficacy and safety outcomes over 4 years compared with a combination of MTX, SSZ and prednisolone bridging.18 However, this protocol prescribed the addition of etanercept in case DAS44 <1.6 was not achieved, which resulted in higher proportions of patients having ever initiated a bDMARD (67% vs 22% in CareRA) Additionally, in the COBRA-Light trial, more patients used glucocorticoids for >3 months during follow-up (42% vs 25% in CareRA).19 Our data are also fully in line with the 5y IMPROVED data20 and both confirm again the EULAR recommendations and contradict the recent draft of the 2020 revision of the ACR recommendations for the management of RA, advising to minimise the use of bridging glucocorticoids.21

Our study population is close to a daily practice early RA population. We included patients with diverse disease characteristics in terms of severity, autoantibody positivity, erosive or comorbidity status, from different types of practices, and followed them regularly for a long period. These features support the external validity of our results and are indicative for a good applicability of such intensive treatment in daily practice.

The vast majority of patients completing the preceding CareRA study reconsented to participate in CareRA-plus. Enrolled patients did not differ in demographics, nor in clinical characteristics from patients not enrolled into CareRA-plus, except for being more ACPA positive. This enrichment for ACPA positivity might have resulted in an underestimation of treatment effect since ACPA is assumed to be a prognostic factor of poor outcome.22 However, ACPA-positive patients were well distributed between treatment arms, so no selection bias influenced group comparison.

In conclusion, we demonstrated that all intensive treatment strategies using bridging steroids showed excellent long-term clinical outcomes without chronic glucocorticoid use in the majority of patients. Initial COBRA Slim therapy with MTX and bridging prednisone demonstrated comparable 5-year effectiveness as more complex combination schemes in high-risk patients with early RA and a better effectiveness than conservative step-up MTX monotherapy in low-risk patients.

Data availability statement

Data are available upon reasonable request. The authors commit to making the relevant anonymised patient level data available for a specified purpose approved by the institution and the principal investigator of the CareRA plus study and with a signed data access agreement.

Ethics statements

Ethics approval

The CareRA plus study was approved by the leading Ethics Committee of the University Hospitals Leuven after consulting the medical ethics committee of each participating centre (ref s53336) and all study participants gave their written informed consent before inclusion.


We would like to show our gratitude to all participating patients, as well as to the investigators of the CareRA study group at all sites. We thank Sylvie Van Vlasselaer and all other study personnel for their contributions to the study. We would also like to thank Tijs Kupers for his help in scoring the X-rays and statistician Anna Ivanova for her advice.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Lay summary

    Disclaimer : This is a summary of a scientific article written by a medical professional (“the Original Article”). The Summary is written to assist non medically trained readers to understand general points of the Original Article. It is supplied “as is” without any warranty. You should note that the Original Article (and Summary) may not be fully relevant nor accurate as medical science is constantly changing and errors can occur. It is therefore very important that readers not rely on the content in the Summary and consult their medical professionals for all aspects of their health care and only rely on the Summary if directed to do so by their medical professional. Please view our full Website Terms and Conditions.
    Copyright © 2021 BMJ Publishing Group Ltd & European League Against Rheumatism. Medical professionals may print copies for their and their patients and students non commercial use. Other individuals may print a single copy for their personal, non commercial use. For other uses please contact our Rights and Licensing Team.


  • Handling editor Josef S Smolen

  • Twitter @sophie_33pl, @DiederikDeCock

  • Contributors PV, JJ and RW designed the study protocol in collaboration with the CareRA study group. Investigators of the CareRA study group, including PV and RW recruited and enrolled patients and were responsible for daily patient management. PV, VS and JJ were responsible for coordination of the trial and of collection of data. VS was responsible for data analysis. All authors contributed to interpretation of the data. Furthermore, VS, PV, DDC and RW drafted the manuscript. SP, KVdE, DB and JJ revised it critically for important intellectual content. All authors have approved the final draft for publication. PV and VS are the guarantors.

  • Funding The CareRA RCT was supported by a grant from the Flemish Governmental Agency for Innovation by Science and Technology (IWT). PV holds the unrestricted Pfizer Chair for ‘Early Rheumatoid Arthritis Management’ at the KU Leuven. Leflunomide was made available for CareRA for free by SANOFI Belgium without any influence on trial design.

  • Disclaimer The funders had no role in study design, data collection, analysis and interpretation or reporting of this study. Patrick Verschueren and Veerle Stouten affirm that this manuscript is an honest, accurate, and transparent account of the study being reported.

  • Competing interests None declared.

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

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.