You are here

  • Home
  • Archive
  • Volume 10, Issue 1
  • What benefit–risk trade-offs are acceptable to rheumatoid arthritis patients during treatment selection? Evidence from a multicountry choice experiment

Article Text

Article menu

Download PDFPDF

Rheumatoid arthritis
Original research
What benefit–risk trade-offs are acceptable to rheumatoid arthritis patients during treatment selection? Evidence from a multicountry choice experiment
  1. Rieke Alten1,
  2. Juan Carlos Nieto-Gonzalez2,
  3. Peggy Jacques3,
  4. Carlomaurizio Montecucco4,
  5. Robert Moots5,6,
  6. Helga Radner7,
  7. Harald E Vonkeman8,9,
  8. Sebastian Heidenreich10,
  9. Chiara Whichello10,
  10. Nicolas Krucien10 and
  11. Katrien Van Beneden11
  1. 1Schlosspark Klinik, University Medicine Berlin, Berlin, Germany
  2. 2Servicio de Reumatología, Hospital General Universitario Gregorio Marañón, Madrid, Spain
  3. 3Department of Rheumatology and VIB Inflammation Research Center, University Hospital Ghent, Ghent, Belgium
  4. 4Division of Rheumatology, University of Pavia and Fondazione IRCCS Policlinico San Matteo, Pavia, Italy
  5. 5Department of Rheumatology, Aintree University Hospital, Liverpool, UK
  6. 6Faculty of Health, Care and Medicine, Edge Hill University, Ormskirk, UK
  7. 7Department of Internal Medicine III, Division of Rheumatology, Medical University Vienna, Vienna, Austria
  8. 8Department of Rheumatology and Clinical Immunology, Medisch Spectrum Twente, Enschede, The Netherlands
  9. 9Department of Psychology, Health & Technology, University of Twente, Enschede, The Netherlands
  10. 10Evidera Inc, London, UK
  11. 11Galapagos NV, Mechelen, Belgium
  1. Correspondence to Dr Rieke Alten; rieke.alten{at}schlosspark-klinik.de

Abstract

Objective Understanding preferences of patients with rheumatoid arthritis (RA) can facilitate tailored patient-centric care. This study elicited trade-offs that patients with RA were willing to make during treatment selection.

Methods Patients with RA completed an online discrete choice experiment, consisting of a series of choices between hypothetical treatments. Treatment attributes were selected based on literature review and qualitative patient interviews. Eligible patients were ≥18 years old, diagnosed with RA, receiving systemic disease-modifying antirheumatic drug therapy, and residents of Europe or USA. Male patients were oversampled for subgroup analyses. Data were analysed using a correlated mixed logit model.

Results Of 2090 participants, 42% were female; mean age was 45.2 years (range 18–83). Estimated effects were significant for all attributes (p<0.001) but varied between patients. Average relative attribute importance scores revealed different priorities (p<0.001) between males and females. While reducing pain and negative effect on semen parameters was most important to males, females were most concerned by risk of blood clots and serious infections. No single attribute explained treatment preferences by more than 30%. Preferences were also affected by patients’ age: patients aged 18–44 years placed less importance on frequency and mode of treatment administration (p<0.05) than older age groups. Patients were willing to accept higher risk of serious infections and blood clots in exchange for improvements in pain, daily activities or administration convenience. However, acceptable trade-offs varied between patients (p<0.05).

Conclusion Treatment preferences of patients with RA were individual-specific, but driven by benefits and risks, with no single attribute dominating the decision-making.

  • arthritis, rheumatoid
  • antirheumatic agents
  • therapeutics

Data availability statement

Data are available on reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/rmdopen-2023-003311

Statistics from Altmetric.com

    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.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Patients with rheumatoid arthritis and their prescribers face challenging trade-offs during treatment selection. To accommodate patients’ circumstances in comprehensive disease management, current recommendations for management emphasise the need to recognise patient preferences.

    WHAT THIS STUDY ADDS

    • Results of this study show that preferences of patients with rheumatoid arthritis were driven by multiple benefits and risks of treatments, with no single attribute dominating the decision making. The trade-offs that patients were willing to make were heterogeneous and varied both between both individuals and subgroups.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • These findings emphasise the importance of considering the entire treatment profile, including benefits, risks and administration, to support shared decision-making between providers and patients.

    Introduction

    Disease modification is the guiding principle for the management of rheumatoid arthritis (RA), with treatments aiming to control the inflammatory cascade and to improve symptoms, physical function, quality of life and work capacity, while inhibiting long-term complications from structural damage.1 With the increasing number of efficacious disease-modifying agents and the rise of precision medicine,2 recommendations on the development of patient-centric treatment targets can help inform RA management to achieve comprehensive disease care. For instance, the European Alliance of Associations for Rheumatology (EULAR) recommendations state that treatment selection should be based on shared decision-making between patient and rheumatologist.1 This shared decision-making implies the recognition of patient preferences, and involves all aspects of the disease, including information on RA and its potential consequences, the modalities of disease assessment, decisions on the therapeutic target and the potential means to reach the target, as well as the development of a management plan and discussions on the benefits and risks of individual therapies.

    Considering the complexity of the RA treatment landscape and the diverse treatment targets, engaging in shared decision-making is key to understanding patients’ treatment priorities and the trade-offs they are willing to make, as this is important for delivering tailored care.1 For example, sulfasalazine and methotrexate are effective first-line conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), but have been linked to adverse events, including gastrointestinal complications, infections, skin reactions, neurotoxicity, malignancies, infertility, and negative effects on sperm that may or may not be reversible.3–7

    Treatment selection is especially challenging among advanced therapies. While the American College of Rheumatology and EULAR align on recommendations for first-line therapies, approximately one-third of patients have an inadequate response to conventional csDMARDs, such as methotrexate and sulfasalazine, and often progress to one of multiple available advanced treatments.1 8 9 These may be biological (bDMARDs), such as tumour necrosis factor alpha (TNFα) inhibitors, T cell costimulation modulators, CD20, and interleukin-6R inhibitors, or targeted synthetic (tsDMARDs), such as Janus kinase (JAK) inhibitors.10 Preventive effects of authorised advanced therapies and undesirable effects, including potential changes in laboratory parameters and increased risk of infections, malignancies and gastrointestinal disorders, have been well characterised.11 12 Therefore, when making a specific treatment decision, risks and benefits can be compared. To facilitate the interpretation of benefit–risk comparisons, patient preference information can provide insights into the trade-offs that patients are willing to make when selecting an advanced therapy for RA. In addition, understanding how acceptable trade-offs may vary in different patient populations can help guide individual shared-decision making processes in doctor–patient interactions. Patient preference information is also increasingly used by decision-makers, such as regulators or health technology assessment agencies, to help interpret clinical data from patients’ perspectives.13–15

    The objective of this study was to elicit benefit–risk trade-offs that patients with RA in Europe and the USA are willing to make. For this purpose, an online discrete choice experiment (DCE) was conducted, in which patients with RA made trade-offs between multiple benefits and risks of various DMARDs, with a focus on advanced therapies. While previous preference research has been undertaken in RA, insights are limited, because these studies grouped adverse events into unclear, overarching categories, outcomes did not relate to clinical endpoints, or studies were conducted in a specific small setting or population that limits insights into preference heterogeneity.16–22

    Methods

    Study design

    An online DCE was conducted from September to October 2021 in adults with RA to elicit the benefit–risk trade-offs they were willing to make. Patients were eligible if they were ≥18 years old, resident of the USA, the UK, France (FR), Germany (DE), Italy (IT) or Spain (ES), and had a RA diagnosis for which they were currently receiving systemic therapy (csDMARDs, bDMARDs, tsDMARDs). Patients were excluded if they reported diagnosis with psoriatic arthritis, axial spondyloarthritis, ankylosing spondylitis or psoriasis, without RA, or were enrolled in a clinical trial at the time of the study. All patients had to provide online informed consent prior to taking part in the study. Male patients were oversampled with a target quota of 50% to allow for subgroup analysis and exploring their willingness to accept the risk of negative effects on semen parameters. All patients were recruited via nationally representative online access panels, physician referrals, patient organisations and social media. Participants were recruited by sending an invitation with information about the study, with contact details or a URL link. Participants who expressed interest and consented to participate were screened for eligibility. On completion of study participation, participants were remunerated either as a direct bank transfer, gift card or panel points.

    Survey design

    A multiphase approach was used to design and test the DCE in compliance with best-practice guidance.23 24 The study was conducted in five phases: first, a targeted literature review identified 26 concepts (ie, 11 benefits, 10 risks, 5 other) potentially relevant for explaining patients’ treatment preferences (online supplemental A.1, table A.1). Second, concepts identified from the literature were discussed in 30 semistructured 60- min virtual interviews that were conducted with patients across the six target countries (online supplemental A.2, tables A.2, A.3 and figure A.1). Third, an initial DCE design was tested and iteratively refined in 30 web-assisted 60-min pretesting interviews with patients across the six target countries (online supplemental A.3, table A.4). The aim of the qualitative pretesting was to ensure that the survey and DCE were clear and that patients were willing and able to make trade-offs between the different attributes of DMARDs.25 Fourth, a quantitative pilot study (online supplemental A.4 and A.5) was conducted with 712 patients (FR: n=111; DE: n=92; IT: n=144; ES: n=114; UK: n=93; USA: n=158). The purpose of the quantitative pilot was to explore the expected data quality and to assess whether patients were able to distinguish between the different risk levels included in the DCE. No changes were made to the instrument following the quantitative pilot and the data were subsequently merged with the main dataset. Fifth, the final data were collected between September and October 2021 (approximately 6 weeks).

    The DCE was part of a wider survey that included a screening questionnaire, an informed consent form, patient information material, the DCE, numeracy and health literacy questions,26–28 and clinical and demographics questionnaires (online supplemental B).

    DCE design

    The final set of DMARD attributes, their definitions and corresponding levels considered in the DCE are shown in table 1.

    View this table:
    Table 1

    Attributes and levels

    A D-efficient design was generated in Ngene V.1.2.1 (ChoiceMetrics Sydney, Australia) and included 45 experimental DCE choice tasks that were split equally across 5 blocks.29 The design was optimised using Bayesian priors obtained from the quantitative pilot.30 Within each DCE choice task, patients were asked to choose between two hypothetical DMARDs described by the attributes and levels as determined by the experimental design. An example choice task is shown in figure 1. The order of choice tasks was randomised across patients to reduce the risk of ordering effects.31 A first practice choice task and two tasks for testing the internal validity of responses were added to the experimental choice tasks.32 The first validity test repeated the 10th choice task as seen by patients to explore choice consistency. The second validity test was a dominance test in which one of the two DMARDS outperformed the alternative on all attributes (ie, administration was identical). In line with best-practice, no patient was excluded from the analysis based on the internal validity tests to avoid introducing selection bias.33 Overall, each patient completed 12 choice tasks.

    Figure 1
    Figure 1

    Example choice task asking participants to select one of two hypothetical treatments.

    Analysis

    The experimental choice tasks were used for the analysis of the DCE data. All statistical tests were two-sided and used a significance level of 0.05. Comparison of statistical performance across models was based on Bayesian information criterion and the adjusted McFadden R2.34 All analyses were conducted in the statistical software R V.4.0.2.

    A correlated mixed logit model was used to analyse the DCE data within a random utility maximisation framework that estimated the effect of changes in attributes on preferences as part-worth utilities.35 36 Compared with classical multinomial logit models, the estimated model implicitly accounted for panel effects, heterogeneity in preferences and variations in choice consistency.37 Two behavioural outputs were obtained from the estimates for the overall sample and by gender: first, relative attribute importance (RAI) scores were calculated to measure the maximal contribution of each attribute to a treatment choice. Second, trade-offs that patients with RA were willing to make between attributes were quantified as maximum acceptable risks (MAR) of blood clots, serious infections and negative effects on sperm. Delta method was used to obtain the SEs and 95% CIs of MAR estimates.38 In addition, subgroup analyses were conducted by estimating RAI scores by age and country using interaction effects included in the mixed logit model. More details on the analysis are included in online supplemental D.

    Results

    Participants

    A total of 44 221 potential participants were invited across the six countries, with 2090 eligible patients (table 2; Online supplemental figure C.1) consenting and completing the survey.

    View this table:
    Table 2

    Sociodemographic and clinical characteristics

    In line with the sampling quota, 42% of the patients were female (n=878), with an overall mean age of 45.2 years (SD=11.3, range 18–83). Female patients were slightly older (mean 46.3 years, SD 12.2) than males (mean 44.4 years, SD 10.4). Most patients had high numeracy (ie, adequate facility with numbers; 79%; n=1656), but fewer patients had high health literacy (ie, adequate facility with reading; 47%; n=973). Half of patients (n=1038, 49%) had been diagnosed 3 or more years ago and most reported 0 to 10 tender or swollen joints (62%; n=1279). The three most reported current symptoms by patients were joint pain (57%; n=1189), joint tenderness (53%; n=1104), and joint swelling (47%; n=988).

    Most patients had children (87%; n=1828) and had not conceived children after their diagnosis (79%; n=1448). The majority were not planning on having more children in the future (77%; n=1602). Of male participants, 29% (n=352) were planning to have more children.

    Preferences

    Within the DCE, most patients passed both the dominance (76%; n=1588) and stability (75%; n=1576) tests, with observed failure rates comparable to other health DCEs in the literature (online supplemental table C.1).32 The mean survey completion time was 17 min (SD=12.5; median=14 min). The data fit for the mixed logit model was good (adjusted McFadden R2=0.583), suggesting it was able to explain the choices that patients made in the DCE. Estimated effects were significant for all attributes (p<0.001), implying that they all influenced patient preferences for DMARDs (online supplemental table C.2).

    On average, patients tended to prefer an oral pill every day over an injection every other week (overall: p<0.001; male: p<0.001; female: p<0.001; figure 2, online supplemental table C.3), or an injection once a week (overall: p<0.001; male: p<0.1; female: p<0.001). Patients valued all reductions in pain and avoiding any of the considered risks (overall: p<0.001; male: p<0.001; female: p<0.001). In addition, treatment preferences of male patients were significantly affected by the risk of negative effects on semen parameters (p<0.001). All estimates were found to vary (p<0.001) between patients, indicating the presence of preference heterogeneity.

    Figure 2
    Figure 2

    Main estimates: the effect of changes in attributes on preferences.

    RAI scores implied by the model estimates are presented in figure 3. While reducing pain was the largest driver of male patients’ treatment preferences (RAI 25%; 95% CI 20% to 30%; Online supplemental table C.4), female patients placed the highest importance on avoiding blood clots (RAI 30%; 95% CI 28% to 32%). Further, while the risk of serious infections was the second most important attribute to female patients (RAI 23%; 95% CI 21% to 25%), it ranked as the fourth most important attribute for male patients (RAI 14%; 95% CI 11% to 17%). Avoiding negative effects on semen parameters was the second most important driver of male patients’ preferences (RAI 23%; 95% CI 19% to 26%). Furthermore, while female patients placed a higher importance on reducing difficulties with daily activities (male RAI 10%; 95% CI 7% to 13%; female RAI 16%; 95% CI 14% to 19%) than they placed on treatment administration (male RAI 10%; 95% CI 8% to 13%; female RAI 9%; 95% CI 7% to 11%), male patients placed a comparable importance on both attributes (p>0.05). Overall, no treatment attribute contributed with more than 30% to treatment preferences of male or female patients.

    The RAI scores can be used to gain insights into differences and similarities of treatment priorities of male and female patients. For example, reducing the risk of blood clots was 1.4 (=19%/14%) times and 1.3 (=30%/23%) times more important than the risk of serious infections to male and female patients, respectively. Reducing pain was 2.6 (=25%/10%) times and 1.3 (=21%/16%) times more important than reducing difficulties with daily activities to male and female patients, respectively. Furthermore, while male patients considered reducing pain as 1.3 (=25%/19%) times more important than reducing the risk of blood clots, female patients considered reducing the risk of blood clots as 1.4 (=30%/21%) times more important than reducing pain.

    Figure 3
    Figure 3

    RAI in the overall population and by sex. Respondents: 2090. Observations: 18 810. Parameters: 64. Null log-likelihood: −13 038.1. Model log-likelihood: −5404.4. Note: The relative importance of negative effects on sperm is based on male responses only. RAI, relative attribute importance.

    Trade-offs

    The MAR estimates obtained from the analysis are presented in table 3 and provide insights into the benefit–risk trade-offs patients were prepared to make.

    View this table:
    Table 3

    Maximum acceptable risk of serious infections, blood clots and negative sperm effects

    For example, patients were willing to accept higher risks of blood clots (male MAR 1.8%, 95% CI 0.9% to 2.8%; female MAR 0.8%, 95% CI 0.4% to 1.2%), serious infections (male MAR 2.5%, 95% CI 1.0% to 4.0%; female MAR 1.0%, 95% CI 0.5% to 1.6%) or negative effects on sperm (male MAR 7.4%, 95% CI 4.2% to 10.7%) for being able to take an oral pill every day instead of receiving an injection once a week. Similarly, patients were willing to accept higher risks of blood clots (male MAR 2.3%, 95% CI 1.7% to 3.0%; female MAR 1.2%, 95% CI 1.0% to 1.5%), serious infections (male MAR 3.2%, 95% CI 2.1% to 4.2%; female MAR 1.6%, 95% CI 1.2% to 2.0%) or negative effects on sperm (male MAR 10.4%, 95% CI 7.9% to 13.0%) in exchange for reducing the amount of pain from 30 to 10 on a scale of 0–100. Similar observations were made for improved performance of daily activities. MAR estimates also provided insights into trade-offs between risks. For example, patients were willing to accept an extra risk of blood clots in exchange for reduced risk of serious infections from 3% to 0% (male MAR 2.2%, 95% CI 1.6% to 2.9%; female MAR 2.3%, 95% CI 2.0% to 2.6%) or from 6% to 0% (male MAR 4.5%, 95% CI 3.2% to 5.7%; female MAR 4.6%, 95% CI 4.0% to 5.2%).

    Subgroup analyses

    In subgroup analyses by age, patients older than 65 were less (p<0.05) concerned with the risk of serious infections (RAI 16%, 95% CI 12% to 20%) than patients aged 18–44 (online supplemental figure C.2). Conversely, younger patients aged 18–44 (RAI 5%, 95% CI 2% to 7%) placed less importance on frequency and mode of treatment administration (p<0.05) than patients aged 45–64 (RAI=12%, 95% CI 10% to 14%) or those older than 65 (RAI 12%, 95% CI 10% to 13%). Similarly, patients aged 18–44 (RAI 10%, 95% CI 7% to 14%) were less (p<0.05) concerned about difficulties with daily activities than patients aged 45–64 (RAI 13%, 95% CI 10% to 16%) or those older than 65 (RAI 15%, 95% CI 12% to 18%).

    Some similarities and differences were observed between countries. Patients from Spain, Italy, the UK and France placed more importance on amount of pain than on difficulty with daily activities, whereas patients from the USA and Germany indicated almost similar importance for amount of pain and daily difficulties (online supplemental table C.5). Patients from France (RAI 32%, 95% CI 25% to 38%) and the UK (RAI 25%, 95% CI 19% to 32%) placed more importance on avoiding pain (p<0.05) than those from Spain (RAI 19%, 95% CI 14% to 25%), the USA (RAI=15%, 95% CI 9% to 21%), Italy (RAI=17%, 95% CI 11% to 24%), or Germany (RAI=13%, 95% CI 2% to 23%). Conversely, participants from Germany (RAI=27%, 95% CI 20% to 33%) and Spain (RAI=27%, 95% CI 23% to 30%) placed higher importance on the risk of blood clots (p<0.05) than patients from France (RAI=14%, 95% CI 10% to 19%), Italy (RAI=22%, 95% CI 18% to 30%), the UK (RAI=22%, 95% CI 18% to 25%) and the USA (RAI=20%, 95% CI 17% to 23%). Patients from Italy and the USA placed more importance on the risk of negative effects on semen parameters than other countries. No differences in RAI were found (p>0.05) when comparing preferences between European (pooled) and US patients.

    Patients without prior experience with advanced therapies placed a higher relative importance on treatment administration (RAI=13%, 95% CI 11% to 15%) and difficulty with daily activities (RAI=18%, 95% CI 16% to 20%) than the average patient in the sample. Patients with prior experience of ≥3 advanced therapies placed a higher relative importance on the risk of negative effects on sperm (RAI=30%, 95% CI 22% to 38%) and a lower relative importance on difficulty with daily activities (RAI=3%, 95% CI 0% to 7%) than the average patient in the sample.

    Discussion

    Summary

    To the best of our knowledge, this is the largest study concerned with the treatment preferences of patients with RA, and specifically designed to capture benefit–risk trade-offs relevant to healthcare decision-making; it is also the first study to include potential impacts on semen parameters. This patient preference study contributed to the literature by offering specific and applicable insights into the benefit–risk trade-offs that patients with RA are willing to make, while accounting for preference heterogeneity. We found that the trade-offs patients were willing to make were heterogeneous and varied both between and within subgroups. A central finding of this study is that none of the considered treatment attributes was a dominant driver for treatment preferences for patients with RA. This highlights the need for a careful consideration of the entire profile of suitable DMARDs, with the aim of weighing relevant benefits, risks and administration aspects.

    While administration contributed less than 10% to treatment preference, patients with RA were also willing to make trade-offs between convenience, benefits and risks. Specifically, male patients were prepared to accept an extra 1.3% and female patients an extra 0.9% risk of serious infections for being able to take an oral pill once daily instead on relying on an injection that is administered every other week. Preferences also varied significantly between age groups and countries. Patients with RA aged 18–44 placed less importance on treatment administration than older patients. Overall, the relative importance that patients placed on difficulties with daily activities tended to increase with age, while the importance that male patients with RA placed on the risk of negative effects on semen parameters decreased with age. The amount of pain was more important to patients from Spain, Italy, the UK and France than difficulty with daily activities, whereas patients from the USA and Germany placed similar importance on pain and daily difficulties. Additionally, patients from Spain and Germany were most concerned about blood clots, and patients from Italy and the USA placed highest importance on risk of negative effects on semen parameters compared with other countries. These spatial differences indicate the need for reflecting on local perceptions and needs when evaluating new treatments and developing treatment guidelines. However, no significant differences were found when comparing a pooled European sample to a pooled US sample.

    Preferences were found to vary significantly in the patient population, at both individual and subgroup levels. Overall, no treatment attribute contributed with more than 30% to treatment preferences of male or female patients. The study demonstrated that patients tended to prefer an oral pill every day over an injection and valued all reductions in pain and avoiding any of the considered risks. At the gender level, female patients placed a higher importance on reducing difficulties with daily activities than they placed on treatment administration, while male patients placed a comparable importance on both attributes. In addition, while pain relief had the largest average impact on male patients’ treatment preferences, female patients placed on average the highest importance on the risk of blood clots. This resulted in differences in the average trade-offs that male and female patients were willing to make. For instance, female patients accepted an extra 1.2% risk of blood clots for reducing their pain score from 30 to 10, compared with male patients who would be willing to accept an extra 2.3% risk of blood clots for the level of pain relief. Thus, women required a higher benefit to compensate for a given level of blood clot risks than men.

    This was also the first study to explore the relative importance that male patients with RA placed on a potential risk of negative effects on sperm from DMARDs. While only 29% of males were still planning to have children, avoiding the risk of negative effects on sperm was valued by male patients. However, male patients were willing to accept higher risks of negative impacts on sperm for additional treatment benefits. For instance, they were willing to accept an extra 17.3% additional risks of negative effects on sperm for reducing pain from 80 to 40 or an additional 9.9% risk of negative effects on sperm for reducing difficulties with daily activities from severe to moderate. While few data on the effects of advanced treatments for inflammatory rheumatic diseases on semen parameters are available, recent data indicated that filgotinib, a preferential JAK1 inhibitor, does not appear to have a negative impact on this aspect of health based on data from the recent MANTA and MANTA/RAy clinical trials.39

    While a number of studies have examined preferences of rheumatologists for DMARDs,21 22 40 41 patient preference data that is suitable for characterising bDMARDs and tsDMARDs is limited. However, findings of this study complement existing evidence. For example, Mathijssen et al conducted a DCE among patients with RA and found that their treatment preferences were affected by route of administration, frequency of administration, and risk of serious infections.42 Similarly, in a DCE conducted by Alten et al, patients were found to prefer ‘oral administration’ over ‘intravenous infusion’.43 Results from two recent systematic reviews evaluating patient preference studies in RA showed variability in preferences across different populations.21 22 Results of the current study align with these findings.

    Strengths and limitations

    This was the largest patient preference study in RA to date, providing a sample size large enough to conduct robust subgroup analyses, with a diverse composition in terms of socioeconomic and clinical characteristics. The study contributes to understanding preferences of patients with RA for DMARDs, with a particular focus on bDMARDs and tsDMARDs (ie, 86% of patients were taking an advanced therapy at the time of the study). The DCE was developed based on best-practice mixed-methods research and provides unique insights into patients’ treatment priorities.

    Despite the advantages of the studies, results must be considered within the context and limitations of the applications. First, by the nature of DCEs, all results are contingent on the considered attributes, which were selected based on qualitative research, clinical data and questions about the effect on semen parameters on treatment decisions. While the consideration of additional attributes may provide a more comprehensive overview of treatments, it may result in overburdening respondents and potential bias from simplifying choice behaviours. Similarly, not all quality-of-life dimensions were captured, as DCEs may specifically not be suitable for assessing psychological elements due to their hypothetical nature (ie, it would require telling participants how to feel). Second, as with most patient preference studies, clinical data were based on self-reports and were not verified by chart reviews. Third, the average age of the sample was lower than that of the general RA population.44 However, the large sample size allowed for a detailed analysis by age group. Fourth, male patients were oversampled in this study to allow for eliciting the relative importance of potential negative effects on semen parameters from DMARD exposure. To test the effect of the oversampling, a model that reweighted the sample composition to one-third male and two-thirds female did not find significant differences in estimates (online supplemental table C.6). Fifth, the data collection was conducted during the COVID-19 pandemic, with unknown effects on patients’ perspectives. For example, blood clot risks were widely discussed in the media as an adverse event of COVID-19 and vaccinations. Sixth, it remains unknown if preferences of patients who participated in this research differed from patients who decided not to take part in the study. Seventh, the risk of negative effects on sperm was specifically included to test its potential impact on treatment decisions in the RA population. Finally, cost was not assessed as willingness-to-pay was considered out of scope for this study.

    Conclusion

    Preferences of patients with RA were driven by multiple benefits and risks of RA treatments, with no single attribute dominating the decision making. Patients were willing to accept higher risks of serious infections and blood clots in exchange for administration convenience, pain relief or improvements in daily functioning. The trade-offs that patients were willing to make were however heterogeneous and varied among individuals and subgroups. Our findings underline the EULAR recommendations on engaging in a shared decision-making process that implies understanding patient preferences and considers patient characteristics as well as the entire treatment profile, including benefits, risks and administration. This comprehensive approach to disease management is essential for optimising patient care. To enable and improve such shared decision-making in routine clinical practice, further research is needed and should consider the development of an engagement process as well as preference-based shared decision-making aids.

    Data availability statement

    Data are available on reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and the study protocol and all patient-facing materials were submitted for review and approval to an independent institutional review board (IRB). Final IRB approval was granted by Ethical & Independent (E&I) Review Services on 13 May 2021 (study number: 20186-01A). Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    Publication coordination was provided by Fabien Debailleul, PhD of Galapagos NV. Medical writing support was provided by Brooke Middlebrook, CMPP (Evidera) and publications management was provided by Aspire Scientific (Bollington, UK), funded by Galapagos NV.

    References

    1. ↵
      1. Smolen JS,
      2. Landewé RBM,
      3. Bergstra SA, et al
      . EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2022 update. Ann Rheum Dis 2023;82:3–18. doi:10.1136/ard-2022-223356
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Garaffoni C,
      2. Adinolfi A,
      3. Bortoluzzi A, et al
      . Novel insights into the management of rheumatoid arthritis: one year in review 2022. Clin Exp Rheumatol 2022;40:1247–57. doi:10.55563/clinexprheumatol/1sjgyr
      OpenUrl
    3. ↵
      1. West-Ward Pharmaceuticals Corp
      . Methotrexate. Eatontown, NJ, 2020.
    4. ↵
      1. Pfizer Inc
      . AZULFIDINE. New York, NY, 2009.
    5. ↵
      1. Akacha A,
      2. Badraoui R,
      3. Rebai T, et al
      . Effect of Opuntia Ficus indica extract on methotrexate-induced testicular injury: a biochemical, docking and histological study. J Biomol Struct Dyn 2022;40:4341–51. doi:10.1080/07391102.2020.1856187
      OpenUrl
    6. ↵
      1. Jensen NB,
      2. Justesen SD,
      3. Larsen A, et al
      . A systematic overview of the spermatotoxic and genotoxic effects of methotrexate, ganciclovir and mycophenolate mofetil. Acta Obstet Gynecol Scand 2021;100:1557–80. doi:10.1111/aogs.14151
      OpenUrl
    7. ↵
      1. Mouyis M,
      2. Flint JD,
      3. Giles IP
      . Safety of anti-rheumatic drugs in men trying to conceive: a systematic review and analysis of published evidence. Semin Arthritis Rheum 2019;48:911–20. doi:10.1016/j.semarthrit.2018.07.011
      OpenUrlPubMed
    8. ↵
      1. Fraenkel L,
      2. Bathon JM,
      3. England BR, et al
      . American College of Rheumatology guideline for the treatment of rheumatoid arthritis. Arthritis Rheumatol 2021;73:1108–23. doi:10.1002/art.41752
      OpenUrl
    9. ↵
      1. Tóth L,
      2. Juhász MF,
      3. Szabó L, et al
      . Janus kinase inhibitors improve disease activity and patient-reported outcomes in rheumatoid arthritis: a systematic review and meta-analysis of 24,135 patients. Int J Mol Sci 2022;23:1246. doi:10.3390/ijms23031246
    10. ↵
      1. Wang Z,
      2. Huang J,
      3. Xie D, et al
      . Toward overcoming treatment failure in rheumatoid arthritis. Front Immunol 2021;12:755844. doi:10.3389/fimmu.2021.755844
      OpenUrlCrossRef
    11. ↵
      1. Min HK,
      2. Kim SH,
      3. Kim H-R, et al
      . Therapeutic utility and adverse effects of biologic disease-modifying anti-rheumatic drugs in inflammatory arthritis. Int J Mol Sci 2022;23:13913. doi:10.3390/ijms232213913
    12. ↵
      1. Riley TR,
      2. George MD
      . Risk for infections with glucocorticoids and DMARDs in patients with rheumatoid arthritis. RMD Open 2021;7:e001235. doi:10.1136/rmdopen-2020-001235
    13. ↵
      1. European Medicines Agency
      . Qualification Opinion of IMI PREFER. 2022.
    14. ↵
      1. FDA
      . Benefit-Risk Assessment for New Drug and Biological Products: Draft Guidance. Center for Drug Evaluation and Research, 2021.
    15. ↵
      1. FDA
      . Patient preference information—voluntary submission, review in premarket approval applications, humanitarian device exemption applications, and de novo requests, and inclusion in decision summaries and device labeling. 2016.
    16. ↵
      1. Ho KA,
      2. Acar M,
      3. Puig A, et al
      . What do Australian patients with inflammatory arthritis value in treatment? A discrete choice experiment. Clin Rheumatol 2020;39:1077–89. doi:10.1007/s10067-019-04843-4
      OpenUrlPubMed
    17. ↵
      1. van Heuckelum M,
      2. Mathijssen EG,
      3. Vervloet M, et al
      . Preferences of patients with rheumatoid arthritis regarding disease-modifying Antirheumatic drugs: a discrete choice experiment. patient prefer adherence 2019;13:1199–211. Patient Prefer Adherence 2019;13:1199–211. doi:10.2147/PPA.S204111
      OpenUrlPubMed
    18. ↵
      1. Bywall KS,
      2. Kihlbom U,
      3. Hansson M, et al
      . Patient preferences on rheumatoid arthritis second-line treatment: a discrete choice experiment of Swedish patients. Arthritis Res Ther 2020;22:288. doi:10.1186/s13075-020-02391-w
      OpenUrlPubMed
    19. ↵
      1. Díaz-Torné C,
      2. Urruticoechea-Arana A,
      3. Ivorra-Cortés J, et al
      . What matters most to patients and rheumatologists? A discrete choice experiment in rheumatoid arthritis. Adv Ther 2020;37:1479–95. doi:10.1007/s12325-020-01258-5
      OpenUrlPubMed
    20. ↵
      1. Martin RW,
      2. McCallops K,
      3. Head AJ, et al
      . Influence of patient characteristics on perceived risks and willingness to take a proposed anti-rheumatic drug. BMC Med Inform Decis Mak 2013;13:89. doi:10.1186/1472-6947-13-89
    21. ↵
      1. Durand C,
      2. Eldoma M,
      3. Marshall DA, et al
      . Patient preferences for disease-modifying antirheumatic drug treatment in rheumatoid arthritis: a systematic review. J Rheumatol 2020;47:176–87. doi:10.3899/jrheum.181165
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Simons G,
      2. Caplan J,
      3. DiSantostefano RL, et al
      . Systematic review of quantitative preference studies of treatments for rheumatoid arthritis among patients and at-risk populations. Arthritis Res Ther 2022;24:55. doi:10.1186/s13075-021-02707-4
    23. ↵
      1. Bridges JFP,
      2. Hauber AB,
      3. Marshall D, et al
      . Conjoint analysis applications in health—a checklist: a report of the ISPOR good research practices for conjoint analysis task force. Value Health 2011;14:403–13. doi:10.1016/j.jval.2010.11.013
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Reed Johnson F,
      2. Lancsar E,
      3. Marshall D, et al
      . Constructing experimental designs for discrete-choice experiments: report of the ISPOR conjoint analysis experimental design good research practices task force. Value in Health 2013;16:3–13. doi:10.1016/j.jval.2012.08.2223
      OpenUrl
    25. ↵
      1. Ryan M,
      2. Watson V,
      3. Entwistle V
      . “Rationalising the ‘irrational’: a think aloud study of discrete choice experiment responses”. Health Econ 2009;18:321–36. doi:10.1002/hec.1369
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Chew LD,
      2. Bradley KA,
      3. Boyko EJ
      . Brief questions to identify patients with inadequate health literacy. Fam Med 2004;36:588–94.
      OpenUrlPubMedWeb of Science
    27. ↵
      1. Fransen MP,
      2. Van Schaik TM,
      3. Twickler TB, et al
      . Applicability of internationally available health literacy measures in the Netherlands. J Health Commun 2011;16 Suppl 3:134–49. doi:10.1080/10810730.2011.604383
      OpenUrlPubMed
    28. ↵
      1. Lipkus IM,
      2. Samsa G,
      3. Rimer BK
      . General performance on a numeracy scale among highly educated samples. Med Decis Making 2001;21:37–44. doi:10.1177/0272989X0102100105
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Rose JM,
      2. Bliemer MCJ
      . Constructing efficient stated choice experimental designs. Transport Reviews 2009;29:587–617. doi:10.1080/01441640902827623
      OpenUrlCrossRefWeb of Science
    30. ↵
      1. Bliemer MCJ,
      2. Rose JM,
      3. Hess S
      . Approximation of Bayesian efficiency in experimental choice designs. J Choice Model 2008;1:98–126. doi:10.1016/S1755-5345(13)70024-1
      OpenUrl
    31. ↵
      1. Carlsson F,
      2. Mørkbak MR,
      3. Olsen SB
      . The first time is the hardest: a test of ordering effects in choice experiments. J Choice Model 2012;5:19–37. doi:10.1016/S1755-5345(13)70051-4
      OpenUrl
    32. ↵
      1. Johnson FR,
      2. Yang J-C,
      3. Reed SD
      . The internal validity of discrete choice experiment data: a testing tool for quantitative assessments. Value Health 2019;22:157–60. doi:10.1016/j.jval.2018.07.876
      OpenUrlCrossRefPubMed
    33. ↵
      1. Lancsar E,
      2. Louviere J
      . Deleting ‘irrational’ responses from discrete choice experiments: a case of investigating or imposing preferences Health Econ 2006;15:797–811. doi:10.1002/hec.1104
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Hess S,
      2. Palma D
      . Apollo: A flexible, powerful and customisable freeware package for choice model estimation and application. J Choice Model 2019;32:100170. doi:10.1016/j.jocm.2019.100170
      OpenUrl
    35. ↵
      1. McFadden D,
      2. Train K
      . Mixed MNL models for discrete response. J Appl Econ 2000;15:447–70. doi:10.1002/1099-1255(200009/10)15:5<447::AID-JAE570>3.0.CO;2-1
      OpenUrlCrossRef
    36. ↵
      1. Revelt D,
      2. Train K
      . Mixed logit with repeated choices: households’ choices of appliance efficiency level. Rev Economic Stat 1998;80:647–57. doi:10.1162/003465398557735
      OpenUrl
    37. ↵
      1. Hess S,
      2. Train K
      . Correlation and scale in mixed logit models. J Choice Model 2017;23:1–8. doi:10.1016/j.jocm.2017.03.001
      OpenUrl
    38. ↵
      1. Hole AR
      . A comparison of approaches to estimating confidence intervals for willingness to pay measures. Health Econ 2007;16:827–40. doi:10.1002/hec.1197
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. Reinisch W,
      2. Hellstrom W,
      3. Dolhain RJEM, et al
      . Effects of filgotinib on semen parameters and sex hormones in male patients with inflammatory diseases: results from the phase 2, randomised, double-blind, placebo-controlled MANTA and MANTA-ray studies. Ann Rheum Dis 2023;82:1049–58. doi:10.1136/ard-2023-224017
      OpenUrlAbstract/FREE Full Text
    40. ↵
      1. Holdsworth EA,
      2. Donaghy B,
      3. Fox KM, et al
      . Biologic and targeted synthetic DMARD utilization in the United States: Adelphi Real World Disease Specific Programme for rheumatoid arthritis. Rheumatol Ther 2021;8:1637–49. doi:10.1007/s40744-021-00357-1
      OpenUrl
    41. ↵
      1. Senbel E,
      2. Durand F,
      3. Roux B, et al
      . Elicitation of rheumatologist preferences for the treatment of patients with rheumatoid arthritis after the failure of a first conventional synthetic disease-modifying anti-rheumatic agent. Rheumatol Ther 2021;8:921–35. doi:10.1007/s40744-021-00311-1
      OpenUrl
    42. ↵
      1. Mathijssen EG,
      2. van Heuckelum M,
      3. van Dijk L, et al
      . A discrete choice experiment on preferences of patients with rheumatoid arthritis regarding disease-modifying antirheumatic drugs: the identification, refinement, and selection of attributes and levels. Patient Prefer Adherence 2018;12:1537–55. doi:10.2147/PPA.S170721
      OpenUrl
    43. ↵
      1. Alten R,
      2. Krüger K,
      3. Rellecke J, et al
      . Examining patient preferences in the treatment of rheumatoid arthritis using a discrete-choice approach. Patient Prefer Adherence 2016;10:2217–28. doi:10.2147/PPA.S117774
      OpenUrlPubMed
    44. ↵
      1. Bajraktari IH,
      2. Teuta B-Ç,
      3. Vjollca S-M, et al
      . Demographic features of patients with rheumatoid arthritis in Kosovo. Med Arch 2014;68:407–10. doi:10.5455/medarh.2014.68.407-410
      OpenUrl

    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.

    Footnotes

    • Twitter @JCNietoGon

    • Contributors KVB, SH, CW and NK were responsible for the planning and conduct of the work described in this manuscript. SH, CW and NK were responsible for data collection and analysis. All authors contributed to the writing and review of this manuscript. KVB is responsible for the overall content as guarantor.

    • Funding This study was funded by Galapagos NV (Mechelen, Belgium).

    • Competing interests RA, consultancy fees from AbbVie, Amgen, Biogen, Bristol Myers Squibb, Celltrion, Gilead, Janssen, Lilly, Medac, MSD, Mylan, Novartis, Pfizer, Roche, Sandoz, Sanofi-Genzyme, UCB, VIATRIS. JCNG, consultancy fees from Lilly, Janssen, Amgen, GSK, AbbVie, Galápagos, MSD; speakers fees from MSD, Pfizer, BMS, AbbVie, UCB Pharma, Janssen, Lilly, Faes Farma, Roche, Celgene, Sanofi, Nordic Pharma, Gebro, Novartis, Biogen, Amgen, Sandoz. PJ, grants support from Pfizer and Roche; speakers fees from Eli Lilly; support for meeting attendance from Galapagos. CM, consultancy fees from BMS, AbbVie, Gilead; speakers fees from BMS, AbbVie, Eli Lilly, Pfizer, Boehringer, Galapagos, Sanofi, Roche. RM, grant support from Novartis; consultancy fees from Ferring; speakers fees from Amgen and Galapagos. HR, speakers fees from Gilead Science, Merck Sharp; Pfizer Cooperation Austria, Janssen. HEV, grants from Galapagos, Boehringer Ingelheim; speakers fees from Galapagos, AbbVie, Boehringer Ingelheim, Novartis, Pfizer, UCB, Janssen. SH, employee of Evidera Inc, which is part of Thermo Fisher Scientific's Clinical Research Group; Evidera received payment for conducting the work outlined in this manuscript; SH is a minority stockholder of Thermo Fisher Scientific, as part of his employment with Evidera. CW, employee of Evidera Inc, which is part of Thermo Fisher Scientific's Clinical Research Group; Evidera received payment for conducting the work outlined in this manuscript. NK, employee of Evidera Inc, which is part of Thermo Fisher Scientific's Clinical Research Group; Evidera received payment for conducting the work outlined in this manuscript; NK is a minority stockholder of Thermo Fisher Scientific, as part of his employment with Evidera. KVB, employee and shareholder of Galapagos NV.

    • 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.