Pragmatic Comparative Effectiveness Trials in Rheumatoid Arthritis-Associated Interstitial Lung Disease

Abstract

BACKGROUND:

Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) portends a devastating prognosis for patients, with survival typically being < 5 to 8 years after diagnosis. Limited clinical trial data exist to guide treatment strategies, and the efficacy of current strategies—immunomodulation and antifibrotics—remains uncertain. Large randomized controlled trials are costly, but pragmatic trial designs could reduce expenses. Establishing equipoise and assessing feasibility from both patient and expert perspectives are essential for developing these trials.

RESEARCH QUESTION:

What are the perceptions of RA-ILD experts and patients with interstitial lung disease surrounding equipoise, feasibility, and trial design?

STUDY DESIGN AND METHODS:

A qualitative study involving a panel of 10 RA-ILD experts and 3 patient panels was conducted. Experts were recruited via snowball sampling, and patient panels included 29 individuals with interstitial lung disease or their caregivers. Discussions were transcribed and analyzed using inductive coding, creating a thematic network based on the Attride-Stirling guidelines.

RESULTS:

Expert themes included variability in treatment strategies, prioritizing patient-reported outcomes and balancing pragmatism with data collection in trial design. Patient themes highlighted outcomes of importance, participation barriers, and the need for patient-centered research.

INTERPRETATION:

Both expert and patient panels endorsed using real-world clinical outcomes and patient-reported outcomes as primary trial end points. Pragmatic trials could reduce costs and expand inclusion criteria, highlighting the potential of patient-centered approaches in RA-ILD research.

In rheumatoid arthritis-associated interstitial lung disease (RA-ILD), there is a paucity of clinical trial data to inform treatment decisions. Patients with RA-ILD typically survive < 5 to 8 years after diagnosis; therefore, it is imperative to optimize treatment selection during this narrow therapeutic window.¹ Establishing the effectiveness of treatments is important given that the standard of care therapy for RA-ILD, immunomodulation, has been shown to increase harm in idiopathic pulmonary fibrosis, a fibrotic lung disease similar to RA-ILD.^2,3 There has been increasing use of antifibrotic medications in patients with RA-ILD.⁴ The The Treatment for Rheumatoid Arthritis and Interstitial Lung Disease 1 (TRAIL1) trial of pirfenidone in RA-ILD produced similar results as has been seen in systemic sclerosis-associated interstitial lung disease (ILD), idiopathic pulmonary fibrosis, and nonidiopathic forms of progressive pulmonary fibrosis: reduced decline in FVC without improvements in quality of life or survival.^4–7 The potential of harm from widespread clinical application of unstudied treatment methods, coupled with financial burdens of treatments lacking in definitive efficacy, underscores the need to study these treatment modalities to determine their clinical effectiveness in RA-ILD.

One of the primary obstacles to generating randomized trial data in RA-ILD has been the cost of large, high-quality studies. Costs are further exacerbated by the relative rarity of RA-ILD, which increases the required number of enrolling centers. These financial barriers have created a clinical trial environment primarily driven by industry-funded phase III studies of novel therapies or novel indications for existing therapies to prove fundamental efficacy in all forms of autoimmune ILD. In RA-ILD, it is imperative that we understand the effectiveness, and safety, of the 2 strategies that are already used in clinics every day for patients (immunomodulatory and antifibrotics) in addition to continuing to support a portfolio of novel therapies that may outperform our current treatment options. One method for lowering per patient costs associated with randomized trials is to implement pragmatic clinical trial designs.

A pragmatic trial design focused on comparative effectiveness of 2 strategies allows for decreased costs by leveraging data for primary and secondary outcomes (eg, pulmonary function test [PFT] values, radiographic features from routine clinical care), thereby reducing the costs of trial administration.^8,9 Additionally, a primary feature of pragmatic trial design is the application of inclusion and exclusion criteria that more closely mimic real-world settings, increasing the number of eligible participants and improving the generalizability of the findings.¹⁰

Establishing underlying equipoise is essential in planning clinical trial study design; equally important is evaluating the balance of trial design tradeoffs and study feasibility based on the perspectives of key stakeholders. To take this important step in trial planning, we convened a panel of expert clinicians and clinician scientists who treat patients with RA-ILD followed by small-group patient discussion panels to assess these important questions regarding pragmatic study design for patients with RA-ILD.

Study Design and Methods

Participants

Expert Panel:

Ten clinicians and clinician scientists were convened virtually. Clinicians and clinician scientists who care for patients with RA-ILD in their routine practice were invited to participate via email invitations and word-of-mouth. All participants in the expert panel had relevant credentials in either pulmonology or rheumatology, current experience treating patients with RA-ILD, and experience with clinical trial design.

Patient Panel:

Patients with RA-ILD were approached by the research team (A. K.-W. and S. M. M.) in ILD clinics at our institution to assess their interest in participating in the discussion. Patients with ILD were also approached at a local pulmonary fibrosis and ILD support group sponsored by the Pulmonary Fibrosis Foundation. Inclusion criteria consisted of patients diagnosed with ILD or their caregivers. Investigators described their study in an institutional review board-approved email which was sent to each support group member before 1 of their monthly meetings and included a copy of the informed consent form to review.

Study Setting/Context

We convened a panel of 10 clinicians and clinician scientists including pulmonologists, rheumatologists, and an advanced practice nurse provider all with clinical trial design experience and expertise in the diagnosis and management of RA-ILD. The panel met remotely for a 2-hour moderated discussion. The moderator of the discussion, also an ILD specialist (S. M. M.), used a preset template (e-Appendix 1) to introduce the concept and the points of discussion but allowed for free-flowing discussion and follow-up questions based on the direction of the discussion. Each provider signed consent via REDCap before logging into the virtual meeting. The discussions were recorded with permission from participants.

Researchers (A. K.-W., S. M. M., and J. R.) attended the monthly support group meeting. At the beginning of the meeting, the investigative team again described their research plan and the goal of eliciting patient perspectives on clinical trial design. Then support group members were invited to ask questions, and if they were interested in participating, to sign the provided informed consent. Three separate small group discussions of 8 to 11 patients (29 in total) were led by research team members (A. K.-W., J. R., and S. M. M.) based on a shared outline (e-Appendix 2), allowing for changes in topic led by direction of the group’s discussion.

Data Collection

We conducted a qualitative study on RA-ILD study designs through 2 small group discussions: 1 expert panel and a patient panel. Each small group discussion was recorded and then transcribed from audio files into text documents using an artificial intelligence transcription software (OtterAI); transcription accuracy was verified and edited by 2 authors (A. K.-W. and S. M. M.).

This study was approved by the University of Kansas institutional review board on April 4, 2023.

Before the expert panel meeting, the research team reviewed guiding questions based on existing challenges in RA-ILD clinical trials based on a literature review, the need to determine equipoise, and to assess the expert consensus on feasibility of pragmatic, decentralized clinical trials in RA-ILD (Table 1). To enable these guiding questions, the moderator (S. M. M.) presented an introductory overview of the current research landscape in RA-ILD and framed several of the important unanswered questions in the management of patients with RA-ILD before introducing the focus of the panel on equipoise and feasibility of future pragmatic trials for patients with RA-ILD.

TABLE 1 ].

Guiding Questions and Panel Discussion Quotations

Guiding Questions	Panel Discussion Quotations
Equipoise  When you see a new patient with RA-ILD in clinic with a fibrotic ILD pattern, assuming they have adequate joint (RA disease) control, what is your typical first treatment strategy (immunomodulation or antifibrotic)?	“I’m going to put this patient on CellCept (mycophenolate).” “We just had this long discussion about this yesterday at our MDD for a patient who had connective tissue ILD, and what do you start with? I’d say if they have a UIP pattern, meaning fibrosis is their big problem, I would actually put them on an antifibrotic, (and) I’d probably put them on OFEV (nintedanib).” “I agree with [redacted], I think I would probably add an antifibrotic at that point if they’re already on immunosuppression with stable joint symptoms.” “(M)y practice is always to take a more like a holistic approach. So, if you can use 1 medication and get away with it, why introduce 2 where you could have more toxicity? So, I would probably, and what we generally do in this kind of scenario at (redacted), is transition to CellCept and monitor for articular disease flares. RA-ILD is still an autoimmune disease and when there’s likely a component that will respond to kind of immunomodulation, I don’t think that there’s strong enough evidence yet for the antifibrotic as a primary treatment strategy.”
Equipoise  In the context of a clinical trial, what patient-specific features would make you uncomfortable to randomize a patient with RA-ILD between these 2 strategies (immunomodulation or antifibrotic)?  Which RA-ILD disease features would be necessary to consider in a stratified randomization schema? For instance, if UIP is unequally distributed in the 2 arms, would the results be internally valid? Or, if RA disease activity is unbalanced in the 2 arms, would the study have internal validity?	“I agree that there’s equipoise. I think whoever takes a side in this, it’s hard to defend your side against the other. So, I think I’d feel comfortable randomizing patients to either 1. The difficulty is going to be when you have someone who’s having worsening fibrosis on immunomodulation strategy, it’s hard to not put those patients on add-on antifibrotic.” “One of the things that makes this space difficult to study is there are so many considerations about their baseline therapy. So, I think the idea of randomizing patients to add something from within 1 of these categories allows for typical clinical freedom to adjust baseline regimens and some needed flexibility.” “I feel like the decision in practice is much more nuanced for the biologic agents, because what’s been highlighted so far does not really include decision making factors such as adverse event profiles for the biologic agents, drug-drug interactions and access to insurance. So practically, these are things we’ll consider before starting patients on any of these medications. If it’s worse than fibrosis, then we may lean towards the antifibrotic. If it’s worse ground glass, for whatever reason, then we might lean towards more immunomodulatory therapy. As a pulmonologist, that’s what we often do, so I think it’s just much more nuanced than black and white like, antifibrotic versus immunomodulation.” “I find myself thinking about why that patient is progressing? Are we in a proinflammatory profile that’s ultimately leading to progressive fibrosis? And therefore, is attenuation of (inflammation) going to be the key to prevent progression? Or have we transitioned into the profibrotic pathway? So, to echo the initial question, in terms of would I feel comfortable randomizing? I think the answer is yes, because I think we just don’t understand very well how to identify which of these spectrums an individual patient is on.” “I also see many patients referred to me with RA that’s under ‘good control’ with an undetectably high anti-CCP [titer] and will have what seems like very active joint disease, but I would be hesitant to put on just 1 or the other drug class if they have UIP based on progression of those patients from the INBUILD data.” “One option to address those concerns would be to stratify randomization based on inflammatory markers or disease activity score.” “RA is one of the few diseases in rheumatology, where we do have remission criteria, the research team could simply perform a DAS (disease activity score), which combines ESR or CRP with a tender swollen joint count. In that case, those patients who are considered in remission by their DAS score would be eligible for randomization between these strategies and that would be more real world in my opinion.”
Feasibility  What outcomes would be important for any clinical trial of therapy in RA-ILD?  How frequently do you assess PFTs and CT scans after initiating treatment for patients with RA-ILD in your routine clinical care?	“Even though this is planned to be pragmatic, I think it is important to include a patient reported outcome (PRO), if you think about the recent PPF definition, it includes respiratory symptoms, and it might be useful to capture that.” “I think that will be very challenging to assess across [the] board in a pragmatic study and what outcome would you pick that’s been validated in the target population for the medication of interest? But I think it’s very important, because I think that’s the way forward in clinical trials and it’s more practical and more meaningful to patients.” “I think it would be easy enough to have a battery, whatever is decided of PROs that patients fill out time to time, including at baseline where they can at least assess their own joint symptom severity as well. Maybe there’s nothing solidly fit for the purpose of assessing symptoms and quality of life in RA-ILD, but I think this would provide a really nice opportunity to gather some really important data.” “In general, we obtain spirometry every 3 months and most of us get a follow up CT at 6 months or 12 months.” “Due to the cost of all these tests, we typically get spirometry, ideally, every 3 to 4 months. But I don’t necessarily repeat CT scans because of trying to be sensitive to cost issues and patient co-pays. I doubt we would repeat a CT more than once a year on most patients without a clinical indication like a nodule that required follow-up.” “CT is more variable, depending on the circumstance with the patients. We usually get spirometry every 3 to 4 months given the recent PPF guidelines.” “I think 1 CT scan per year is very reasonable. We do quantitative image analysis on our regular standard of care patients, and we see changes as early as 6 months. So, we do that to assess treatment response. But I think for the purposes of this study, it would be more feasible to just do it at 1 year, since you’re not going to be paying for the CT scans.”

CCP = citrullinated protein antibodies; CRP = C-reactive protein; DAS = disease activity score; ESR = erythrocyte sedimentation rate; ILD = interstitial lung disease; MDD = Multidisciplinary Discussion; PFT = pulmonary function test; PPF = Progressive Pulmonary Fibrosis; PRO = patient reported outcome; RA = rheumatoid arthritis; RA-ILD = rheumatoid arthritis-associated interstitial lung disease; UIP = Usual Interstitial Pneumonia.

Data Analysis

For the provider expert panel, the research team (A. K.-W. and S. M. M.) extracted overall themes from the transcription of the discussion and summarized each point in tabular format using grounded theory coding. A priori categories for summarizing the discussion included the areas of equipoise, opportunities for trial design, open questions that need to be answered, and strengths and weaknesses of the proposed trial approach.

Given the larger number of participants and broader discussions in the patient perspective panels compared with the expert panel discussion, the team elected to generate a thematic network analysis for the patient perspective themes. To extract general themes and thematic concepts from the patient perspective discussions, 2 research team members (D. P. M. and A. K.-W.) performed inductive coding on grounded theory principles to perform thematic network analysis from the patient panels.¹¹ These coauthors (D. P. M. and A. K.-W.) discussed the codebook, resolving differences in their understandings of definitions via consensus. Two authors (A. K.-W. and D. P. M.) coded the 3 small group discussions followed by another author (D. P. M.) to establish consensus and code saturation.

Results

Expert Panel

The expert provider panel included predominantly providers associated with pulmonary clinics and predominantly physicians with 1 advanced practice registered nurse. All providers cared for patients in academic or academic-affiliated medical centers and had an average ± SD of 7.4 ± 6.7 years of experience in independently managing patients with RA-ILD (Table 2).

TABLE 2 ].

Expert Provider Panel Demographics (n = 10)

Characteristic	Value
Gender, women	3 (30)
Specialty
Rheumatologist	1 (10)
Pulmonologist	9 (90)
Credentials
APRN	1 (10)
MD/DO	9 (90)
Years in ILD clinical practice	7.4 [6.7]
Practice setting
Academic medical center	8 (80)
Academic/private hybrid model	2 (20)
Location of practice in United States
West	3 (30)
Midwest	6 (60)
East	0 (0)
South	1 (10)
Academic rank
N/A	1 (11.1)
Assistant professor	6 (66.6)
Associate professor	2 (22.2)
Professor	1 (11.1)

Values are presented as mean [SD] or No. (%). APRN = advanced practice registered nurse; ILD = interstitial lung disease; MD/DO = Medical Doctor or Doctor of Osteopathic Medicine; N/A = not applicable.

The expert panel was not designed to develop consensus on these topics; the focus was to foster an open discussion and a discussion of each of these points to inform future study design. Table 1 highlights several quotations from members of the expert panel for each of the discussion points based on these guiding questions focused on equipoise and feasibility. Key themes included conflicting use of immunomodulatory vs antifibrotic initial treatment strategies, the need to consider background disease state and prior treatments, including patient-reported outcomes for respiratory symptoms alongside typical measures of PFTs and chest CT scans, and reducing the frequency of PFT and chest CT testing from what is typically done in ILD clinical trials. Based on responses to the guided questions, the expert panel identified priority areas for consideration when designing pragmatic trials in RA-ILD (Table 3).

TABLE 3 ].

Expert Panel Areas of Particular Focus for Pragmatic Trial Designers in RA-ILD

Expert Panel Areas	Summary
RA disease activity	A trial of RA-ILD therapy that is unable to reliably capture baseline and longitudinal RA disease activity will have limited external validity due to the lack of expert comfort in randomizing patients with high RA activity. A trial of RA-ILD without measures of RA disease activity will have limited internal validity if randomization cannot be balanced for RA disease activity scores.
Crossover contamination	The expert panel highlighted the need to consider how a research team would handle crossover in study design between the 2 treatment strategies. The balance should consider the need to maintain separation between experimental groups while accounting for routine clinical care decision-making and patient safety given the current state of care in RA-ILD where both strategies are often used in sequence.
ILD pattern	The expert panel highlighted the need to balance UIP pattern disease in both groups to maintain internal validity.
Patient-specific features that predict treatment response	One feature that true pragmatic studies would lack is the prospective collection of RA-ILD patient biospecimens. Traditionally designed clinical trial design would be well positioned to leverage baseline patient samples to determine molecular predictors of treatment response in each arm. The expert panel thought that a research team designing a pragmatic study should consider the balance between pragmatism and the collection of important data and samples that could inform these important clinical questions.
Outcomes	The expert panel thought that collection of patient-reported outcomes was a fundamental element of study design that would need to be accounted for in a pragmatic study design. The lack of validated patient-reported outcome instruments for RA-ILD and these specific treatments will require attention in study design. Focusing on validation of patient-reported outcomes in this study could contribute to an important contribution to the field regardless of the primary outcomes of the study.
Treatment delivery	A fully pragmatic study design would allow providers, after randomization, to choose therapies within the strategy they were assigned (antifibrotic: nintedanib vs pirfenidone) and then leverage real-world prescribing methods to get the medications to patients. However, there are barriers to receiving costly therapies specific to each patient related to copays, provider comfort, and prior authorization time frames that may influence the likelihood of the patient remaining within the randomized strategy that will limit the internal validity which will need to be addressed in study design.

ILD = interstitial lung disease; RA = rheumatoid arthritis; RA-ILD = rheumatoid arthritis-associated interstitial lung disease; UIP = Usual Interstitial Pneumonia.

Patient Panel

In the patient small group discussions, there were 29 participants with ILD and their caregivers (Table 4). The patient small groups provided insight into patient perspectives on important outcomes for clinical trials in RA-ILD, and incentives and barriers to clinical trial involvement. Additionally, our study explored patient perspectives on patient-centered outcomes in RA-ILD clinical trials. The coding and identification of themes are highlighted in Table 5. Themes were arranged into the thematic network (Fig 1) based on the Attride-Stirling guidelines for thematic network diagrams to describe thematic relationships.¹²

TABLE 4 ].

Patients With ILD Small Group Member Demographics (n = 29)

Characteristics	Value
Gender, women	11 (38)
Currently on antifibrotics	12 (41)
Currently on immunomodulation	15 (52)
Years with ILD	2.4 [3.2]
Provider locationa
Academic medical center	17 (59)
Community clinic	20 (69)
Highest degree obtained
High school or equivalent	9 (31)
Some college or associates	8 (28)
Bachelor’s degree	8 (28)
Master’s degree	3 (10)
Doctorate degree	1 (4)

Values are presented as mean [SD] or No. (%). ILD = interstitial lung disease.

^aPatients could have both academic and community providers.

TABLE 5 ].

ILD Patient Small Group Discussion on Reseach Focus and Priorities

Codes	Issues Highlighted	Themes Identified
Patient-centered barriers to participation
Side effects	Side effects represent a significant concern	Interventions have potential to decrease QOL in patients with terminal diagnosis
Cost	Cost/travel/time represent a significant burden	Cost can be high and benefit is unknown for patients participating in research
Risk of randomization to placebo	Risks involved with new treatments and risk of no treatment at all	Potential for harm from participating
Physician-centered barriers to clinical trial enrollment
Lack of knowledge	Coordination between centers is difficult	Shared decision-making: patient relies on physicians to present information on available trials
Lack of advertisement
Lack of engagement	Physician shortage especially in rare diseases	Patient value and expect interfacility communication and collaboration between local and specialty doctors
Access to physician
General barriers
Inclusion/exclusion criteria	Older population with multiple comorbidities	General barriers can lead to a lack of external validity in clinical trials and make pragmatic trials difficult to develop
Access to supplements
Lack of research studies	Rare diseases
Distrust of pharmaceutical companies
Lack of understanding of pragmatic trials
Benefits		Pragmatic trials are relatively unknown
Challenges
Confirming info
Logistics
Existing treatment options	Concern that PT does not lead to novel treatments
Patient-centered research
Focus on genetics	Desire to protect family	Patient generally wants research to not only benefit themselves, but their families
Focus on risk factors
Partnership/coordination	Desire for education/information about PF	Transparency and informed consent valued
Focus on personalized medicine
Symptoms	Results should show clear benefit
Tracking activity/data	Disease progression is unpredictable and difficult to monitor	Surrogate outcomes must reflect data that are relevant to patients
Spirometry	Measurements do not accurately reflect symptoms/QOL
Side effects, placebo	Unexpected costs of participation	For this terminal condition, time is an exceedingly valuable commodity. Side effects and delays in treatment result in decreased QOL
Incentive behind participation in research
Access to specialty physicians
Altruism
Access to information about disease
Hope for new drug/cure
QOL/PRO
ADLs, IADLs		QOL measures are not conventionally included in clinical trials primary outcomes
Dissatisfaction with current and projected QOL
Exercise/activity intolerance		Clear benefit should reflect improvements/preservation of QOL
Impact on relationships
Impact on sleep
Inability to travel/fear of traveling

Blank cells denote no discussion of the theme. ADL = activity of daily living; IADL = instrumental activity of daily living; PF = Pulmonary Fibrosis; PT = Pragmatic Trial; PRO = patient-reported outcome; QOL = quality of life.

Figure 1–

Organizing themes for pragmatic rheumatoid arthritis-associated interstitial lung disease trials from the patient panels.

Organizing Themes From Patient Panels

Organizing Theme 1:

Patient-Centered Barriers to Clinical Trial Involvement: Many patients noted the time and cost required to travel to the nearest academic medical center on a frequent basis to be involved in clinical trials and the barriers from their lung and systemic diseases that made such travel difficult. The risk of randomization to a placebo was a primary concern for many patients in clinical trial involvement. There was also concern regarding randomization to an untested drug and the potential of harm and side effects that led many patients in our group to be wary of clinical trial involvement.

Organizing Theme 2:

Incentives for Patient Participation in Research: The predominant incentives for involvement in clinical research from the patient perspective were access to novel therapies, access to specialty care more frequently through the study visits, and altruism for others with the same condition in the future.

One patient discussed the patient perspective on enrollment in comparative effectiveness studies vs typical phase II or III trials of new and novel compounds: “The challenge from my perspective, you’ve got all these drugs that all these companies are working on, and they’re sort of competing for patients with maybe more state-of-the-art drugs right? Now you’re taking the old thing that’s been out for a while and saying, hey, it might work. For me, I want to go to the new drug.”

Organizing Theme 3:

Physician-Centered Barriers to Clinical Trial Involvement: Patients perceived barriers to clinical trial enrollment based on their providers. In many cases, patients were unable to be seen in academic medical centers despite having been referred, from their perspective, to discuss research trial options. Patients thought most providers and clinics did not have access to clinical trials and were unaware of existing studies for patients like them. One patient stated, “So, it’s kind of in your hands, I thought the doctor would send my name to the research department and I would just wait for a call, but now I see it’s kind of like, in your hands to call and say, I’m interested, what do you got? Then I had to wait just to get seen by the doctor at the research department and things were getting worse the whole time.”

Organizing Theme 4:

Lack of Understanding of Pragmatic Clinical Trials: Pertinent to comparative effectiveness trials, the patient perspective was primarily focused on the lack of understanding the specifics of this study design, where most patients were familiar with traditional placebo trials of new and novel compounds. Many patients expressed skepticism about the benefits of this approach given that these strategies are already available to patients and their providers. For example, “I have a question. These would be studies for existing drugs? For RA-ILD? Why aren’t we taking them already if there’s something that’s beneficial? If you get on 1 particular drug, and it’s not working for you, do you get to switch and get on the other 1? Or are you locked into this one?”

Organizing Theme 5:

Quality of Life and Patient-Reported Outcomes: Most of the discussion of patient-centered outcomes noted the ways in which patient’s lung disease had impacted their life. The themes were loss of ability to travel, inability to participate in daily activities of living, the impact on relationships because of their symptom burden and difficulty leaving home, and loss of exercise tolerance. One patient stated the following:

Could the trial perhaps provide pedometers or oximeters initially when they were putting it to trial if they didn’t have them or ask them to do a log or a journal? Because, if we look at, say, going up and down the stairs and use the guidelines like we use in pulmonary rehab, using pulmonary exercise to see how difficult this activity and how much exertion is it. We use those scales of 1 to 20, or 1 to 10 range. If we use those scales, we can then see, at this point in my life right now, this is where I meant doing that certain activity, here’s where I was 6 months ago, or when you call us back 6 months from now you can see if there’s a change.

In terms of the use of PFTs as the primary outcomes of trials, the patient perspective supported the use of patient symptoms or survival. For instance, “Because it can vary. If I were to have it done. Once a day for 3 days, it would vary greatly. Yeah. I think it depends on what was my quality of sleep like that night. Did I have a lot of stairs to climb? Did I have a long distance to walk just to get into the doctor? And then have the test done? Yeah, I think it’s, it’s not real accurate.”

Organizing Theme 6:

Patient-Centered Research: Several participants pointed out their hope that this type of research could identify specific factors that would predict treatment response. Additionally, many participants reiterated their desire for precision medicine in RA-ILD with a particular focus on genetic risk. For instance,

Yeah, there’s a couple of things that I am really curious about, that I think could be breakthroughs for us. One is the approach to the research. I’m not an expert at this, because I’m not a scientist or a doctor. I read about this very specific, almost down to the individual patient approach where they, you know, they’ve got your gene analysis and everything else and they can they can pinpoint like a cancer right now to a specific gene for a person. If you both are diagnosed with the same thing, you’re gonna get two different forms of treatment because of who you are, and your makeup and your genome, and I can I think there’s a lot of opportunity that we’re just not spending money on.”

Organizing Theme 7:

General Barriers: The final organizing theme was the description of general barriers to clinical trials in RA-ILD globally. Patients thought that compared with other diseases, pulmonary fibrosis and rheumatoid arthritis more specifically had very few clinical trials to participate in at all. Several patients had been interested in clinical trials but when they asked their providers, they were told they did not meet the inclusion or exclusion trial criteria due to the nature of their diagnosis, the severity of their lung condition, comorbid conditions, or age. Finally, there was a distrust of pharmaceutical companies’ ability to put patient well-being ahead of profit.

Discussion

This qualitative study collected expert clinician and patient perspectives on pragmatic comparative effectiveness trials in RA-ILD. The focus of this work was to systematically gather these impressions, assemble them into informative themes, and incorporate discussions with these important stakeholders into study design at the earliest stages. It is imperative that comparative effectiveness studies involve patients in the process of study design to align measured outcomes with patient perspectives. Finally, before investing invaluable patient time in a study, it is vital for the research team to explore equipoise, to explore feasibility, and to gather wide ranging input from experts on possible barriers and trade-offs of elements of study design.

For patients with RA-ILD, there is only 1 dedicated prospective, randomized controlled trial of treatments.¹³ This study, TRAIL1, which evaluated the safety and efficacy of pirfenidone, failed to reach its primary end point; however, secondary findings suggested the medication could slow FVC decline. The primary treatment approach of immunomodulation in RA-ILD is supported by observational data,¹⁴ but has not been studied in a randomized trial, limiting meaningful conclusions about safety and efficacy. Patients with RA-ILD deserve to have treatments proven to be safe and effective through high-quality clinical trials. Without such trials, all guideline recommendations for treatment in RA-ILD will remain conditional given the limited evidence. One possible avenue to obtain high-quality necessary data is through pragmatic comparative effectiveness studies of the 2 primary strategies most commonly in use.

The patient perspective in selecting pragmatic trial outcomes is of paramount importance. In the patient small groups, the discussion about outcome choice in RA-ILD elicited themes that considered the importance of articular disease control and a desire to include outcomes beyond pulmonary physiology changes. This perspective was corroborated by the expert panel, which noted that treatment decisions are based on balancing the patient’s lung disease phenotype, articular disease, and side-effect profiles of the available agents. Pragmatic study design allows for, and encourages, the flexibility to track these outcomes.

These results highlight the equipoise for this question and broadly supports the feasibility of leveraging routine, real-world clinical outcomes to measure PFT changes and radiographic progression. However, the patient and expert panel perspective support the use of patient reported outcomes and functional assessments as primary outcomes in comparative effectiveness study design, a finding similar to other published reports supporting the use of patient-centered outcomes in clinical trials.^15,16 It is imperative for a pragmatic comparative effectiveness study in RA-ILD to balance pragmatism of clinical outcomes with validated and exploratory measures of patient outcomes and quality of life (eg, validated quality of life measures related to dyspnea in patients with RA-ILD).¹⁷ Digital platforms (eg, study applications, telephone calls from study coordinators, emails, traditional coordinator-based outcome collections) can be combined to create multiple layers of patient-reported outcome collection in pragmatic study designs and should form the basis of a robust patient-centered outcome approach in these trials.

Regarding general barriers to research, patients expressed their desire for expanded access to clinical trials. This theme was similarly identified in another qualitative study of ILD care delivery models.¹⁸ Pragmatic trials have a primary advantage over traditional clinical trial infrastructure of less stringent inclusion and exclusion criteria to increase the external validity of the trial. By nature, this study design can increase the number of eligible participants in clinical trials via this design feature. Furthermore, pragmatic study designs offer researchers the opportunity to innovate by expanding decentralized approaches to patient enrollment.¹⁹ Decentralized trials democratize enrollment beyond siloed academic medical centers with research infrastructure support. By incorporating decentralized structures into pragmatic research trials in RA-ILD, several important questions could be pursued by leveraging a more diverse and representative sample of patients for significantly less financial burden.

Interpretation

In conclusion, we described and analyzed clinician and patient perspectives on pragmatic trial design in RA-ILD. To our knowledge, this is the first study aimed at soliciting clinician and patient perspectives on trial design in RA-ILD. With the availability of immunomodulatory and anti-fibrotic therapies for RA-ILD, the themes and concepts of trial design identified in this study can be used to design patient-centered, pragmatic RA-ILD trials.

Take-Home Points.

Study Question:

What are the perceptions of patients and experts on pragmatic trials for rheumatoid arthritis-associated interstitial lung disease?

Results:

Experts highlighted the current variability in treatment strategies, lack of relevant high-quality data, and necessary focus on patient-reported outcomes. Patients and caregivers centered on quality of life, barriers to participation, and desire for patient-centered research.

Interpretation:

Pragmatic trials could represent an opportunity to find evidence-based, patient-centered approaches to treating rheumatoid arthritis-associated interstitial lung disease.

ABBREVIATIONS:

ILD

interstitial lung disease

PFT

pulmonary function test

RA-ILD

rheumatoid arthritis-associated interstitial lung disease