Table 1shows the clinic, patient, and Rheumatologist’ characteristics, as well as trial-related questions. Among 26clinics, 58% were general rheumatology (42% had separate RA clinics), 69% were academic-center clinics. The median number of rheumatologists at each potential trial site was 10 with a support staff of 6. The median number of RA patients seen in one clinic per year was 1,000 with a median of 10 RA patients per day. The median rheumatologist age was 45 years. Half of the rheumatologists had 6-10 years of clinical practice experience since completing their rheumatology fellowship.
All participants indicated interestin participating in this pragmatic trial.Almost all (96.2%) clinics performed at least one lab monitoring for RA patients. Most participants (73%) thought the plan for the 8-month trial with 28 joint counts was feasible, and 23 participants (88.5%) indicated that there were no other critical barriers than discussed below (Table 1).
In terms of patient reported outcomes measures, three participants (11%) indicated that they were satisfied with standard RA measures without additional suggestions. Other suggestions for RA measures are listed in Table 2.
Twomain themes were coded, consistent with the research questions identified and includedpatient barriers, and clinic barriers(Figure 1).
Theme 1: Patient Barriers
Physicianperceived patient barriers offer challenges to a RA pragmatic trial. This theme coveredthree sub-themes: 1) patients’ personal barriers, 2) patients’ treatment-related factors, and 3) trial-related factors (e.g., patient recruitment, side effects, mode of use, etc.).
Patients’ personal factors.
Patients’ personal factors, the largest sub-theme, included insurance status, language barriers, and travel-related factors. Insurance status and plan was the biggest factorthought toinfluence patients’ medication affordability and thus, trialparticipation. Seventeen out of 26 (65%) physicians mentioned health insurance as one of the top three patient barriers. Different insurance plans (i.e., private commercial plan, Medicare, Medicaid) have different rules forRA medication- switch, when the MTX did not work. This may limit physicians to prescribe only the insurance -preferred new medication. Dr. D stated “let’s say they start on one TNF (tumor necrosis factor) and they want to crossover, but their insurance only allows them another TNF, so, they usually make us do two TNFs before being able to move on. … we have Medicare which is probably about 30-35% of our population.” Insurance plans have restrictions and requirements on crossover medications and havedifferent co-pays. Medicare patients accounted for about 30% to 35% patient population in someclinics, but the medication coverage rules also differed by the region. Dr. Y said, “it can vary according to Medicare secondary. If it is straight Medicare, they are going toan 80/20 split, with that 80/20 split, the cost of the TNF is much less then say Actemra, where it is a much higher, so 20% of you know, I am just going to throw out, it’s not $40,000 but I mean 20% of a $2000 infusion versus 20% of a $1000 is more out of pocket for them.” Medicaid has restrictions too.. Dr. Y explained, “Medicaid doesn’t let us have; some insurance companies may let you have it.”
Patients with limited English proficiency have communication barriers, which limits trial participation. Clinics across the country serve different population, such as Hispanic and Asian population in California, and Hispanic population in Texas. Dr. L said, “We do have a large number of patients that are Spanish speaking. We do have bilingual coordinators, also bilingual physicians but, you know, they may or may not be attracted to (the trial), but we would obviously need to get consent form and other materials in Spanish.”
Physicians were also concerned about the travel time forpatients. A longer distance from home to clinic might limit patients fromvisiting clinic regularly. Dr. X stated, “Distance from home to visit us, sometimes they don’t want to come as often as we would like.”
Treatment-related factors.
The second sub-theme was treatment-related factors, including patients’ medical condition and complexity, comorbidities, treatment, and patient outcomes. Patients differ by types and numbers of current medications, and their behavioral and physical health conditions. These factors providedadditional challenges for physicians when they identify patients for pharmaceutical trials. Dr. O. explained, “Sometimes patients are started on dual therapy, MTX (methotrexate) +HCQ (hydroxychloroquine), for a physician not to push dose higher, triple therapy in 25% of my patients.” Patients on single therapy are optimal for the pragmatic trial. This makes physicians easy to examine the effects of the medication and adjust another medication if it failed. Dr. K said, “If they have another serious infection with one type of drug, then switch to another drug.” Some physicians were concerned that infectionslead to discontinuation. Dr. M asked, “If they have infection, can they cross-over to the other arm sooner? They can also drop out, if the other arm has a similar contraindication.” Other physicianshad different opinions about the comorbidities and thought seriousinfections are uncommon, “Comorbidities shouldn’t be an issue (Dr. L).” One physician was concerned about the influence of some biologics on patients’ outcome. Dr. B asked, “If DAS28 (Disease Activity Score in 28 joints) is the outcome measure, how do you measure the outcomes with IL-6 drugs (tocilizumab) that are very effective against CRP and compare outcomes across groups?”
Trial-related factors.
The third large sub-theme was trial-related factors, and included patient recruitment, preferences, burden, and side effects. Almost all physicians reported patient recruitment asone of the top 3 patient barriers, including few eligible rheumatoid arthritis (RA) patients (e.g., new RA patients), unwillingnessin trial participation, patients from different cities, patient identification, other competing study in the clinic limiting patient availability. Dr. R explained, “A lot of time research visits are on non-clinic days, interestingly, even for studies that I am doing it’s hard to stop and identify patients in the middle of a busy clinic.”
Patient preferences also play an important role in recruitment, such as unwillingness to be randomized, research participation, orstrong patient preference for oral therapy versus injection or vice versa. Dr. G said, “Patient preference- they have a strong choice, based on what they want, either the drug or the route.” There are different patients, some patients are willing to join clinical trial. Dr. W explained, “for a lot of people placebo-based trials, it is going to be more, you know people who may not have insurance that are willing to get on a trial. Since it is a pragmatic trial, they are eligible for any of the biologics is my assumption.”
Physicians also reported concern about patient attrition from the study.Patient burden of completingthe questionnaire and extra time added to their usual clinic visit time was a perceived barrier. However, some physicians thought committing extra time during usual clinic visits will increasepatient participation. Dr. A stated, “Well, it depends on patient not wanting to commit the extra time, they may not see what’s in for them.”
Medication side effects were also a concern, including gastrointestinal (GI) side effects, upper respiratory infection (URI), and injection site reaction. However, there were mixed opinions on side effects. Some physicians were worried about the side effects, while, other physicians did not expect too much side effects from the medications in RA pragmatic trial. One physician suggested a solution to cope from side effects. Dr. P explained, “GI side effects, viral URIs are the most common issues with these drugs, hold it for one dose. If Jak kinase, hold for 5 days.”
In summary, physicians’ perceived patient barriers reflected concerns about insurerrequirement, patient preferences and recruitment, trial-related factors, treatment and comorbidities, and available resources to work with patients.
Theme 2: Clinic Barriers
Physician perceived clinic barriersfrom the clinic or the healthcare system perspective to RA pragmatic trial enrollment. This theme covered four sub-themes: 1) clinic-related factors, 2) patient-related factors, 3)research personnel, and 4) facilitators (positive factors of the clinic) (Table 2).
Clinic-related factors
The clinic-related factors include patient/clinic flow, lack of resources (e.g., patient room, infrastructure, and time), ethics committee approval, competing trial/studies, clinical equipoise, and recent changes to a new health care record system (e.g., EPIC).
Abig concern among clinic-related factorswas the patient/clinic flow. “Hard to integrate clinical research into the routine clinic flow, not to say it can’t be done, we see patients pretty quickly, in the real world, 20-minute follow up visit, it will take a little bit more for the patient” (Dr. R).Lack of resources was a close second, including inadequate resources for adequate coordinator time, room and time to do the study, technology, and logistics. Dr. S stated, “Logistics, how it is set up in the clinic. If you had a small shop, it might be a problem for those sites.”Within the lack of resources sub-theme, the number of rooms for RA patients remained a concern for four participants although the median number of patient rooms in clinic was 10 (Table 1).
Institution review board (IRB) application and other ongoing studies in the clinic were considered as other clinic barriers. Five participants reported delay in IRB application approval as a barrier for their trial participation.Dr. X said, “We may have competition from another study. Easy to deal with our own IRB. Central IRB was difficult.”Participants thought that other trials that they were conducting might compete with the pragmatictrial for patient recruitment. Dr. Y explained, “I have another study that is a bit of a conflict because everybody wants patients with methotrexate exposed and not anything else.”
Clinical equipoise was also considered a clinic barrier. Participants were concerned about the clinical equipoise for patients who will participate in the trial. Dr. P said, “Whole issue of equipoise is we don’t’ realize that we do not know. Worry that if they are being randomized, they may be harmed by not getting the ‘preferred treatment’.”Two participants reported that recent changes to their electronic health record system (EHR), EPIC, raised some challenges for them to recruit patients, however, other two participants thought the changes would help patient recruitment in this trial.For example, Dr. Y explained, “We had used RAPID-3 (Routine Assessment of Patient Index Data with 3 measures), we have now switched to EPIC and we only got the upgrade to use CDAI (Clinical Disease Activity Index).” However, Dr. T held the opposite opinion and said, “Getting EPIC implemented this weekend, that should be a big plus.”
Patient-related factors
Among all patient-related factors, insurance was considered as the key clinic barrierby12(46%) participants, including 7 participants who thought that insurance was a barrier for both clinic and patients. As mentioned early in the section of patient barriers, different insurance plans have different requirement or restrictions on the TNF vs. Non-TNF biologic vs. Jak-kinase inhibitor initiation for RA patients who experienced MTX failure. The variation in patient insurance plansbygeographyposes different levels of challenges to the clinics. Dr. J, one of the seven participants who considered insurance as both clinic- and patient-level barrier,explained “Underinsured, probably, I think, 35-40% of our patients. Medicare would probably be, overall, 60%. Commercial 10-15%. That is going to be a problem.”
Another patient-related factor was clinic-level recruitment, which was mentioned by 8 (31%) participants, including 4 participants who treated patient recruitment as a both clinic- and patient-level barrier.Participants were concerned about patient recruitment because they saw few early or new RA patients, for example, Dr. N said, “Typically, these are early RA patients, very few, Not common.” Participants thought that patient randomization might be problematic. Dr. P explained, “Can keep track of who is enrolled in the study but cannot be randomized.”Multiple sites of the clinic might be a barrier to implement the trial. Dr. W said, “We have obviously patients coming from many different sites, so I would have to think about where we would maximize our personnel to be at the site that sees the most of these patients, right?”
Research personnel
Research personnel was the third sub-theme underclinic barrier theme, including coordinator efforts and provider engagement. Nearly half of participants (46%) reported that coordinator effort was a concern, including decision on full time vs. part time coordinator, budget fora new coordinator, and coordinators were stretched between several studies. Dr. M explained, “would need coordinator to help with all the paperwork and keeping on track with you guys, that would be my concern.” Dr. Y said,”…the problem for us is most of the bureaucracy of clinical trials because we do not have a dedicated clinical trial coordinator.” Another concern wasrheumatologist engagement mentioned by 10 participants (38.5%) because they thought that every physician has a busy schedule. Dr. S explained, “I just think everyone is overwhelmed in the clinic, you know, gotten harder and harder to make a living in rheumatology and so, people are squeezing more patients in, so the rate limit taking the time to begin that discussion, we have a clinical research center, your budget is not going to be what we are used to.”
Facilitators (Positive factors)
Although the goal of the interviews were to identify factors that contributed to the patient recruitment in the pragmatic trial, many participants pointed out some positive factors that may facilitate patient recruitment. These positive factors include 1) havingan experienced research team, 2) having developed a good network, 3) having technology support, and 4) the trial design rigor. Seven participants were confident that they had an experienced research team, including pharmacists and nurses. Dr. F explained, “We have done 100 clinical trials, we have a full support of our nursing team.”Five participants reported that they have developed a good network of research that can help in trial implementation.Dr. O stated, “in the past few years, no RA studies, patient pool is larger and less fatigued than before, mostly PsA (psoriatic arthritis) studies. They trust our coordinators and our research teams.”
In summary, physicians perceived clinic barriers reflect concerns in insurance (e.g., proportion of patients with Medicare and Medicaid), influence on clinic routine, coordinator efforts, and provider engagement, and patient recruitment.In addition to barriers, participants also reported some facilitators for trial implementation.
Insurmountable Barriers
After discussing the top 3 patient barriers and top 3 clinic/system barriers, participants were asked which barriers were insurmountable. Eleven(42%) participants said there were no insurmountable barriers at their clinic. Fifteen (58%) participants pointed out that28 barriers were insurmountable, including clinic-related barriers (i.e., clinic flow, change to EPIC, IRB), patient-related barriers (i.e., insurance, patient recruitment, patient stipend, provider engagement, patient language barriers, and coordinator effort). Among all insurmountable barriers, the leading barrier was patient insurance (7 out of 15 participants, 47%), followed by patient interest/recruitment (6 of 15, 40%), patient stipend (3 of 15, 20%), provider engagement (3 of 15, 20%), patient language barriers (2 of 15, 13%), coordinator effort (2 of 15, 13%), clinic flow (2 of 15, 13%), change to EPIC (7%), and the IRB (7%).