UBC Faculty Research and Publications

Perspectives of patients, first-degree relatives and rheumatologists on preventive treatments for rheumatoid… Munro, Sarah; Spooner, Luke; Milbers, Katherine; Hudson, Marie; Koehn, Cheryl; Harrison, Mark Jul 5, 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-41927_2018_Article_26.pdf [ 441.7kB ]
JSON: 52383-1.0368838.json
JSON-LD: 52383-1.0368838-ld.json
RDF/XML (Pretty): 52383-1.0368838-rdf.xml
RDF/JSON: 52383-1.0368838-rdf.json
Turtle: 52383-1.0368838-turtle.txt
N-Triples: 52383-1.0368838-rdf-ntriples.txt
Original Record: 52383-1.0368838-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessPerspectives of patients, first-degreerelatives and rheumatologists onpreventive treatments for rheumatoidarthritis: a qualitative analysisSarah Munro1,3, Luke Spooner2, Katherine Milbers2,3, Marie Hudson4, Cheryl Koehn5 and Mark Harrison2,3,6*AbstractBackground: There is growing evidence that it may be possible to identify people at high risk of developingrheumatoid arthritis (RA). Assuming that effective interventions were available, this could mean that treatmentsintroduced in the pre-symptomatic phase could prevent or delay the onset of the disease. Our study aimed toidentify the potential attributes involved in decision-making around whether or not to take preventive treatmentfor RA, in order to inform the development of a discrete choice experiment (DCE) to ascertain consumerpreferences for a preventive treatment program for RA.Methods: We conducted a focus group study to develop conceptual attributes, refine their meaning, and developlevels. Participants included RA patients, first-degree relatives of RA patients, and rheumatologists who were18 years of age and over, could read and speak English, and could provide informed consent. Candidate attributeswere refined through iterative rounds of data collection and analysis. All focus groups were audio-recorded andtranscribed, and then analyzed using the Framework Method to identify, compare, and contrast key conceptualattributes.Results: Attributes identified from analysis included: accuracy of the test, certainty in estimates, method ofadministration, risk of RA and risk of reduction with treatment, risk and seriousness of side effects, person recommendingthe test, and opinion of the health care professional. Patients with RA, first-degree relatives of patients, andrheumatologists all valued the accuracy of testing due to concerns about false positives, and valued certainty inestimates of the test and preventive treatment. Patients and first-degree relatives desired this evidence from arange of sources, including discussions with people with the disease and health care professionals, and theirpreferences were modified by the strength of recommendation from their health care professional.Conclusions: The role of the person who recommends a test and the opinion of a health care professional are novelpotential attributes involved in decisions around whether or not to take preventive treatment for RA, that have notbeen included in previous DCEs.Keywords: Arthritis, Decision making, Discrete choice experiment, Focus groups, Health/economics, Rheumatology/economics, Rheumatoid/therapy* Correspondence: mark.harrison@ubc.ca2Faculty of Pharmaceutical Sciences, University of British Columbia,Vancouver, BC, Canada3Centre for Health Evaluation and Outcome Sciences, St. Paul’s Hospital,Vancouver, BC, CanadaFull list of author information is available at the end of the articleBMC Rheumatology© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Munro et al. BMC Rheumatology  (2018) 2:18 https://doi.org/10.1186/s41927-018-0026-7BackgroundThere is growing evidence that it may be possible to iden-tify people at high risk of developing rheumatoid arthritis(RA) based on clinical and genetic risk factors, such asfamily history of RA [1], female sex [2–4], and lifestyle riskfactors including smoking [5, 6]. Multiple studies haveshown that RA has a prolonged and identifiable asymp-tomatic pre-clinical development phase, during whichcharacteristic biomarkers appear, in particular antibodiesagainst cyclic citrullinated peptide (CCP) [7–10]. Theseautoantibodies have been shown to precede the onset ofRA symptoms by many years and are highly specific toRA [7, 11, 12]. The increasing evidence of a pre-clinicalphase of RA has led to interest in the early identificationof people at high risk of RA, and a possible window of op-portunity for treatment in the pre-symptomatic phase toprevent or delay onset of the disease [13].There is considerable uncertainty about when andwhether those who are predicted to develop RA will ac-tually develop RA [14], and if those individuals would bereceptive to preventive treatment [15]. In this context, itis critical to explore the preferences of stakeholders whowould be directly affected by preventive RA treatment inorder to predict uptake of treatment options, and to in-form policies for resource allocation [16, 17]. To the bestof our knowledge, there are currently two qualitativestudies involving first-degree relatives from the UK,Germany, and Austria [18, 19] and one exploratory bin-ary choice experiment, which was not developed usingqualitative methods [15], that explores or describeswhich attributes of a potential preventive treatment pro-gram for RA are valued by those who might be asked torecommend, consider, and provide or accept preventivetreatment [17].Discrete choice experiments (DCEs) are a rigorousmethod for exploring consumer preferences in healthcare services [18, 19]. In a DCE survey, the participantreviews a set of hypothetical scenarios and makes theirpreferred choice among two or more health care op-tions. This technique allows the researcher to identifyparticipants’ willingness to make trade-offs between thedifferent characteristics, or “attributes,” of a health careservice [19–24]. Undertaking qualitative work in partici-pant populations before designing and administering aDCE is critical to identify important attributes and de-scribe them in a way that is relevant and understandable,and to reduce the risk of inaccurate or biased DCE re-sults [22].Our study aimed to identify the potential attributes in-volved in decisions around whether or not to take pre-ventive treatment for RA, to inform the development ofa DCE that would subsequently be used to ascertain thepreferences of people at risk of developing RA for devel-opment of a preventive treatment program for RA.MethodsStudy designFollowing principles for qualitative attribute develop-ment in the design of DCEs [22], we conducted a focusgroup study to develop conceptual attributes, refine theirmeaning, and develop attribute levels.Setting and participantsFocus groups were conducted January to March 2016 intwo large Canadian cities. Participants included RA pa-tients, first-degree relatives of RA patients, and rheuma-tologists who were 18 years of age and over, could readand speak English, and could provide informed consent.Patient and first-degree relative participants were re-cruited through the marketing and communications listsof the Arthritis Consumer Experts/Joint Health groupand the Arthritis Research Canada Arthritis Patient Ad-visory Board mailing list, whose subscribers are a generalaudience of arthritis patients. All first degree relativeswere recruited via the mailing list or snowball samplingthrough the patient participants. Rheumatologists wererecruited through personal invite from the rheumatolo-gist on our team (MHu) and selected to represent arange of experience and geographic setting. Participantswere screened for eligibility and provided with a consentand study information form in advance of the focusgroup. Ethical approval for the study was granted by theUBC Behavioural Research Ethics Board (H15–01948).Data collectionIterative rounds of data collection and analysis wereconducted to create the list of candidate attributes:1) Round 1: One focus group each with RA patients,first-degree relatives, and rheumatologists toidentify and refine the preliminary list of attributes,discuss clarity of description of RA, and considerpotential treatment decisions around preventivetreatment to be included in the DCE.2) Analysis of focus group data and revision of the listof attributes based on participants’ feedback fromthe first round;3) Round 2: One focus group each with patients andfirst-degree relatives (different participants for bothgroups to those in round 1) to provide feedback onthe list of attributes and survey design. At the endof these focus groups participants were asked torank the list of potentially important attributes aspriorities.Before commencing each focus group, the moderatorreviewed the study information and consent form, andreceived written consent from each participant. Asemi-structured interview guide was developed for theMunro et al. BMC Rheumatology  (2018) 2:18 Page 2 of 10study consisting of questions and probes that exploredpotential candidate attributes, which were identifiedthrough consultation with a patient partner, rheumatolo-gist and pharmacist experts, researchers, and a literaturereview. The review included existing choice-based stud-ies of treatments for RA, either from the published lit-erature or published as abstracts from the 3 previousyears of American College of Rheumatology, EuropeanLeague Against Rheumatism, and British Society forRheumatology meetings. Conceptual attributes – topicsthat may potentially, but not necessarily, be included ina DCE – that were identified prior to the first focusgroup included effectiveness at preventing RA [15], gen-eralized and local side effects/adverse events [15, 23, 24],strength of evidence [24], how the treatments are taken[15, 23, 24], how often they are taken [15, 23, 24], howlong they would be taken for [15], and cost [23]. Thesetopics were explored through focus group discussion sothat we could identify candidate attributes – conceptsthat are important, relevant, and understandable topeople involved in decisions around whether or not totake preventive treatment for RA. Expert consultationconsisted of discussion with clinical members of the re-search team in Vancouver regarding the types of infor-mation they would expect to communicate to a personmaking the decision.Round 1 focus groups with first-degree relatives ex-plored the potential impact of RA on their life, personalexperiences of RA with relatives, their informationneeds, with whom they would discuss preventive treat-ment, and the importance of uncertainty in risk predic-tion, benefits, and incidence of side effects. In the secondround of focus groups, first-degree relatives and patientgroups reviewed and discussed a list of candidate attri-butes and asked to rank the importance of attributes withregard to the decision of whether to undertake preventivetreatment and to eliminate attributes they felt to be unim-portant. They also discussed the decision-making themesthat arose from analysis of Round 1 focus groups and pro-vided further feedback on the survey design. All focusgroups were audio-recorded and transcribed for analysis.Finally, expert consultation consisted of discussion withclinical members of the research team in Vancouver re-garding the types of information they would expect tocommunicate to a person making the decision.Data analysisQualitative analysis was guided by the FrameworkMethod [25] to identify the key concepts emerging fromeach of the interview transcripts and to compare andcontrast these across focus groups of patients,first-degree relatives, and rheumatologists. Analysis con-sisted of a) immersion in the transcripts, b) coding, c)development of a working analytical framework, d)applying the framework, e) charting the data into amatrix, and f) interpreting the data. Two researchers(LS, KM) read and re-read the transcripts to gain famil-iarity and then open-coded them independentlyline-by-line to identify salient themes. The two re-searchers then met to discuss their codes with a qualita-tive methods expert (SM), assess interrater reliability,make changes to labels for semantic consistency, andthen grouped similar codes together to create a workinganalytical framework. This process of discussing thecodes, testing them with the transcript data, and refiningthe framework was repeated until no new codes weregenerated. The coded transcripts were then entered intoNVivo qualitative analysis software (version 10).To complete the framework analysis, after data hadbeen coded and organized, one researcher (LS) chartedthe data in a matrix table and identified illustrativequotes with a and b to indicate how well they repre-sented the sub-theme. For each key theme, the matrixconsisted of one row per participant group and one col-umn per sub-theme. Potential attributes were generatedfrom the matrices by interpreting them as a team (LS,KM, SM) and identifying patterns between participantgroups and themes.The final stage of analysis involved development of po-tential attributes and attribute-levels that would be im-portant in considering whether to take or recommendpreventive treatment. Through face-to-face discussion,the researchers (LS, KM, SM) engaged in data triangula-tion following constant comparison techniques [26]. Wecompared the framework analysis matrix with the poten-tial conceptual attributes that had been identifiedthrough expert consultation and literature review. Thistriangulation sought to enhance the representativenessand legitimacy of attributes and levels developed.Throughout data collection and analysis, we engaged instrategies for quality and validity, including keeping anaudit trail of memos, independent coding to reduce bias,and constant comparison of data with data.ResultsSummary of qualitative analysis and key themesIn total, 5 focus groups were conducted with 25 partici-pants (13 patients [7 in first round, 6 in second round],5 first degree relatives [3 in first round, 2 in secondround], and 7 rheumatologists [second round]). Themean age among patients and first degree relatives was50.4 (SD 18.3) and 29.4 (SD 12.4), respectively. Patientshad experienced their disease on average 14.5 years (SD9.1) and most (82%) were currently taking treatment forRA. The rheumatologists currently practiced in Alberta,British Columbia, Ontario, and Quebec. Analysis identi-fied two key themes: “Living with Rheumatoid Arthritis”and “Preventing Rheumatoid Arthritis.” A summary ofMunro et al. BMC Rheumatology  (2018) 2:18 Page 3 of 10major themes and related subthemes is found in Table 1and discussed below.Living with rheumatoid arthritisFor patients, living with rheumatoid arthritis involvedcoping with symptoms of joint swelling, without whichone “might think [they] don’t have the disease.” Thesymptoms of the disease had an emotional impact oneveryday life; for example, one patient felt “devastated”when they “couldn’t even for an afternoon get dressed,couldn’t pull up the zipper on [their] pants.” Positiveemotional impacts included patients being proactiveabout their health and making lifestyle changes such asexercising or quitting smoking. In the second round offocus groups, patients further expressed that efforts toimprove their quality of life also involved trying to avoidthe side effects of medication, and some had strong per-ceptions that RA treatment had damaged their kidneys,liver, and heart. Patients had few concerns about the im-pact of the test, that is the potential impact of a positiveresult, for risk of RA and felt that any informationgleaned from test results would enhance their healthcare experience. The results may also signal a wake-upcall to change their relative’s health behavior and miti-gate the future effects of living with RA, or seek furtherexamination and testing before undertaking preventivetreatment.First-degree relatives’ perceptions of RA were verysimilar to that of patients and were shaped by their ex-perience witnessing the effects of RA on their relative.One first-degree relative described RA as “not just a dis-ease of the joints. It just starts in the joints. It can affectevery organ in your body.” Some discussed wanting abetter quality of life for their relative and described thelifestyle changes that happen when your family memberlives with RA, such as no longer being able to travel. Aminority of patients noted, however, that early informa-tion on their risk for RA would have negatively impactedtheir ability to enjoy their lifestyle while still asymptom-atic. First-degree relatives indicated they would be inter-ested in testing for RA, if it gave them concrete ways tobe proactive about their health such as changing theirlifestyle or taking preventive treatment to maintain theirhealth. However, they had concerns about the impact ofthe test, particularly the potential damage of a false posi-tive result on their mental wellbeing and their insurancecoverage.The rheumatologist focus group discussion suggestedthat one theme of primary importance to clinicians wasconcerns about the impact of the test for high-risk indi-viduals. The lifestyle concerns raised by patients andfirst-degree relatives notably did not emerge amongstrheumatologists. However, they did share the practicalconcerns about the costs related to the test, as well asinsurance and anxiety concerns stemming from a falsepositive result. Additionally, rheumatologists’ discussionshighlighted the perceived challenge of supporting RApatients to be proactive about their health, particularlyregarding medication adherence. One rheumatologistsuggested that first degree relatives may be no more ad-herent to a preventive treatment regimen than theirfamily members living with RA, questioning the utilityof preventive testing and treatment. Rheumatologistsagreed that maintaining current quality of life was animportant factor in the consideration of treatment,though their preferences primarily involved avoiding sideeffects in the long term (Table 2).Preventing rheumatoid arthritisAll participant groups questioned if preventive treatmentis appropriate in first-degree relatives, due to the poten-tial for serious side effects. Patients suggested to “startTable 1 Summary of Major Themes and Related SubthemesTheme Sub-ThemesLiving with Rheumatoid Arthritis Living with RABeing Proactive about My HealthWanting a Better Quality of LifeTrying to Avoid the Side Effects of Medications On Health (not deteriorating for patients)On Lifestyle (not getting worse for bothpatients and first-degree relatives)Having Concerns about the Impact of the TestPreventing Rheumatoid Arthritis Questioning if Preventive Treatment Is AppropriateNeeding More Evidence Due to uncertainty about the treatmentDue to gaps in knowledge about RAImplementing Preventive Treatment for RA In clinical practice with patientsAt the health system levelWanting Alternatives to Medication: For Preventive TreatmentMunro et al. BMC Rheumatology  (2018) 2:18 Page 4 of 10small” and “if the RA factor is rising, then look at some-thing” rather than have a first-degree relative immedi-ately take medication. Focus groups with first-degreerelatives highlighted that if a trusted health care profes-sional recommended the preventive treatment theywould believe it was appropriate. All sought more evi-dence on the evidence for potential preventive treat-ments, including related to the dosing schedule,methods of administration, how it has been tested (i.e.clinical trials), short and long-term side effects of treat-ment, and the evidence on the effectiveness of the treat-ment in preventing RA. Patients were particularlyhesitant about preventive treatment because they per-ceived there to be significant gaps in current knowledgeand ability to diagnose and understand the causes of RA.Finally, when discussing how to implement preventivetreatment in practice, analysis of all focus groups sug-gested cost was important and that counseling shouldemphasize the purpose and effectiveness of the treat-ment, to enhance adherence. At the time of diagnosismany patients were initially in denial of RA or believedthat diet and lifestyle changes were needed, rather thanmedication.First-degree relatives similarly focused on wanting al-ternatives to medication, such as natural “herbal” treat-ments. First-degree relatives also needed more evidencedue to uncertainty about the treatment. This includedtopics such as the extent of which RA was hereditary,and the differences between RA and other types of arth-ritis, such as osteoarthritis. They perceived that the ef-fectiveness and side effects of treatment for preventingRA might be similar to the treatment for their relativesliving with RA.The rheumatologist discussion focused on concernsabout the appropriateness of preventive treatment dueto the side effects of medications, and the need for moreevidence on the probability that treatment would preventRA. Analysis suggested rheumatologists have concernsabout the marginal benefit of treatment compared to notreatment. Their discussions suggested that, in the ab-sence of high quality data on the effectiveness of emer-ging pharmaceutical treatments, weight loss or smokingcessation would be reasonable treatments to recom-mend. When discussing how to implement a preventivetreatment program in practice, the concerns of rheuma-tologists centered on the ability to identify people athigh risk of RA, whether rheumatologists had capacityto take on more patients, and whether rheumatologistswere well suited to provide preventive treatment inter-ventions. Nevertheless, some rheumatologists were will-ing to consider providing preventive treatments incertain high-risk populations (i.e. first-degree relativesTable 2 Thematic Framework for “Living with Rheumatoid Arthritis”Living with RheumatoidArthritisBeing Proactiveabout My HealthWanting a BetterQuality of LifeTrying to Avoid theSide Effects ofMedicationHaving Concerns aboutthe Impact of the TestPatient “I was just devastated, Icouldn’t even, for anafternoon get dressed,couldn’t pull up thezipper on my pants, andI just lost so muchweight.”b“You want to dosomething about it [RA].It’s not just a matter ofkind of thinking you’regoing to get over it”a“[Treatment would]minimize sort of like thelong term impact ofmaybe like hunchingand succumbing to thepain. So that was reallygood.”b“And of course themedication is affectingall the other things, theliver, the kidneys. Yourskin, your hair, likeeverything. Eyes.”b“And you could havesome kind of decisiontree or flow chart. Okay,so I’ve got this positivemarker, now what?”bFirst DegreeRelative“She [family member]had a life and then oncethe disease came andtook it from her, shedidn’t [anything]anymore. She couldn’tdo things.”b“If there were perhaps atreatment that wereextremely preventiveand very effective atlessening the risk ofdeveloping such adisease, I absolutelywould take the testbecause that to me leadsto something that ispreventive. That leavesme being able to takesome action”a“If that was a risk for themedication, it’s also arisk for the RA. You’realmost guaranteed toget serious infectionsand TB is completelylikely. So would I ratherget those now when I’mstrong enough andhealthy enough tofight them”“Especially because ofwatching my mom withprednisone, if there’sanything that increasethe mental risk, thatwould be like hugefor me.”a“And for me adding anykind of anxiety to it, notbecause [a test result]necessarily jars me into arealism that I’m notcomfortable with, butbecause I don’t think itadds anything.”bRheumatologist “They [first degreerelatives] want to know[about RA risk] becausethey think that they canprevent disease inthemselves.”“Well, if I know I’m goingto have Lupus then myinsurance goes into thetoilet, you know, and Idon’t want that, so Idon’t want to know. Idon’t want my family toknow.”ba and b next to quotes indicate moderate and high importance/representativeness, respectivelyMunro et al. BMC Rheumatology  (2018) 2:18 Page 5 of 10with RA, smokers, Aboriginal [First Nations, Inuit, andMétis] persons) (Table 3).Discrete choice experiment attributesResults of iterative focus group data collection and ana-lysis suggested seven candidate attributes were involvedin the decision to consider or recommend preventivetreatment for RA: 1) accuracy of the test, 2) certainty inestimates of the risks and benefits of treatments andtests, 3) method of administration, 4) the initial baselinerisk of RA and risk reduction with treatment, 5) risk andseriousness of side effects, 6) who recommends the per-son to consider treatment (e.g., a friend, or a health careprofessional), and 7) opinion of the health care profes-sional about a preventive treatment option. Table 4 pro-vides a summary of the final attributes and levelsgenerated and modified through analysis of focus groupdata, as well as direct quotes from patients, first-degreerelatives, and rheumatologists to support the selection ofthese attributes.Patients and first-degree relatives valued the “Accuracyof the Test,” which they perceived to mean having a lowrate of false positives. Rheumatologists were also con-cerned with the accuracy of the individual componentsof the test in predicting RA (i.e. genetics, presence ofRA specific antibodies). Similarly, “Certainty of Risk Es-timates” all groups agreed that the strength of evidenceto support testing and preventive treatment was import-ant. Rheumatologists particularly valued evidence fromplacebo-controlled trials in high-risk populations, andpreventive treatments with the greatest risk reduction.These attributes were closely related to the themes ofPreventing Rheumatoid Arthritis discussed above.In considering how to operationalize the sub-themeWanting a Better Quality of Life into an attribute for theDCE, our data suggested that potential candidates in-cluded risk and type of side effects, ability to maintainTable 3 Thematic Framework for “Preventing Rheumatoid Arthritis”Questioning if PreventiveTreatment Is AppropriateNeeding More Evidence:Due to uncertaintyabout the treatmentNeeding More Evidence:Due to gaps inknowledge about RAImplementingPreventive Treatmentfor RA: In clinicalpractice with patientsWanting Alternativesto Medication: Forpreventive treatmentPatients “Because it [RAtreatment] is going tostop your pain when youtake it anyways, whywould you want to takethat before if it has a lotof risk involved?”“How the treatmentaffects or it works, downto a cellular level. Themethods and results oftesting. All possible sideeffects, short-term,long-term, andcomplementary lifestylechoices.”a“If we don’t know thecause [of RA], everythingis suspect that we do.You know? Andespecially all thetreatments”a“People should knowwhy they should takethe drugs because, forpeople like me whowere in denial or justthought I would eatbetter and exercise anddo yoga and whatnotI’d be fine and I don’tneed all these drugs.”“Your whole generationjust looks at so manydifferent options.”aFirst DegreeRelatives“From where it would becoming from, Dr.— waslike, ‘Hey, you know,there’s this treatment.You know, I know howbadly it effects yourmother. I think that youare possibly at risk forhaving it,’ and hesuggested it to me, Iwould definitely take alook at it.”b“There would always bethat little bit in the backof my mind that wouldgo, ‘Okay, how far is thetreatment going to beadvanced by the timethat I get there.’ Youknow, like in another 15,20 years of medicalscience how much is thetreatment for peoplewith it going to beadvanced?”b“And I’ve heard theories,everything from it [RA]skips generations to it’simmediate, to you knowit only affects thewomen in one side ofthe family. I’ve heard awhole bunch of differentcrazy different things.”b“So let’s say that it’s a60% chance that it’sabsolutely going toprevent rheumatoidarthritis later in my life,and there’s a herbaltreatment which is, like,55%, 50%. Thatmassively changes whatmy personal treatmentplan is.”bRheumatologists “But from our point ofview is it safe to saythough that we, too, ifthere was goodevidence thatnormalizing endosmosis,or that weight loss orsmoking cessationreduces [RA], we wouldbe more at ease withthat sort of interventionthan an intervention thatinvolves medicationswith toxicity?”b“I think that a really, reallystrong, good solidscientific placebo controlor analyzed control, let’sdo it, I’ll push for it. Butbefore that it is do noharm and that is how Iapproach my patient.”a“I think that if you’reable to profilerheumatoid as to thosepatients who have reallyterrible diseases, youknow, you can get itunder control … andyou were able to givesomething really, Iwould feel that thosepatients that I would bewilling to do [preventivetreatment].”a“Is there a marketingapproach that wouldchange actual behavioror compliance in allthat sort of thing. Ithink that’s one thingmedicine really hasn’t –you know, drugcompanies do it all thetime, but that is to selldrugs to us not to thepatient”b“Patients want a cure,and patients want acure naturally, right?And natural is perceivedas being with no risk,which is not alwaystrue.”a and b next to quotes indicate moderate and high importance/representativeness, respectivelyMunro et al. BMC Rheumatology  (2018) 2:18 Page 6 of 10employment, and ability to maintain an active lifestyle.These attributes, however, had the potential to overlap withone another and not be mutually independent. Thus, in oursecond round of focus groups we were able to explore thistheme further with participants to identify two key mutu-ally exclusive attributes: “Method of Administration” and“Risk and Seriousness of Side Effect.” These attributesreflected our finding that all participant groups valued in-formation on whether preventive treatments would begiven as a tablet or injection, require regular monitoringfrom a health care professional and have side effects bothin the short term and long term. They focused largely onwanting to know if treatments would limit their travel, em-ployment, and sports/activities. All groups valued havinginformation on the first-degree relative’s “Risk of RA andRisk Reduction with Treatment,” and rheumatologists inparticular noted the value of information on the marginalbenefit of treatment compared to no treatment.Table 4 Summary of potential attributes, their levels, and supporting quotes compiled from the Framework AnalysisFinal Attributes Lay Terminology Key Quotations from Qualitative Data Suggested Labelsof Possible Levels1) Accuracy of Test How accurate is the testin predicting rheumatoidarthritis“I guess I want to know how accurate the testis, and if there is any chance that you couldmaybe be told like oh, there is a very goodchance of you getting it, but maybe findingout later that that actually wasn’t true.” –First-degree relative“Because if you don’t get IgA [rheumatoid factor],for example, up to 50%, first degree relatives wouldbe positive and I doubt all those are going to getarthritis, so.” - Rheumatologist• High• Medium• Low2) Certainty in Estimates How strong is the evidencefor the test and preventivetreatments“Whether there was enough evidence to showthat that treatment actually has a chance ofpreventing.” – Patient“Is there any data saying that coming from ahigh risk situation, what is the reduction?” –Rheumatologist• Moderate• Limited• Very limited3) Method of Administration Whether it is an infusion,injection, tablet.“You know, I went to Europe last year with mywife. We were gone for, you know, half a year.Now if I wasn’t able to do that because I hadto go to a specific doctor twice a week to getthis thing, no thanks. I’m good.” – First-degreerelative• Infusion• Injection• Tablet4) Risk of RA and Risk Reductionwith TreatmentThe risk of developingrheumatoid arthritis withoutvs. with treatment“Me personally, never [would consider testing].Unless it’s 100% positive. Just with the testturn out.” - Patient“[I would consider testing] if there were perhapsa treatment that were extremely preventive andvery effective at lessening the risk of developingsuch a disease.” – First-degree relative• High• Medium• Low5) Risk and Seriousness ofSide EffectThe risk of a side effectfrom treatment“And I’ve had side effects with - I had a heartattack. I had my kidneys at stage - just thestage before. I needed to have dialysis, so.You know, there is side effects that youget that you have to watch out for.” - Patient“Especially because of watching my mom withprednisone, if there’s anything that increase themental risk that would be like huge for me.” -First-degree relative• Major irreversible;minor reversible• Major reversible;minor reversible• Minor reversible6) Who Recommends Whether it is a healthcare professional, patient,or relative who recommendsit“[If I] learn that I had a high risk of developingRA, I would probably talk about it to peopleand then that is why I came up with who recommendsit being important. And I think I would have to hear itfrom at least two sources to act on it,” – First-degreerelative• Health care professional• Patient• Relative7) Opinion of Health CareProfessionalWhether a health careprofessional or patientsupports/wants to take testand/or preventive treatment“I think that I also have a lot of trust at this point inwhat health care professionals say. And a lot of myown opinions, and ultimately in the end, like it wouldbe my own opinion, but I just think a lot of my ownopinion would come from what the doctor said” -First-degree relative• Health care professionaldoesn’t prefer• Health care professionalis neutral• Health care professionalprefersMunro et al. BMC Rheumatology  (2018) 2:18 Page 7 of 10Patients and first-degree relatives placed high value on“Who Recommends” the testing and treatment program,as well as “The Opinion of the Health Care Profes-sional.” First-degree relatives wanted multiple perspec-tives from individuals who had experience with RAincluding patients, nurses, physicians, and rheumatolo-gists. Data from our Round 2 focus groups withfirst-degree relatives reinforced this observation that theopinion of the health care professional was highly valuedby first-degree relatives. These attributes were related tothe sub-theme of Needing More Evidence, discussedabove. The cost of testing and treatment was an attri-bute directly identified by both first-degree relatives andpatients that we chose to exclude from our proposedDCE as there is significant variation in costs associatedwith medications in Canada depending on patient insur-ance coverage.Finally, analysis of focus groups revealed how partici-pants might interpret the attributes and levels, what layterms are most understandable, their thought process inmaking trade-offs between different attributes, and per-sonal experiences that may modify perceptions andvalues of RA. These decision-making processes andpersonal characteristics have been integrated into oursurvey development and analysis. For instance, demo-graphic questions about age, sex, ethnicity, and, if sur-veying first-degree relatives, which of their relativeshave RA have been included. We also include questionson attitudes to risk, preventive behaviours, and the im-portance of shared decision-making and discussion ofcost in treatment decision-making. The background in-formation also includes recorded information sectionson RA, the attributes of the DCE and the process ofcompleting the DCE survey, based on focus group dis-cussions and feedback.DiscussionOur study suggests that patients with RA, first-degreerelatives of patients, and rheumatologists may value andmake trade-offs between seven different attributes whenmaking the decision for testing and preventive treatmentfor RA. All valued the accuracy of testing, due to con-cerns about false positives, and valued certainty in esti-mates of the test and preventive treatment, due toperceptions that treatment and testing for preventingRA is based on emerging evidence. Patients andfirst-degree relatives desired this evidence from a rangeof sources and their preferences were modified by thestrength of recommendation from their health care pro-fessional. The method of treatment administration wasimportant for patients and first-degree relatives due tolifestyle concerns, while rheumatologists were concernedwith how method influences adherence to treatment.Previous DCEs involving RA decisions have been ad-ministered to the general population [24], those at highrisk of developing RA [15], patients [23, 27, 28], andrheumatologists [29], with the majority focusing ontreatment preferences [15, 23, 24, 27, 28]. One explora-tory binary choice experiment sought the preferences ofthose at high risk of developing RA, finding that thoseindividuals preferred treatments that had high reductionin the risk of developing RA and low risk of a seriousside effect [15]. This study was informed by a qualitativeexploration of the perspectives of first degree relatives inthe context of willingness to enroll in preventive treat-ment trials, which was different from the populationcompleting the choice experiment [32]. Nonetheless,their qualitative findings are similar to ours and suggestrelatives prefer to avoid injections/impact on lifestyleand risks of side effects, experience uncertainty aboutwhether they will get disease/if a treatment will help,and prefer prevention via routes other than drugs. How-ever, the study did not explore the potential inaccuraciesin testing for risk of RA. Furthermore, it did not seek toexplore the tradeoffs made between treatment attributes,which may vary depending on the treatment available. Asemi-structured interview study involving 34 firstdegree-relatives of RA patients in the UK, Germany, andAustria found, similar to our study, concerns about thecertainty of predictive testing and expectations for needfor support and opinions when considering and inter-preting test findings [16]. However, this study did not in-vestigate the preferences of RA patients and careproviders. Thus, our research provides critical, novel evi-dence on the attributes of choice of preventive RA test-ing and treatment, not only from the perspective ofthose at risk of developing RA (first-degree relatives),but also from RA patients and rheumatologists who maybe involved in the decision-making process. While ourliterature review of previous DCE’s included five of theseven candidate attributes identified through our study,our analysis further identified who recommends the test,and the opinion of the health care professional as poten-tial attributes important to patients and their first-degreerelatives. This reflected participants’ trust in their healthcare provider and increased willingness to try a test orpreventive treatment that provider recommended.The variation we observed between the values of rheu-matologists and patients/first-degree relatives highlightsthe importance of eliciting patient preferences at thepoint of clinical decision-making and incorporatingthose preferences into care. Patient preferences for test-ing and treatment for preventive RA may be influencedby their firsthand experiences, the advice of importantothers, the strength of the recommendation from theircare provider, and the quality of the evidence. A modelof shared decision-making may prove useful for creatingMunro et al. BMC Rheumatology  (2018) 2:18 Page 8 of 10a mutual understanding between patients and providersof the factors that modify their preferences such as, forinstance, the lifestyle factors that may influence adher-ence to treatment [30].This study is strengthened by its systematic qualitativeapproach to the development attributes and attribute-levelsfor a DCE, a process that is recommended but typically ei-ther not conducted or is underreported in the literature[31–38]. A previous systematic review of 254 publishedDCEs in health care found that 44% (n = 111) did not men-tion at all any use of qualitative methods [38]. Our ap-proach may be adapted for future DCE developmentstudies. Further, including multiple perspectives in ourfocus groups may allow for the development of a genericDCE that can be administered with patients, first-degreerelatives, and health care professionals and allow for directcomparison of preferences. Inclusion of multiple perspec-tives also mitigates the risk of excluding or missing import-ant attributes, and avoids researcher bias which couldoccur if attributes were elicited from a literature reviewalone [22, 33].One potential limitation, as with all qualitative studies,is that findings may not be adaptable to all jurisdictions,only to populations and health systems similar to oursampling frame. The first-degree relatives in this samplemay be more highly engaged and educated because offamiliarity with their relative’s disease, however thepatient population was drawn from the general public ofarthritis patients. We did not initiate a third round offocus groups to conduct a final review of the selected listof attributes and their levels, and instead relied on theinvolvement of our patient partner to ensure that thelanguage we used was appropriate for a lay audience,evoked the conceptual labels identified in our qualitativeanalysis, and to confirm that the attributes were mutu-ally exclusive with no conceptual overlap. The small sizeof the first-degree relative groups was due to drop out(two participants did not show and were lost tofollow-up). To ensure saturation in spite of small samplesize, probing and discussion in the first-degree relativegroups continued until the researchers felt they had afull understanding of the participants’ perspectives andthere were sufficient examples in each category to iden-tify the characteristics of attributes [39]. We will alsoassess the comprehensibility and usability of the laylanguage descriptions when piloting our DCE survey ina small online sample.Our focus group method offers a novel approach to at-tribute development and, in comparison to one-on-oneinterviews alone, offers the opportunity for participantsto share different experiences and generate sharedunderstanding. At the analysis stage, our FrameworkAnalysis approach allows for synthesizing multiple per-spectives and clear documentation of choices made byour research team about inclusion/exclusion of potentialattributes and meaning of attributes. This meticulousprocess may be adopted by other research teams to en-hance the rigour and transparency of choice experimentdesigns.ConclusionThis qualitative study involved exploring patient,first-degree relative, and rheumatologist preferences forpreventive treatments and provides insight into theprocess of eliciting attributes through qualitativemethods. Findings have shaped the development and ad-ministration of a DCE for choice of testing and treat-ment for RA, which will inform the development,design, and implementation of a preventive treatmentprograms for RA.AbbreviationsCCP: Cyclic Citrullinated Peptide; DCE: Discrete Choice Experiments;RA: Rheumatoid ArthritisAcknowledgementsWe would like to acknowledge the patients, first-degree relatives andrheumatologists who participated in these focus groups.FundingFinancial support for this study was provided by a 2015 grant from theCanadian Initiative for Outcomes in Rheumatology Care (CIORA).Availability of data and materialsThe datasets generated and/or analysed during the current study are notpublicly available due the requirements of our ethical approval for the studybut are available from the corresponding author on reasonable request.Authors’ contributionsMHa, CK and MHu designed and led the study. SM, LS, KM analyzed andinterpreted the data. LS and SM prepared the manuscript. All authors readand approved the final manuscript.Ethics approval and consent to participateEthical approval for the study was granted by the UBC Behavioural ResearchEthics Board (H15–01948). Written consent was obtained from allparticipants.Consent for publicationNot applicable.Competing interestsCK, KM and LS have no competing interests to declare. SM holds a MichaelSmith Foundation for Health Research Trainee Award 2016 (#16603). MHu isfunded as a Clinician Scientist by the Fonds de recherché en Santé du Québec.MHa is supported by The Arthritis Society Young Investigator Salary Award2016 (YIS-16-104) and a Michael Smith Foundation for Health ResearchScholar Award 2017 (#16813). MHa holds the UBC Professorship inSustainable Health Care, which is funded by AmgenCanada, AstraZenecaCanada, Eli Lilly Canada, GlaxoSmithKline, Merck Canada, NovartisPharmaceuticals Canada, Pfizer Canada, Boehringer Ingelheim (Canada),Hoffman-La Roche, LifeScan Canada, and Lundbeck Canada.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Department of Family Practice, University of British Columbia, Vancouver,BC, Canada. 2Faculty of Pharmaceutical Sciences, University of BritishMunro et al. BMC Rheumatology  (2018) 2:18 Page 9 of 10Columbia, Vancouver, BC, Canada. 3Centre for Health Evaluation andOutcome Sciences, St. Paul’s Hospital, Vancouver, BC, Canada. 4Division ofRheumatology, Jewish General Hospital and Lady Davis Institute, andDepartment of Medicine, McGill University, Montreal, QC, Canada. 5ArthritisConsumer Experts/JointHealth, Vancouver, BC, Canada. 6Arthritis ResearchCentre of Canada, Richmond, BC, Canada.Received: 26 February 2018 Accepted: 11 June 2018References1. Hemminki K, Li X, Sundquist J, Sundquist K. Familial associations ofrheumatoid arthritis with autoimmune diseases and related conditions.Arthritis & Rheumatism. 2009;60(3):661–8.2. Linos A, Worthington JW, O’Fallon M, Kurland LT. The epidemiology ofrheumatoid arthritis in Rochester Minnesota: a study of incidence,prevalence, and Mortality. American Journal of Epidemiology. 1980;111(1):87–98.3. Kvien TK, Uhlig T, Ødegard S, Heiberg MS. Epidemiological aspects ofrheumatoid arthritis. Ann N Y Acad Sci. 2006;1069(1):212–22.4. Humphreys JH, Verstappen SMM, Hyrich KL, Chipping JR, Marshall T,Symmons DPM. The incidence of rheumatoid arthritis in the UK:comparisons using the 2010 ACR/EULAR classification criteria and the 1987ACR classification criteria. Results from the Norfolk arthritis register. AnnRheum Dis. 2013;72(8):1315.5. Costenbader KH, Feskanich D, Mandl LA, Karlson EW. Smoking intensity,duration, and cessation, and the risk of rheumatoid arthritis in women. AmJ Med. 2006;119(6):503.e1–9.6. Källberg H, Ding B, Padyukov L, Bengtsson C, Rönnelid J, Klareskog L, et al.Smoking is a major preventable risk factor for rheumatoid arthritisestimations of risks after various exposures to cigarette smoke. Ann RheumDis. 2011;70(3):508–11.7. Deane KD, O’Donnell CI, Hueber W, Majka DS, Lazar AA, Derber LA, et al.The number of elevated cytokines and chemokines in preclinicalseropositive rheumatoid arthritis predicts time to diagnosis in an age-dependent manner. Arthritis Rheum. 2010;62(11):3161–72.8. Karlson EW, van Schaardenburg D, van der Helm-van Mil AH. Strategies topredict rheumatoid arthritis development in at-risk populations.Rheumatology. 2016;55(1):6–15.9. Nielen MMJ, van Schaardenburg D, Reesink HW, Twisk JWR, van de Stadt RJ,van der Horst-Bruinsma IE, et al. Simultaneous development of acute phaseresponse and autoantibodies in preclinical rheumatoid arthritis. Ann RheumDis. 2006;65(4):535.10. Demoruelle MK, Deane KD. Treatment strategies in early rheumatoidarthritis and prevention of rheumatoid arthritis. Curr Rheumatol Rep. 2012;14(5):472–80.11. Nielen MMJ, van Schaardenburg D, Reesink HW, van de Stadt RJ, van derHorst-Bruinsma IE, de Koning MHMT, et al. Specific autoantibodies precedethe symptoms of rheumatoid arthritis: a study of serial measurements inblood donors. Arthritis Rheum. 2004;50(2):380–6.12. Sokolove J, Bromberg R, Deane KD, Lahey LJ, Derber LA, Chandra PE, et al.Autoantibody epitope spreading in the pre-clinical phase predictsprogression to rheumatoid arthritis. PLoS ONE. 2012;7(5):e35296. MatloubianM, editor13. van Nies JAB, Krabben A, Schoones JW, TWJ H, Kloppenburg M, van der HMil AHM. What is the evidence for the presence of a therapeutic window ofopportunity in rheumatoid arthritis? A systematic literature review. Ann.Rheum. Dis. 2014;73(5):861–70.14. Visser H, le Cessie S, Vos K, Breedveld FC, Hazes JMW. How to diagnoserheumatoid arthritis early: a prediction model for persistent (erosive)arthritis. Arthritis & Rheumatism. 2002;46(2):357–65.15. Finckh A, Escher M, Liang MH, Bansback N. Preventive treatments forrheumatoid arthritis: issues regarding patient preferences. Curr RheumatolRep. 2016;18(8):51.16. Falahee M, Simons G, Raza K, Stack RJ. Healthcare professionals’ perceptionsof risk in the context of genetic testing for the prediction of chronicdisease: a qualitative metasynthesis. J. Risk Res. 2018;21(2):129–66.17. Bayliss K, Raza K, Simons G, Falahee M, Hansson M, Starling B, et al.Perceptions of predictive testing for those at risk of developing a chronicinflammatory disease: a meta-synthesis of qualitative studies. Journal of RiskResearch. 2018;21(2):167–89.18. Stack RJ, Stoffer M, Englbrecht M, Mosor E, Falahee M, Simons G, et al.Perceptions of risk and predictive testing held by the first-degree relativesof patients with rheumatoid arthritis in England, Austria and Germany: aqualitative study. BMJ Open. 2016;6(6):e010555.19. Falahee M, Simons G, Buckley CD, Hansson M, Stack RJ, Raza K. Patients’perceptions of their relatives’ risk of developing rheumatoid arthritis and ofthe potential for risk communication, prediction, and modulation. ArthritisCare Res (Hoboken). 2017;69(10):1558–65.20. Louviere J, Hensher D, Swait J. Stated choice Methods : analysis andapplications. Cambridge: Cambridge University Press; 2000.21. Street DJ, Burgess L, Viney R, Louviere J. Designing Discrete ChoiceExperiments for Health Care. In: Ryan PM, M.Sc RKG, Amaya-Amaya RFM,editors. Using Discrete Choice Experiments to Value Health and Health Care.Dordrecht: Springer Netherlands; 2008. p. 47–72. (The Economics of Non-Market Goods and Resources).22. de Bekker-Grob EW, Ryan M, Gerard K. Discrete choice experiments inhealth economics: a review of the literature. Health Econ. 2012;21(2):145–72.23. Clark MD, Determann D, Petrou S, Moro D, de Bekker-Grob EW. Discretechoice experiments in health economics: a review of the literature.PharmacoEconomics. 2014;32(9):883–902.24. Ryan M, Gerard K, Amaya-Amaya M. Using discrete choice experiments tovalue health and health care. Dordrecht: Springer; 2008.25. Coast J, Al-Janabi H, Sutton EJ, Horrocks SA, Vosper AJ, Swancutt DR, et al.Using qualitative methods for attribute development for discrete choiceexperiments: issues and recommendations. Health Econ. 2012;21(6):730–41.26. Augustovski F, Beratarrechea A, Irazola V, Rubinstein F, Tesolin P, Gonzalez J,et al. Patient preferences for biologic agents in rheumatoid arthritis: adiscrete-choice experiment. Value Health. 2013;16(2):385–93.27. Harrison M, Marra C, Shojania K, Bansback N. Societal preferences forrheumatoid arthritis treatments: evidence from a discrete choiceexperiment. Rheumatology. 2015;54(10):1816–25.28. Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the frameworkmethod for the analysis of qualitative data in multi-disciplinary healthresearch. BMC Med Res Methodol. 2013;13(1):117.29. Leech NL, Onwuegbuzie AJ. An Array of qualitative data analysis tools: a callfor data analysis triangulation. Sch Psychol Q. 2007;22(4):557–84.30. Poulos C, Hauber AB, González JM, Turpcu A. Patients’ willingness to tradeoff between the duration and frequency of rheumatoid arthritis treatments.Arthritis Care Res. 2014;66(7):1008–15.31. Kuijper TM, Folmer R, Stolk EA, Hazes JMW, Luime JJ. Doctors’ preferences inde-escalating DMARDs in rheumatoid arthritis: a discrete choice experiment.Arthritis Res. Ther. 2017;19(1):78.32. Novotny F, Haeny S, Hudelson P, Escher M, Finckh A. Primary prevention ofrheumatoid arthritis: a qualitative study in a high-risk population. Joint BoneSpine. 2013;80(6):673–4.33. Makoul G, Clayman ML. An integrative model of shared decision making inmedical encounters. Patient Educ Couns. 2006;60(3):301–12.34. Vass C, Rigby D, Payne K. The role of qualitative research methods indiscrete choice experiments: a systematic review and survey of authors.Med Decis Mak. 2017;37(3):298–313.35. Helter TM, Boehler CEH. Developing attributes for discrete choiceexperiments in health: a systematic literature review and case study ofalcohol misuse interventions. J Subst Use. 2016;21(6):662–8.36. Vosper J, Coast J, Flynn T. Qualitative methods in discrete choiceexperiments: Two case studies. In: Coast J, editor. Qualitative methods forhealth economics. London: Rowman & Littlefield; 2017. p. 175–92.37. Coast J, Horrocks S. Developing attributes and levels for discrete choiceexperiments using qualitative methods. J. Health Serv. Res. Policy. 2007;12(1):25–30.38. Hazlewood GS, Bombardier C, Tomlinson G, Thorne C, Bykerk VP, ThompsonA, et al. Treatment preferences of patients with early rheumatoid arthritis: adiscrete-choice experiment. Rheumatology. 2016;55(11):1959–68.39. Saunders B, Sim J, Kingstone T, Baker S, Waterfield J, Bartlam B, et al.Saturation in qualitative research: exploring its conceptualization andoperationalization. Qual Quant. 2017;52(4):1–15.Munro et al. BMC Rheumatology  (2018) 2:18 Page 10 of 10


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items