Article Text

Original research
Patient and healthcare team recommended medication adherence strategies for hydroxychloroquine: results of a qualitative study informing intervention development
  1. Shivani Garg1,
  2. Betty Chewning2,
  3. David Gazeley3,
  4. Shelby Gomez2,
  5. Noah Kaitz2,
  6. Amanda C Weber1,
  7. Ann Rosenthal4 and
  8. Christie Bartels1
  1. 1Department of Medicine, Division of Rheumatology, University of Wisconsin-Madison School of Medicine and Public Health, Madison, Wisconsin, USA
  2. 2Department of Pharmacy, University of Wisconsin-Madison School of Pharmacy, Madison, Wisconsin, USA
  3. 3Department of Medicine, Rheumatology, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
  4. 4Department of Medicine, Division of Rheumatology, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
  1. Correspondence to Dr Shivani Garg; sgarg{at}medicine.wisc.edu

Abstract

Objective Patients identified as black and from disadvantaged backgrounds have a twofold higher hydroxychloroquine (HCQ) non-adherence, which contributes to worse lupus outcomes and disparities. Yet, most adherence interventions lack tailored strategies for racially and socioeconomically diverse patients who face unique challenges with HCQ. We aimed to examine a broadly representative group of patients with SLE and physician perspectives on HCQ adherence and adherence strategies to redesign an adherence intervention.

Methods We conducted four virtual focus groups (90 min each) with 11 racially and socioeconomically diverse patients with SLE recruited from two health systems. Additionally, we hosted two focus group meetings with nine healthcare advisors. In focus groups, patients: (1) shared their perspectives on using HCQ; (2) shared concerns leading to non-adherence; (3) discussed strategies to overcome concerns; (4) prioritised strategies from the most to least valuable to inform an adherence intervention. In two separate focus groups, healthcare advisors gave feedback to optimise an adherence intervention. Using content analysis, we analysed transcripts to redesign our adherence intervention.

Results Worry about side effects was the most common barrier phrase mentioned by patients. Key themes among patients’ concerns about HCQ included: information gaps, logistical barriers, misbeliefs and medication burden. Finally, patients suggested adherence strategies and ranked those most valuable including co-pay assistance, personal reminders, etc. Patient and healthcare advisors informed designing a laminate version of an adherence intervention to link each barrier category with four to six patient-recommended adherence strategies.

Conclusion We developed a patient stakeholder-informed and healthcare stakeholder-informed tailored intervention that will target non-adherence at the individual patient level.

  • lupus erythematosus, systemic
  • health services research
  • patient care team

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Most of the analysed data were included in the paper, tables, figures and supplemental files.

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

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

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Most adherence interventions lack tailored strategies for diverse patients who face unique challenges with hydroxychloroquine.

  • Providers and staff perspectives have not been included in current adherence interventions that explains why these are not routinely used in clinical settings.

WHAT THIS STUDY ADD

  • Our study delivers a patient stakeholder-informed and healthcare stakeholder-informed brief tailored adherence intervention that targets non-adherence at the individual patient level and will have high uptake in busy clinical settings.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • Our patient-informed and healthcare-informed intervention could promote and facilitate adherence discussions via shared decision-making in patients with lupus with their healthcare providers.

  • Future studies will aim to implement and examine the impact of this intervention on improving adherence and outcomes in clinics with limited resources.

Introduction

Hydroxychloroquine (HCQ) is the cornerstone of SLE therapy as it improves damage-free survival in all patients with SLE.1 However, non-adherence to the pivotal therapy—HCQ—is alarming.2–4 HCQ non-adherence is strongly correlated with a sixfold higher risk of severe lupus flare, 45% higher rate of lupus hospitalisations and an eightfold higher mortality risk.2–6 Moreover, patients from black racial group face a twofold higher HCQ non-adherence risk,2 7 and those from disadvantaged backgrounds face even worse adherence rates, highlighting that such groups could face unique barriers leading to non-adherence.3 8 However, most adherence interventions lack diverse patients’ insights and tailored strategies to address their unique challenges with HCQ.

Investigating patient insights and recommended strategies could address the discordance between patients’ and clinicians’ perceptions regarding disease activity and treatment adherence.9–11 Previous studies highlight that patients’ perceptions of and experiences with the health system, physicians, medication effectiveness and side effects influence their adherence to treatment.12–14

Medication non-adherence involves an interplay of different factors and thus requires a qualitative synthesis of different patient experiences and perceptions to develop a conceptual framework model that explains underlying themes and interactions that may lead to non-adherence. Sun et al classified medication barriers according to components of the Capability, Opportunity, Motivation, Behaviour change theory.15 A recent study highlighted that effective patient-physician communication and understanding patients’ desires and suggested strategies could improve medication adherence.8 Building on this work, an intervention was developed based on the key domains of two gold-standard general adherence interventions (Medication Adherence Self-Report Inventory-Visual analogue scale and Brief Medication Questionnaire).16 17 The prior adherence intervention was developed using published information on categorised barriers to assess non-adherence and address barriers (online supplemental figure 1A–B).17–19 Furthermore, these barrier categories were linked with key adherence strategies to help clinicians facilitate adherence counselling in visits. The intervention involved a two-step process in which patients completed the brief intervention and reported barriers/concerns followed by clinicians individualising their conversation to address patient responses. This adherence intervention was created to improve clinicians’ skills to identify and target SLE non-adherence to improve care; however, feedback from patients and healthcare stakeholders was not included.20 Given that current validated adherence interventions are rarely used during clinic visits, our earlier first step was to determine the feasibility of administering the intervention during routine SLE visits.21 In that prior feasibility study, we found that the prior intervention was completed in 97% of 112 consecutive visits and the time spent to address non-adherence was 2.6 min. After establishing feasibility, we aimed now to address the next critical gap by incorporating patient input on adherence strategies that would work for them. We also aimed to collect patient and healthcare advisors’ feedback to optimise implementation of the intervention without burdening clinics.

Supplemental material

Methods

Study design

This focus group study was designed to elicit patients’ experiences and perceptions regarding lupus medications and to obtain a list of patient-suggested and ranked strategies, and healthcare staff and provider feedback to inform our intervention. Consolidated Criteria for Reporting Qualitative Research was used to designing this study.22

Setting and participants

Patient advisors: clinical staff at two universities screened patients with lupus on HCQ using a patient self-report visual analogue scale or pharmacist review during their routine visit to identify eligible patients. Patients on HCQ and reporting concerns with the medication, those who had active or a history of gaps in refills or were missing >5–6 doses per month (≤80% adherence) were eligible to participate in this study. Their provider asked if they would be interested in participating, and interested patients received a study flyer. They were then contacted by the research team via phone. The research team gave: (1) further details about the study, (2) verified eligibility and (3) obtained verbal consent. An information sheet was mailed to recruited patients.

The sample size selection and the number of focus groups were determined based on our team’s prior qualitative work,23 24 and feedback from our community advisors and expert facilitators. We purposefully recruited representative patients from two racially and socioeconomically diverse Wisconsin cities: Madison and Milwaukee. We recruited a broadly representative group of 11 patients from different age groups. We sought to include at least one male member, from non-white racial/ethnic groups, with severe lupus defined as ≥1 vital organ involvement, and with social challenges identified using social determinants of health documents in the electronic health record (EHR), as such factors can affect adherence. All 11 patients were invited to participate in all 4 focus group meetings.

Healthcare advisors: we recruited nine healthcare team members with equal representation from four clinics from two diverse academic centres (University of Wisconsin (UW)-Madison and Medical College of Wisconsin). Members from each centre included a registered nurse (RN), a medical assistant (MA), a physician, a pharmacist and the medical director (MD) of the UW rheumatology clinics.

Data collection

We developed six semi-structured virtual focus groups; four patient focus groups and two healthcare advisor focus groups. The interview guide was developed based on a literature review of patients’ and providers’ experiences with HCQ and other medications,13–15 and by adapting a focus-group guide used in a prior study on smoking cessation in patients with autoimmune diseases and communication in paediatric diabetes clinics.23 24 Leveraging blueprints from our team’s prior qualitative work23 and expert facilitators, we planned an onboarding session and a series of four meetings for continuous stakeholder engagement. Three investigators facilitated six consecutive focus group meetings. Each virtual focus group meeting lasted for 90 min and was audio recorded. All audio recordings were subsequently transcribed. During focus groups, notetakers (NK, SGo) documented the order of speakers and non-verbal cues (eg, body language) to supplement the transcriptions.

Patient focus group: interviewers used a semi-structured interview guide to gather information regarding patient experiences with lupus medications, especially HCQ, their beliefs and assumptions, challenges encountered with HCQ use, emotions elicited regarding lupus and medications and experiences with the healthcare team. Furthermore, patients discussed and prioritised their desired interventions, facilitators and motivators that they believed would be the most valuable and actionable to address underlying concerns to improve adherence. Patient feedback was incorporated to redesign our prior adherence intervention (online supplemental file 1A–B).

Healthcare advisors focus group: the objectives of two sequential healthcare staff and provider focus groups were to obtain feedback to optimise the adherence intervention. We shared the revised adherence intervention after incorporating patient feedback and current workflows with healthcare advisors to obtain feedback on content, language, format of use, feasibility and other recommendations to optimise and deliver a stakeholder informed adherence intervention.

Final survey: a final survey was sent to healthcare and patient advisors to endorse the final revision of the adherence intervention and implementation workflow.

Analytic framework and plan

Descriptive statistics were calculated for patient and healthcare staff demographics.

  1. Patient focus group: we sought information from patient focus group meetings in two major domains: (1) patients’ insights and experiences with lupus medications (HCQ); (2) patient-recommended adherence strategies. An a priori coding scheme reflecting these domains was used. Consistent with a content analysis approach, we used the structurally coded text as an entry point into understanding the range of experiences, perspectives and needs that were identified by patients. Three study team members (SGa, NK, SGo) developed and tested the coding scheme by applying it independently to all focus group transcripts. Adjustments to the coding scheme were made iteratively between each reading until thematic saturation was reached. Discrepancies in coding, codebook structure and generated themes were resolved via discussion to enhance trustworthiness and rigour. Furthermore, we reviewed our analysis with the focus group members during subsequent focus group meetings to obtain advisors’ feedback on the findings. Transcripts were coded using NVivo software.25 Listed frequencies of coded categories for patient insights and patient-recommended adherence interventions were calculated. A list of the frequency of themes was generated from the analysis of patients’ insights about HCQ adherence and the presence of codes for two or more themes in a single patient phrase. From these, a conceptual framework was developed to show the inter-relationships between themes and how these lead to non-adherence. Finally, patients suggested adherence strategies to target non-adherence during clinic visits. These strategies were coded by the study team and a list of adherence strategies was created. Using member checking, we assured that patients endorsed all strategies and their associated placement in respective adherence barrier categories.26 Patients then ranked these adherence strategies in each category from most valuable to least valuable in addressing underlying barriers in the real world.

  2. Healthcare team focus group: healthcare team data were analysed to identify key steps using content analysis to optimise an adherence intervention to deliver a patient-informed and healthcare team-informed adherence intervention and an implementational workflow for clinics.

Results

In total, across six focus groups we engaged 20 stakeholders (11 patient advisors and 9 healthcare staff) from four clinics in two health systems (table 1).

Table 1

Patient and healthcare advisors’ characteristics (n=20)

  1. Patient focus groups: the group of 11 patients included 90% females, 55% identifying as black, 45% with social barriers and 55% with severe lupus (table 1A); all 11 patients were invited to participate in 4 sequential focus group meetings. All 11 patients attend first three focus groups, and 10 patients attended the fourth meeting (90% attendance).

Analysis across the two domains of adherence barriers and strategies generated eight barrier themes and six strategy themes.

Patient barriers and concerns about HCQ and lupus medications

‘Long-term concerns’, ‘worry’ and ‘side effects from HCQ’ were the most frequently mentioned words or phrases during the initial focus group meeting in which patients shared their perspectives and insights about HCQ (figure 1). Qualitative analysis identified eight themes regarding adherence based on patients’ concerns about HCQ as shown in table 2 by frequency.

Figure 1

Patient perspectives and insights about hydroxychloroquine. The size of the word/phrase matches the listed frequency.

Table 2

Eight themes highlighting patients’ insights and concerns about HCQ

Theme 1: medication information gaps and conflicting information

Patients described that receiving incomplete or conflicting information about medications profoundly impacted their decision to start HCQ or discontinue after starting. Furthermore, patients reported that if they were given incomplete information about HCQ, they often resorted to unreliable information sources, like ‘the internet’ (table 2), to gather more information which fueled anxiety about HCQ. A patient stated that “I had to go to the internet… not a great place to get information!”

Theme 2: reasons for stopping or logistical barriers

Commonly, patients starting HCQ reported they experienced early side effects with HCQ and a lack of early perceivable benefits. One patient mentioned that “I started losing hair, I thought it is from HCQ, so I stopped it.” This reduced their motivation to continue HCQ and led them to interrupt therapy. Moreover, patients with severe lupus (eg, renal and central nervous system lupus) were often started on several medications at the same time, which resulted in the logistical barrier of having complex medication schedules. Finally, another major logistical barrier leading to non-adherence was change in social situation such as losing a job or health insurance, lacking transportation to get refills or having unstable housing.

Theme 3: misbeliefs and assumptions

As an overarching theme, patients shared that the role of HCQ in lupus was not clear. Several had knowledge gaps and one patient said that she was completely unaware that HCQ improves survival and prevents blood clots in lupus. Moreover, patients were confused whether the routine blood and urine labs were to monitor lupus or HCQ toxicity. Thus, some assumed that routine labs were to monitor for HCQ toxicity which amplified their concerns that HCQ might lead to multiorgan damage.

Theme 4: emotions and attitudes

Worry regarding potential eye toxicity with HCQ was the most common emotion described by most of the patients. Other patients described how symptoms such as upset stomach or presumed allergic reactions with HCQ led to negative emotions about its use.

Theme 5: comorbidities

Patients also reported their concerns regarding HCQ’s safety with other chronic medical conditions. In particular, they reported that chronic symptoms such as fatigue and brain fog did not improve with HCQ use, or change with missed doses, decreasing their motivation to take HCQ regularly.

Theme 6: medication burden and timing

Patients reported that medication schedules were not individualised based on their preferences and lifestyle. Thus, medication regimens often conflicted with their work schedules leading to non-adherence.

Themes 7–8: duration of use and young age

Patients expressed unsettling feelings and doubts about short-term and long-term HCQ use, especially when they were diagnosed at a younger age and when they had to continue HCQ over the life course even when their symptoms improved.

Conceptual framework of how concerns with HCQ lead to non-adherence

The flow diagram in figure 2 depicts current workflows in our clinics and how patients’ concerns were inter-related and affect adherence at different points. For example, we found that initial gaps in information or ineffective patient-provider communication or insufficient time to describe safety and benefits of HCQ, resulted in negative emotions contributing to reluctance to start HCQ. Moreover, adherence and logistical barriers were not assessed during follow-up visits, thus, lingering concerns and difficulties led to misbeliefs, fear and non-adherence over time. Finally, changes in their appearance, for example, weight gain, hair loss, that started after taking HCQ were commonly attributed as side effects from the medication rather than lupus or steroids. Lack of provider discussion about side effects or concerns during follow-up visits accelerated their worries about continuing medication. Additionally, we noted that each patient faced different challenges that were multifaceted, particularly in those experiencing social barriers, comorbidities and/or severe lupus at a young age.

Figure 2

Flow diagram showing current workflow in clinics, and how patient concerns and themes are inter-related (shown with double arrow) and affect adherence at different time points. HCQ, hydroxychloroquine.

Patient recommended and ranked strategies to improve SLE medication adherence

This domain included six themes highlighting patient-suggested adherence strategies to address underlying barriers and enhance adherence including motivators, filling information gaps and resolving conflict, facilitators, personal reminders, improving communication and building rapport and trust (table 3).

Table 3

Six themes highlighting patient-suggested and ranked strategies to address patient concerns

Theme 1: motivators

Patients perceived positive symptom response with HCQ use as the most valuable motivator to take HCQ regularly. Framing the time and outcome expectancy can help that. One patient mentioned that “knowing that HCQ can take several weeks to months to help my symptoms, encouraged me to take HCQ even when I felt my lupus was active.”

Theme 2: filling information gaps and resolving conflict

Patients suggested that sharing complete or adequate medication information could address patient concerns about HCQ and positively impact adherence. Likewise, sharing reliable curated medication information resources could reduce conflicting information that accentuates worry and misbeliefs about HCQ.

Theme 3: facilitators

Patients expressed that personalisation of adherence strategies facilitated adherence, such as placing their pillbox near their toothbrush. They also conveyed that receiving reassurance on the safety of HCQ and data regarding medications and lupus outcomes were other important facilitators of adherence, ‘Basically, it came down to the lesser of two evils, med is a lot less bad than active lupus.’ Furthermore, they noted that easier medication schedules tailored to their lifestyle and work schedules could help patient adherence to complex medication regimens. Receiving care in multidisciplinary clinics with skilled nurses, social workers and pharmacists facilitated adherence as their concerns and social barriers were addressed swiftly by the team.

Theme 4: personal reminders

Simple strategies such as embedding taking doses at the same time as current habits at bedtime was viewed positively as was using a pillbox to help improve adherence. Using two strategies together helped reduce their forgetfulness to take medications.

Theme 5: improving communication

Patients noted that a non-judgemental approach encouraging positive communication with their healthcare team, and personalised adherence discussions to address concerns at the individual level could improve adherence (table 3).

Theme 6: building rapport and trust

The importance of psychosocial support from a healthcare team for patients was emphasised, given their central role in their lupus care.

Patient ranked barriers and recommended strategies for clinics

Eight categories of adherence barriers were identified based on literature and patient experience, and ≥4 patient-suggested strategies were linked to each category of adherence barrier. During ranking of adherence strategies, patients unanimously ranked discussion on SLE labs and symptom response with HCQ as the most important strategy to address understanding about HCQ adherence barrier. Ninety per cent of patients ranked co-pay assistance and 90-day prescription as the most valuable strategy in the pay or refill barrier category. Patients shared positive experiences with co-pay assistance programme and highlighted the importance of discussing this strategy during visits. Patients shared personal examples of visual reminders to take meds (eg, keeping the pillbox lids open) and prioritised visual and personal reminders to address timing and schedule barriers. Highly ranked strategies for other barrier categories are shown in figure 3B.

Figure 3

(A) Final version of our patient stakeholder-informed and healthcare stakeholder-informed adherence intervention with eight items to assess adherence and categorise adherence barriers. (B) Eight barrier categories crosswalk to top four to six patient-recommended adherence strategies to facilitate adherence discussions between clinician and patient and develop an individual adherence plan using shared decision-making.

Healthcare advisors focus group

A group of nine healthcare advisors from four clinics (table 1B) participated in two sequential focus group meetings. Four clinics included two lupus clinics and two general rheumatology community clinics from two institutions. Providers in lupus clinics were supported by an MA, RN and a pharmacist, while in general rheumatology clinics, they were supported by an MA and RN only. All advisors participated in both meetings to review our modified adherence intervention that patient advisors had revised, and shared feedback to optimise the adherence intervention for clinical use.

Content

All advisors reported that they often face difficulty starting a non-judgemental adherence discussion with patients and thus, they noted that the content of the intervention would be helpful during visits. Additionally, they suggested avoiding adding extremely rare side effects to prevent undue anxiety. All advisors recommended reducing content by including the top four to six most valuable strategies for each barrier category, and combining categories with similar strategies, such as side effects and drug interactions (figure 3A–B).

Language and readability

Advisors recommended simplifying and revising the language of the intervention using the summary for patients with low literacy by the Agency of Healthcare Research and Quality.27 Additionally, advisors recommended reducing verbiage to improve readability. Finally, advisors recommended including specific numbers (natural frequencies) in adherence assessment and strategies for easy understanding.

Format of use

Compared with paper and online versions of adherence intervention, most advisors (90%) recommended using a laminate version of the intervention with a sliding bar to highlight strategies for each specific reported adherence barrier category. Additionally, most advisors suggested using a paper version along with the laminate so that patients and clinicians could complete the forms during the visit, which could be scanned in the EHR for future reference. A few recommended using electronic forms and sharing forms via patient message portals.

Feasibility

Important questions were raised regarding feasibility, including the timing of sharing the adherence form with patients during visits and who should review the findings. All advisors recommended sharing the paper version of the intervention with the patients at the time of check-in so that patients have enough time to review and complete the form. Clinician advisors suggested leaving the completed paper version of the adherence intervention on the keyboard as a visual clue for the clinician to review during the visit. Additionally, physicians reported that pharmacists, when available, should follow-up with the patients between visits after a new medication was started and to discuss and address barriers between visits. Ultimately, clinicians felt that an intervention that takes <5 min to assess and address non-adherence would be feasible for use in busy clinics. This feedback informed the five-step workflow to implement our adherence intervention in clinics.

Other recommendations

After reviewing the intervention, no significant concerns were raised and there was good consensus support. All advisors agreed that the intervention with eight categorised barriers each cross-walking to four to six specific patient-recommended adherence strategies would be a highly useful in-clinic intervention. As a minor change, a recommendation was made to incorporate other medications in the intervention as well.28 Finally, all healthcare advisors recommended including a short (5–10 min) case-based video training for all clinicians including MA, RN, MDs, fellows, pharmacists or residents, before implementing this intervention in clinical settings.

All feedback was incorporated to revise the final adherence intervention and implementation workflow, which were approved by the group. Finally, designers developed the final laminated (printed copy covered with plastic for use) and paper versions for single use.

Final survey

Qualitative feedback was elicited in the final electronic survey which was incorporated to deliver the final version of the intervention. All patient and healthcare advisors endorsed the final version of the adherence intervention (figure 3A–B) and workflow (online supplemental file 2). Most advisors, 90%, reported that the intervention was feasible based on the proposed workflow and knowing that the time spent to complete the intervention was <3 min in our pilot study across 112 consecutive visits.21

Supplemental material

Discussion

Medication non-adherence in chronic diseases like lupus is a key clinical gap according to the National Institutes of Health and Centers for Disease Control and Prevention leading to worse clinical outcomes, higher mortality and perpetuating health disparities.2–4 29 Yet, adherence is often not routinely assessed or addressed during clinic visits.30–32 Current general adherence interventions used in patients with chronic diseases are time consuming and not tailored to individual patient needs.31 Furthermore, adherence intervention, including our prior adherence intervention, lack patients’ and healthcare advisors’ perspectives to increase usability in rheumatology clinics including those with low resources (staff, time, pharmacy or social support).31 32 These are the existing gaps that limit use of these interventions in clinics that are addressed in our study. Our study is one of the few studies that directly incorporated patient and healthcare advisors’ feedback to deliver a stakeholder-informed adherence intervention by: (1) incorporating patient insights on barriers contributing to non-adherence; (2) incorporating patient-recommended and ranked adherence strategies for each barrier category to target non-adherence with strategies that work in the real world; (3) incorporating healthcare advisors’ feedback to optimise the intervention to enhance feasibility, usability and adoption of the intervention in busy clinics; (4) facilitating personalised adherence discussion via shared decision-making at the individual patient level during visits.

Validated adherence interventions are not routinely used in busy clinical settings as they take >10 min and lack patient and healthcare staff informed implementation strategies.16 31–35 Clinicians report that planning adherence strategies in clinics is challenging, noting single ready-made solutions are insufficient to address the adherence barriers that are unique to each patient.31 36 Objective adherence measurement, such as HCQ blood levels, can effectively measure non-adherence.37–41 However, clinicians reported that they face difficulty in starting non-judgemental conversations with patients about non-adherence during visits. Thus, the overarching goal of our intervention development process was to develop a patient-centred approach beginning with the patient identifying their concerns followed by facilitating and tailoring adherence discussions using patient-recommended strategies between patients and clinicians. Additionally, we focused on respecting clinician concerns about consultation time.

Similar to the published literature, we found that worry and anxiety about side effects and long-term use of HCQ were the most common barrier phrases mentioned by patients.13 14 42 Moreover, patients reported that without provider discussions about HCQ’s role in lupus, anticipated time to response and outcome expectancy, patients sometimes received conflicting information from other resources exacerbating their worry and leading to non-adherence. We found that each patient faced different challenges and most of the patient concerns were multifaceted. This could explain why previous interventions that used a single ‘one-size-fits-all’ adherence strategy did not report sustained improvement in medication adherence over time.36 43 We found that developing a multifaceted, personalised adherence plan with individual patients during clinic visits did improve and sustain adherence in a pilot study.44 Thus, findings support the need for a tailored multifaceted adherence intervention that targets non-adherence at the individual patient level.

To develop a generalisable intervention, incorporating racially and socioeconomically diverse patient feedback is important since medication adherence is lower, and outcomes are worse in such groups.2 45 We purposefully oversampled patients with lupus to recruit a broadly representative group with racial and socioeconomic diversity. Similar to previous findings, we found that patients in our study reported that changes in social situation, such as lack of transportation, high co-pays or unstable housing, contributed to higher non-adherence.15 Furthermore, patients in our study mentioned that clinics offering social and pharmacy services provided them with immense support to address logistical barriers and improve adherence. These findings support the need for a multifaceted approach that assesses and addresses social barriers along with knowledge to target non-adherence, which informed the development of our brief adherence intervention.

Limited information is available regarding adherence strategies that would work for patients with lupus in real-world clinical settings.43 This study is one of the first to report patient-suggested strategies, motivators and facilitators of adherence. We noted that the key strategies highlighted by the patients included effective communication with the healthcare team to explain response time and outcome expectancy with HCQ use and personalise a plan to address barriers. Similar to the findings from a previous lupus study, we found that patients with severe lupus often felt more motivated if they were included in medication decision-making, and when the healthcare team valued their input.8 This supports the need for an intervention that facilitates non-judgemental dedicated adherence discussions via shared decision-making with patients, including those who are the highest risk of non-adherence. Most adherence interventions currently lack any guidance for clinicians on how to start feasible personalised adherence discussions with patients. Having patients complete a brief form immediately before the visit gave the foundation for a meaningful, individualised interaction. Thus, using these findings and engaging our patient advisors, we included the top four to six patient-recommended adherence strategies mapped for each of the eight patient-informed adherence barrier categories in our revised adherence intervention.

Despite patient and healthcare staff engagement from the two health systems, we acknowledge limitations. Although our study included racially and socioeconomically diverse members with lupus, we did not compare barriers by race or other characteristics. We used anonymised transcripts based on community advisors’ feedback to enhance patient engagement. However, we incorporated insights and recommended strategies from a broadly representative group of patients so that the refined version of the intervention could assess barriers faced by different groups of patients with lupus and can tailor adherence strategies at the individual patient level. Next, we specifically examined patients’ experiences and perception regarding HCQ use, although based on advisor feedback other lupus medications were included in the final version of the intervention. This is consistent with high non-adherence and similar barriers for other medications.13 14 28 Future studies will examine the effectiveness of this intervention in assessing and addressing non-adherence to HCQ and other lupus medications. Third, only 10% of patient advisors were men and patients with important factors, such as pregnancy, breast feeding or depression, were not specifically included in our study. Adherence strategies could differ in such groups and should be explored in future studies. Fourth, although we included two healthcare advisors from community rheumatology practices, our findings might not be generalisable to all rheumatology practices and will be addressed in future studies. Additionally, we did not include objective adherence assessment, such as HCQ blood levels, as an alternative method to measure non-adherence in SLE.37–41 Measuring only patient-reported adherence could miss some patients who do not report medication non-adherence because of social desirability effect. Thus, future studies will aim to examine comparative effectiveness of using this intervention with and without therapeutic drug monitoring to improve adherence and outcomes in lupus.

In conclusion, our study delivers a healthcare staff-informed and patient-informed and endorsed adherence intervention that assesses adherence and facilitates tailored adherence discussions using patient-recommended strategies to clarify misbeliefs and encourage HCQ use. Our multifaceted intervention is informed by eight patient-recommended adherence barrier categories and top four to six patient-recommended strategies to guide clinicians to target non-adherence.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Most of the analysed data were included in the paper, tables, figures and supplemental files.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by Institutional Review Boards (University of Wisconsin-Madison and Medical College of Wisconsin): 2020-0979 and PRO00038640. Participants gave informed consent to participate in the study before taking part.

References

Supplementary materials

  • Supplementary Data

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

Footnotes

  • Contributors All authors critically reviewed the study proposal. SGa, NK, SGo, BC, DG and CB were involved in collecting or supervising data collection and analysing data. All authors participated in review of analysed data and summarising findings. All authors reviewed and approved the final version of the manuscript. SGa is the guarantor.

  • Funding UW School of Medicine and Public Health from the Wisconsin Partnership Program through a grant to the UW Institute for Clinical and Translational Research (UW ICTR). Additional support was provided by the National Institutes of Health-National Center for Advancing Translational Sciences through a Clinical and Translational Science Award (1UL1TR002373) to UW ICTR.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the 'Methods' section for further details.

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

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