Healthcare professionals’ perceptions on barriers and facilitators to DMARD use in rheumatoid arthritis

Semi structured in-depth interviews were held with HCPs involved in the treatment of patients with RA. The Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist was used to ensure complete and transparent reporting [22]. Eligible HCPs were rheumatologists, nurses and pharmacists, experienced in the care for patients with RA. Purposive sampling was used to obtain the perspectives of HCPs with varying professions, sex, age and regions to increase transferability.

Based on an adjusted Theoretical Domains Framework (TDF) [23] an interview guide was developed by the researchers to address the perceptions of HCPs on potential barriers and facilitators to DMARDs that influence patients’ adherence (see Table 1). Questions were formulated reflecting the following 11 domains: knowledge, emotions, attention/memory, decision making, social influences, beliefs about capabilities, beliefs about consequences, motivation and goals, goal conflict, environmental context and resources, skills. The interview guide was pilot tested among three HCPs. After the pilot test, no changes to the interview guide were made. Therefore, the data obtained in the pilot test were used for the final study and data analysis. All interviews were conducted by the first female author (MV) who was fully trained and experienced in qualitative interviewing. The interviews were conducted between April and September 2019. A summary of each interview (member check) was sent to each HCP to ensure their views and beliefs have been interpreted correctly by the researcher.

Purposive sampling was used to obtain the perspectives of healthcare professionals with varying jobs, sex, regions, age and work experiences to increase transferability. No relationship was established between the interviewer and the participants prior to the study. Participants were invited by e-mail, followed up by a phone call. They were interviewed in their own environment at a time and place of their convenience. Written informed consent to participate as well as to record the interview was obtained prior to the start of the interview. The interviews were audio-recorded and field notes were made during the interviews. Each interview was scheduled for 60 minutes. Data saturation determined the number of interviews. Intermediate analyses were conducted by two researchers (JV, MV) to decide whether new themes had surfaced. Relevant quotes of the participants were selected and translated to support the findings.

A demographic questionnaire was completed before the start of the interview (age, sex, region, work experience in years, working in Academic/peripheral hospital).

All interview-tapes were transcribed verbatim by a professional agency. The qualitative data analysis software MAXQDA 2018 (MAXQDA, 1989–2018) was used by two independent researchers (JV and MV) to code meaningful fragments in the text, conform the analysis method described by Boeije [24]. This method uses a structured way of coding [25], the principles of constant comparison [26], and in this study, an inductive thematic analysis. This method was chosen because it facilitates a theoretically flexible approach to the coding of these data. Applying this method to this study implied that the researchers interpreted the data with theoretical sensitivity to the adjusted domains of the TDF framework. Before starting the coding process, the transcripts were read and reread for familiarization with the data. The coding was done in 3 steps: open, axial and selective coding. Open coding is the part of the analysis concerned with identifying, naming, categorizing and describing phenomena found in the transcripts of the HCPs that were interviewed. These open codes were categorized (axial coding). From these axial codes the core themes and interrelatedness between themes were identified (selective coding). To support the coding process, field notes made during the interviews were used. In addition, the researchers made reflective notes about their thoughts and views regarding the identification of themes. Data collection and data analysis alternated in a cyclic process. Throughout this process, 3 authors (MV, BvdB and JV) continuously and repetitively reflected on, compared, discussed, and adjusted the coding in order to carefully determine the number and wording of themes in an iterative design (to ensure confirmability) until consensus was reached (Boeije, 2009).

As HCPs seem to be well aware of the potential barriers interfering with the use of medication by patients, the question arises why these potential barriers to medication use are often not the topic in consultations, and even described as “conspiracy of silence” [33]. Especially given the fact that most modifiable barriers can be addressed in patient education and self-management interventions, such as having trouble to understand instruction inserts, not believing in (the need for lifelong) prescribed medication, not believing medication will always be available, or believing that the condition can be cured [34]. Furthermore, effective (e-health) interventions that are relatively easy to use, can help to address patients’ health literacy, patients’ self-efficacy to manage their condition and may help to overcome the limited consultation time of physicians by offering other supportive ways for patients to discuss their barriers for medication adherence [35]. As for interpersonal factors, a trusting bond between HCP and patient, to be able to discuss adherence barriers and facilitators and providing tailored information in all phases of the treatment process (e.g. before treatment initiation, in the treatment initiation phase and in the treatment persistence phase) were mentioned as facilitating medication adherence. Several studies report that the relationship between healthcare professional and patient is key to enhance adherence and health outcomes [8, 36]. In a recent study of Roodenrijs et al., the presence of adverse events was prioritized as the most important adherence barrier and a good relationship with the HCP was prioritized as the most important adherence facilitator [37]. Although in this study, we have explored, but not prioritized, which barriers and facilitators are important for medication adherence according to HCPs, findings can be interpreted as comparable since the experience of side effects and creating a bond/empathy/trust with the patient, were factors that were mentioned by most HCPs. A good relationship with the HCP, described in the study of Nota et al., involves themes such as mutual respect between HCP and patient, trust and an open style of communication [38].

HCPs endorsed the view that they should create the conditions for shared decision-making (SDM) during their encounters with the patient to create a common playing field with the patient while discussing the pros and cons of medication and medication adherence in different stages of the condition. Although most RA patients prefer SDM, their preference may vary according to the situation they are in (type of treatment and severity of complaints) and the extent to which they experience barriers in getting more actively involved [38]. Unawareness of having a choice is still a major barrier for patient participation. The HCP has an important role as facilitator in enhancing patient participation by raising awareness and offering options, but implementing SDM is a shared responsibility; all parties need to be involved and educated [38].

In a previous maximum difference scaling exercise, patients identified the most important barriers and facilitators [28]. Reducing symptoms, maintaining independency and shared decision-making were found to be the three most important contributors to medication adherence. Factors identified in the current study are in line with these latter observations, although a difference scaling exercise could add knowledge of the relative importance to the factors identified by the HCPs in this study.

The strength of this study is that the qualitative theory-based in-depth interviews were conducted with different health professionals. This provided insight from different healthcare professionals’ perspectives into patients’ facilitators and barriers to medication use.

As limitations, the use of a predetermined interview guide based on an adjusted TDF model [9, 39, 40], could be considered as a stringent way to identify possible barriers and facilitators for medication use, however, by adding a question if there were factors that HCPs felt important to mention beside the factors already discussed, this potential limitation has been accounted for. No additional factors have been mentioned at the end of each interview.

In addition, the sample size of this study was limited to 15 healthcare professionals. Yet, saturation was reached in the 13th interview. This is in line with the findings of a review study by Guest et al. [27] into saturation and variability. They found that saturation can occur within 12 interviews and elements of key themes can already be presented after six interviews. Data collection and analyses were alternated to ensure that saturation could properly be established.

Furthermore, this study was conducted in one healthcare system of a single European country, which may limit the transferability of the findings, particularly to non-Western countries.