Stakeholder perceptions of preventive approaches to rheumatoid arthritis: qualitative study of healthcare professionals’ perspectives on predictive and preventive strategies

Abstract

Background

There is increasing research interest in the development of preventive treatment for individuals at risk of rheumatoid arthritis (RA). Previous studies have explored the perceptions of at-risk groups and patients about predictive and preventive strategies for RA, but little is known about health care professionals’ (HCPs) perspectives.

Methods

One-to-one semi-structured qualitative interviews were conducted (face-to-face or by telephone) with HCPs. Audio recordings of the interviews were transcribed, and the data were analysed by thematic analysis.

Results

Nineteen HCPs (11 female) were interviewed, including ten GPs, six rheumatologists and three rheumatology nurse specialists. The thematic analysis identified four organising themes: 1) Attributes of predictive and preventive approaches; 2) Ethical and psychological concerns; 3) Implementation issues and 4) Learning from management of other conditions. Theme 1 described necessary attributes of predictive and preventive approaches, including the type and performance of predictive tools, the need for a sound evidence base and consideration of risks and benefits associated with preventive treatment. Theme 2 described the ethical and psycho-social concerns that interviewees raised, including the potential negative economic, financial and psychological effects of risk disclosure for ‘at-risk’ individuals, uncertainty around the development of RA and the potential for benefit associated with the treatments being considered. Theme 3 describes the implementation issues considered, including knowledge and training needs, costs and resource implications of implementing predictive and preventive approaches, the role of different types of HCPs, guidelines and tools needed, and patient characteristics relating to the appropriateness of preventive treatments. Theme 4 describes lessons that could be learned from interviewees’ experiences of prediction and prevention in other disease areas, including how preventive treatment is prescribed, existing guidelines and tools for other diseases and issues relating to risk communication.

Conclusions

For successful implementation of predictive and preventative approaches in RA, HCPs need appropriate training about use and interpretation of predictive tools, communication of results to at-risk individuals, and options for intervention. Evidence of cost-efficiency, appropriate resource allocation, adaptation of official guidelines and careful consideration of the at-risk individuals’ psycho-social needs are also needed.

Background

Rheumatoid arthritis (RA) is a chronic inflammatory joint disease, which causes joint pain, swelling, stiffness and fatigue, as well as joint damage [1, 2] and extra-articular features such as cardiovascular and pulmonary disease which may reduce life expectancy [3, 4].

Advances in understanding of the biological mechanisms underlying the development of RA [5,6,7] and prediction of who is likely to develop RA in the future [8,9,10,11] have laid the foundations for increased research focus on interventions for at-risk individuals to prevent or delay disease development and progression [12, 13]. A number of clinical trials of pharmaceutical preventive treatments are either completed or underway to investigate this, including in asymptomatic first-degree relatives [14]. Early findings are promising and may signify a shift towards RA prediction and prevention, rather than treatment of established RA in future years.

At present, the responsibilities of healthcare professionals (HCPs) focus primarily on the diagnosis and management of RA, including identification of signs and symptoms of RA, communication of treatment options, and provision and monitoring of pharmacological treatments [15]. HCPs may also recommend lifestyle changes such as smoking cessation and weight loss that support the effective management of RA. As the introduction of predictive and preventive approaches would likely affect HCPs’ roles and responsibilities in both primary and secondary care, it is important to understand their views around such approaches for RA. Understanding these views will help to identify potential barriers and facilitators, and support needs that would need to be addressed to inform the design and implementation of effective approaches.

A small number of studies have explored the perceptions and preferences towards predictive and preventive strategies for RA among several different stakeholders. These stakeholders include members of the public, at-risk groups such as first-degree relatives (FDRs) of patients with RA and those with clinically suspect arthralgia (CSA), and RA patients. Stakeholders identified several concerns relating to the accuracy and certainty of the risk information provided by predictive tools, as well as the potential for these tools to cause psychological harm to a person or their family [16,17,18,19]. Compared to at-risk groups who were asymptomatic (such as FDRs), those with symptoms were more likely to take a predictive tool or preventive treatment for RA [18]. Stakeholders also reported a preference for lifestyle compared to pharmacological interventions to reduce the risk of RA [20, 21]. The efficacy of preventive treatments in reducing RA risk was identified as important in stakeholder’s decision-making surrounding the uptake of these treatments [20, 21].

A very small number of studies have included assessment of rheumatologists’ perceptions about predictive and preventive approaches for RA in addition to FDRs and RA patients. Rheumatologists highlighted concerns regarding the cost and accuracy of predictive tools as well as a lack of evidence surrounding the efficacy and safety of pharmacological treatments to reduce risk [22,23,24]. Similar to other stakeholders, this group also expressed a preference for lifestyle interventions [23, 24]. However, differences in preferences between rheumatologists, FDRs and RA patients were also identified; rheumatologists placed more emphasis on the certainty of evidence surrounding preventive treatment compared to FDRs and RA patients, whereas FDRs and RA patients placed greater importance on how a treatment is taken [23]. Rheumatologists were also more likely to choose a preventive treatment for at-risk individuals over no treatment compared to FDRs and RA patients [23]. It is therefore important that the perspectives of all stakeholders are explored and understood.

To our knowledge, no study to date has examined the perspectives of other relevant HCPs, including those from primary care services, towards predictive and preventive approaches to RA. Consequently, the aims of the present study were to explore the perceptions of rheumatologists, specialist nurses and GPs regarding the utility of predictive and preventive approaches for RA, and factors that may affect their implementation within healthcare services.

Methods

Design

This study was a qualitative interview study using a semi-structured interview schedule [25]. Data were analysed thematically using the approach developed by Braun and Clarke, [26] with codes and themes identified using an inductive approach, based on the data obtained in this study [26,27,28].

Ethical approval was granted by the University of Birmingham Science, Technology, Engineering and Mathematics Ethical Review Committee (ERN_18-1781). The Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines [29] were used to report on the methods and results of this study.

Participant recruitment

HCPs were eligible to take part in the interviews if they: (1) managed patients with RA within primary or secondary care settings, and (2) were proficient enough in English to participate in an interview. HCPs who work predominantly with patients under the age of 18 were not eligible to participate.

A sample size of around 10–20 interviews has been suggested as sufficient to achieve data saturation for this type of study [30, 31]. In the current study, interviews were conducted until data saturation was achieved, determined by consensus among the research team that no new information was generated in the interviews, and no new codes or themes were identified in the analysis [32].

GPs were identified either through the NIHR CRN (West Midlands), or with support from CM (an NIHR professor of general practice) at the Midlands Partnership NHS Foundation Trust (MPFT), who identified individuals who met the inclusion criteria for the study and contacted them via email on behalf of the research team. Rheumatologists and rheumatology nurse specialists were identified by KR (a senior consultant in rheumatology) at Sandwell and West Birmingham (SWB) NHS Trust, and by research staff at MPFT, who contacted potential participants either face-to-face or via email.

Participants were recruited using a convenience sampling technique, [33] and were made aware that the interviewer was a PhD student supervised by KR (a senior rheumatology consultant) and MF (a research psychologist specialising in rheumatology).

Procedure

Participants were provided with a background questionnaire and consent form to complete prior to their interview. The background questionnaire assessed gender, professional role, years since qualification and whether the participant had a specialist interest in rheumatology. All interviews were conducted by IW, a female PhD student with a background in health psychology and experience in conducting one-to-one interviews. The interviews were conducted either face-to-face at a mutually agreed location or over the phone and only IW and the participant were present during the interviews. No repeat interviews were carried out. Field notes were made by IW after each interview.

The interviews were guided by a semi-structured interview schedule. The initial draft of the interview schedule was developed by IW, with input from MF and KR. Several open-ended questions were developed to address the study aims, with the first half of the schedule focused on predictive approaches for RA, and the second half addressing preventive treatment. A small number of vignettes relating to the provision of predictive strategies for those at various stages of RA risk (those who presented with RA symptoms, and those who presented with a family history of RA only) were also included.

The questions generated were informed by previous related studies [16, 17, 34]. and through involvement of a panel of stakeholders including an FDR of an RA patient, a GP, a rheumatology nurse specialist, and a rheumatologist, none of whom were participants in the study. As a result of the input of the stakeholders, the interview schedule was updated to [1] clarify the type of preventive interventions the research team wished to discuss, including pharmaceutical treatments, lifestyle interventions, or both, and [2] include prompts relating to the type of predictive tool that HCPs may consider (for example inflammatory markers or imaging). In addition, a face-to-face pilot interview with a rheumatology nurse specialist was also conducted to determine the effectiveness of the interview schedule at addressing the aims of the study. Following feedback from this interview, the schedule was modified further to include interview prompts to further clarify questions relating to perceived challenges and benefits associated with preventive interventions. The final interview schedule is provided in supplementary materials 1.

All interviews were audio-recorded and transcribed verbatim using an independent transcription company (The Transcription Company). Transcripts were not returned to participants for comment and/or correction but were sense checked by the interviewer.

Analysis

The analysis was facilitated by the NVivo software (version 12.0), which enabled coders to record codes identified from the raw data and arrange them into categories and overarching themes [20]. Three researchers (IW, JPK (a medical student) and GS (a health psychologist)) read the transcripts in full to familiarise themselves with the data, and then coded the data line by line. There was an overlap of three of the 19 transcripts which were independently coded by IW, JK and GS without a codebook to assess consistency. The independent coding was compared for the selected transcripts and discussed amongst the three coders to identify inconsistencies. It was agreed that all codes were comparable in meaning, and any differences related only to the description of the codes. A common terminology was agreed and reviewed by a fourth researcher (MF). This was then used to guide further coding.

The resulting codes, and the field notes made after each interview were used to develop initial themes and subthemes, which were continuously refined and developed through regular discussions with KR and MF. This was facilitated by a document collating coded data extracts from all interviews organised into overarching categories. The themes were then further refined by GS using this collated document. IW reviewed these additional refinements, and the final themes were decided upon through discussions with GS, IW and MF. Participants did not provide feedback on the findings. Data analysis was conducted in parallel with data collection to facilitate assessment and revision of the semi-structured interview schedule, if necessary, as well as to assess whether data saturation had been achieved.

Results

Participants

Nineteen HCPs (11 female) from the Midlands (UK) took part in a one-to-one interview either over the phone (n = 18) or in person (n = 1) between November 2019 and July 2021. No new codes or themes were identified from transcripts of the last three interviews conducted, which was interpreted as evidence of data saturation. Participants included ten GPs, six rheumatologists and three rheumatology nurse specialists with an average of 14 years of experience in healthcare post qualification at the time of the interviews. Interviews lasted between 30 and 80 min. Although most interviews were conducted prior to the onset of the COVID pandemic, three interviews with rheumatologists were conducted during the pandemic in July 2021. The characteristics of individual HCPs are summarised in Table 1.

Table 1 Participant characteristics

Thematic data analysis

The analyses resulted in four organising themes, each with a number of subthemes (Table 2). The first theme, ‘Attributes of predictive and preventive approaches’, encompasses HCPs’ views on those aspects of predictive tools, preventive treatments and other interventions that were considered important. Theme two, ‘Ethical and psychosocial concerns’, deals with HCPs’ views about the potential psychological and social/ ethical consequences of predictive approaches. Theme three, ‘Implementation issues’, covers what HCPs described as necessary to integrate predictive and preventive approaches into RA care and the impact it might have on healthcare services. Finally, theme four, ‘Learning from management of other conditions’, describes the lessons that can be learned from HCPs’ experiences of predictive and preventive approaches in other disease areas.

Table 2 Overview of organising themes and subthemes

1) Attributes of predictive and preventive approaches

This theme deals with the attributes of predictive and preventive approaches the interviewees considered important. The four subthemes are outlined below and supporting quotes can be found in the text and in Table 3.

Table 3 Quotes illustrating Theme 1 ‘Attributes of predictive and preventive approaches’

Type and performance characteristics of predictive tool. HCPs, and in particular rheumatologists, discussed a number of factors they deemed important for predictive tools. They considered factors such as: What type of predictive tools might be appropriate, describing blood tests including autoantibodies (Table 3, Quote 1; T3Q1); Potential issues with interpretation of test results, for example an elevation in inflammatory markers not being specific for inflammatory arthritis (T3Q2); And the performance characteristics of available tools themselves including the need for appropriately high positive and negative predictive values (T3Q3). Complicating matters for many interviewees, was a recognition that RA itself was a heterogenous disease and that tests may not be able to predict who might be likely to face a severe disease course versus those who would be likely to face a mild course (T3Q4).

Availability of, and evidence base for, preventive approaches. Several HCPs stated that risk assessments should only be carried out if preventive interventions are available (T3Q5). Others mentioned that they would need to be assured that such intervention would have sufficient benefit and were worried about exposing individuals who might not actually require treatment to unnecessary side effects from drug treatments (T3Q6). Interviewees also discussed the need for evidence of the preventive treatment’s effectiveness, and concerns about how long the preventive treatment may need to be given for:

“I think for preventive interventions, you probably wouldn’t want to treat somebody for longer than a year or two years without clear evidence to do that”. PPN 08, Rheumatologist.

Risks associated with the preventive treatment. GPs, rheumatologists, and nurse specialists all recognised that preventive treatments (some of which currently under consideration are treatments for established RA) come with risks of (potentially severe) side effects:

“So, the DMARDs, are, you know, particularly unpleasant drugs and I regularly see patients in practice who experience significant side effects from them and end up having two, three before they, you know, get put onto something, you know, immunological agents.” PPN06, GP.

Some suggested they would not expose patients to such treatments if they did not have symptoms (T3Q7-8). Some also recognised the burden that the preventive treatment might pose for at risk individuals, including having to take the medicine, time commitment and having to be monitored (T3Q9), which might prevent treatment uptake.

Perceived benefits of predictive tools and preventive treatments. HCPs also described potential positive outcomes of predictive testing and preventive treatments. Some HCPs hoped that risk assessment would influence people’s lifestyle choices and lead to changes in behaviour such as smoking cessation (T3Q10) which in turn might reduce their chances of developing RA. Preventing future pain and suffering for patients was also a priority for interviewees (T3Q11). HCPs also recognised that by reducing the risk of RA you also reduce the risk of co-morbidities of RA such as cardiovascular disease (T3Q12). Another positive consideration, identified in particular by GPs, was that if patients knew about their risk, they would be more likely to be vigilant for symptoms and present early if they developed them (T3Q13). For many the possibility of earlier intervention if RA were to develop, and potential prevention of RA, were seen as the main benefit of predictive tools.

2) Ethical and psychosocial concerns

This theme describes potential psychological and social/ ethical consequences of predictive approaches. The three subthemes are outlined below and supporting quotes can be found in Table 4.

Table 4 Theme 2: Ethical and psychological concerns

Social/ ethical consequences. Interviewees discussed the social consequences of predictive tools and subsequent treatment including the potential implications for health or life insurance availability or costs (T4Q1-2) or potential impact on future employment prospects (T4Q3). Some participants expressed concerns that the results of predictive tools could raise problematic questions, for example if it were possible to determine someone’s future risk of RA at a (very) early age, should we offer treatment at that very early point:

“Let’s say you develop a predictive test, but it’s actually genetic based and you could even look at the foetus to determine their risk of rheumatoid arthritis in due course. If you could say that they are at risk of developing severe rheumatoid arthritis from their early 20s onwards, there’s then very serious questions about, well, do we offer treatment from birth, or do we even allow the birth to proceed? There’s potentially quite severe knock-on consequences for these sorts of tests.” PPN 18, Rheumatologist

Psychological consequences. Interviewees discussed the potential for negative psychological consequences of risk communication, (T4Q4-5), especially in relation to uncertainty around disease development:

“I really wouldn’t want to do a test that would give them a nebulous risk of you know developing a condition that they might never get […] and that would cause anxiety and aside from the lifestyle changes I’d recommend anyway”. PPN 01, Rheumatologist.

Uncertainty. HCPs described concerns about the uncertainty associated with risk information, particularly in relation to the inability to guarantee that RA would or would not develop (T4Q6-7). They also described additional uncertainty around issues such as the likely severity of RA if it were to develop (T4Q8), and lack of specificity of predictive tools that are currently available (T4Q9). Such considerations were associated with concerns around the potential for overtreatment of individuals identified as being at risk to prevent RA development (T4Q10).

3) Implementation issues

This theme centres around what would be needed for effective integration of predictive and preventative approaches to RA, and the potential impact of these approaches on the health service. The five subthemes are outlined below and supporting quotes can be found in Tables 5 and 6.

Table 5 Theme 3: Implementation issues (Knowledge and training needs; HCP roles)

Table 6 Theme 3: Implementation issues (Resources; Guidelines; Patient factors)

Knowledge of prediction and prevention and training needs. A number of HCPs indicated that they were not aware of existing risk prediction tools for asymptomatic individuals (T5Q1), whereas others mentioned using the currently available diagnostic tools for RA (T5Q2). Some rheumatologists described existing predictive tools but often noting that these are not integrated into routine care (T5Q3-4).

Whereas some HCPs described a lack of awareness of preventive strategies for RA (T5Q5), other than the possibility for lifestyle interventions (T5Q6), several rheumatologists in the current study were aware of clinical trials looking at RA prevention (T5Q7). Some were also aware that the evidence base for preventive interventions is very limited at present (T5Q8).

HCPs further identified specific training needs to be met for them to deliver predictive and preventive approaches for RA effectively. These include training in communication of risks and benefits associated with preventive interventions (T5Q9-10) and providing psychological support to patients (T5Q11). HCPs also stated they would need training on how to use predictive tools, and interpret the results:

“We would need training and education in how to use the tests and what the results actually mean, and then we would require capacity to do that from a clinic setting.” PPN 19, Rheumatologist.

HCP roles. HCPs identified potential responsibilities they could take on in the prediction and prevention of RA, as well as those for other HCPs. GPs felt they were well placed to prescribe lifestyle interventions, (T5Q12-13) pharmacological interventions were perceived to be more appropriately provided in secondary care:

“If it’s drugs, then I would say that currently, unless the methotrexate and rituximab come with very, very, very specific instructions, then I’d still suspect that that would need to be done in secondary care, or certainly initiated in secondary care” (PPN 09, GP).

Specialist nurses were seen as particularly well suited to have discussions with patients around risk information and preventive treatment for RA, as they may be likely to spend more time with patients than other HCPs and have the relevant skills to discuss this information in a sensitive manner (T5Q14-15). Interviewees also identified other HCPs such as occupational therapists and physiotherapists who could play a role in facilitating lifestyle interventions to prevent RA development (T5Q16). HCPs felt that tests could be interpreted accurately by a range of HCPs as long as they received appropriate training (T5Q17). HCPs also described how they would refer a patient/ an individual at high risk of RA to a rheumatologist (T5Q18-19).

Resources. This subtheme incorporates discussion of the potential costs (T6Q1-2) and funding implications (T6Q3-5) of the integration of predictive and preventive measures into the UK health care systems. HCPs invariably felt that the measures need to be cost effective and that there needed to be robust evidence of cost effectiveness (T6Q6). However, the potential for significant benefit for patients and the wider society was highlighted:

“Because in terms of all risk, if it can be treated sooner or treatment before it kicks in, it’s obviously better for the patient and the healthcare economy in the long term.” PPN 05, Rheumatology clinical nurse specialist.

HCPs further discussed that appropriate resources would need to be allocated to support the increased demands that would be needed to support effective implementation of preventive strategies, such as staff time and expertise, workload, clinic space, and administrative support (T6Q7-8). They also highlighted that requirement for monitoring patients at risk of developing RA would be associated with costs both for the health service and patients themselves (T6Q9). Some HCPs further suggested there might be a need for dedicated RA prevention clinics:

“So, again, as mentioned before, whether we need to then run dedicated clinics or whether we can swap this in amongst our early arthritis or general clinics. I suspect, given that it’s a completely different way of thinking for us, it might be helpful to have its own dedicated clinic, in which case do we need a trained consultant who only deals with that, or a trained nurse who only deals with that?” PPN 18, Rheumatologist.

Guidelines. The need for a clear pathway for management of predictive and preventive interventions alongside existing pathways for RA was discussed (T6Q10-11). To effectively integrate these approaches into the healthcare system, it was noted that they would need to be addressed in extended national treatment guidelines (T6Q12-13).

Patient factors influencing predictive approaches and preventive treatment prescription. HCPs also described patient factors that could affect both HCP and patient decision-making about predictive tools or preventive treatment. These included family history of RA (T6Q14), age (T6Q15) and treatment preferences of the individual at risk:

“I think if we can prevent it, it would be good as long as the patient is happy to take those medications to prevent it.” PPN 04, Rheumatology clinical nurse specialist.

HCPs further shared ideas on how the results of a predictive tool, or risk associated with a treatment plan would need to be communicated to the at-risk individual in a way that would suit the individual and was tailored to their level of understanding/ education and experience (T6Q16).

4) Learning from the management of other conditions

The final overarching theme centres around what HCPs felt could be learned from their experiences of predictive and preventive approaches in other disease areas. The four subthemes are outlined below with supporting quotes in Table 7.

Table 7 Theme 4: Lessons learned from other disease areas

Knowledge of disease. HCPs referred to their knowledge of research into preventive strategies and experience of the clinical translation of that research in other disease areas (T7Q1). HCPs stated that members of the public tended to be less knowledgeable about RA compared with other diseases such as diabetes mellitus (DM) and cardiovascular disease (CVD) and highlighted the impact that this lack of knowledge may have on the uptake of preventive approaches (T7Q2). It was further suggested that the lower prevalence of RA compared to other disease areas limited the amount of evidence for preventive strategies:

“We’ve been able to research that [CVD] quite robustly with the tools we’ve got because it’s such a common disease. Whereas, with rheumatoid arthritis, it’s quite rare actually, isn’t it?” PPN 07, GP.

Treatment. HCPs described their experiences of prescribing preventive interventions for other (chronic) diseases, for example statins to reduce risk of CVD and identified similarities in how they would approach the issue of risk-benefit trade-offs in the context of decisions about preventive treatment for RA (T7Q3). Some favoured lifestyle interventions (T7Q4) or regular monitoring of patients identified as being at risk over preventive pharmacological interventions, describing applications of such approaches in practice or research studies for other conditions, such as diabetes:

“I’m aware of a study being done in type 1 diabetes where they look for a genetic marker and if that genetic marker is present then they do a blood test every year to see if antibodies start to be developed. So, they’re doing sort of blood testing as opposed to exposing patients to treatment, so I see that as being perhaps a better and less burdensome way of managing potential risk.” PPN 06, GP.

Guidelines and tools for other diseases. HCPs discussed their experience of working with existing guidelines for the prediction and prevention of other diseases such as bowel cancer and CVD and used that experience to make suggestions for guidelines for the prediction and prevention of RA as well as pointing out the potential pitfalls of such guidelines. HCPs highlighted the importance of personalised approaches and effective risk/benefit communication to facilitate shared decision making but also warned of over-medicalising healthy individuals, by putting them on preventive treatment (T7Q5). HCPs further worried about burdening the health system, highlighting guidelines that suggested the need to further screen individuals with what they described as non-specific symptoms. (T7Q6)

Existing tools that assess risk were also discussed as examples of predictive tools already integrated into clinical practice, in particular the widespread use of the QRISK score for classifying those at risk of developing CVD:

How are they [HCPs] going to identify it [RA risk] so is it a clinical scoring tool, in cardiovascular, you use QRISK, is there a clinical scoring risk tool for that…? PPN 15, GP.

Risk communication. The final lesson learned from other disease areas was about how risk information and risk reduction strategies could be communicated to those at risk of RA most effectively. HCPs made suggestions based on their experiences of communicating risk for chronic diseases such as CVD and osteoporosis (T7Q7-8), for example, using ‘smiley/sad face’ pictograms or descriptions:

“Well, I haven’t done this within the context of inflammatory arthritis. I mean I do a lot about communication in osteoporosis. … I think there’s lots of sort of generic challenges, even though I don’t know much about the, you know the early inflammatory arthritis example. I mean the first thing is explaining what the condition is that you’re trying to predict and what the significance of that is. And then depending on how the risk is, then you’d be trying to explain that in an understandable way as possible, not using percentages or anything but talking, you know the numeric risks using some simple frequencies.” PPN 17, Rheumatologist.

Discussion

The findings from the current study increase our understanding of the views of HCPs who would likely be involved in the prescription of predictive and preventive approaches for RA. The interview data show that in order to successfully implement predictive and preventive approaches for RA in the current UK healthcare system a number of factors need to be considered. HCPs had clear views about the necessary attributes of predictive and preventive approaches, including the sensitivity of the predictive tool and the need for a robust evidence base for the preventive approach as well as consideration of both risks and benefits associated with preventive treatment. They further raised ethical and psycho-social concerns that they felt needed to be taken into consideration, including the potential negative effects of risk disclosure for the individual, existing uncertainty around the risk of developing RA, the potential for harm (side-effects) and the potential for benefit associated with the treatments being considered. The interviews also revealed a number of implementation issues, including the need for appropriate resource allocation, guidelines, and training around predictive tools and treatment, including interpretation and communication of risk results to patients. Interviewees’ responses were informed by experiences of preventive approaches in other disease areas.

The concerns regarding the accuracy and certainty of RA risk information provided by predictive tools identified in the current study are consistent with results from previous studies examining the perceptions of rheumatologists, members of the public, RA patients and their relatives [16,17,18,19, 22,23,24]. The need to develop tools that provide high positive and high negative predictive values are important to ensure the success of these approaches. However, given the heterogenous nature of RA, this may be difficult to achieve [35]. The need to establish the cost-effectiveness of preventive approaches for RA was also consistent with previous research, where cost-effectiveness was identified as an important factor in decision-making around preventive treatment for RA among RA patients and those at risk [20, 21].

Similarly, HCPs’ concerns around the potential for predictive tool results to cause psychological harm to patients align with previous findings from studies examining perceptions of members of the public, RA patients and their relatives [16,17,18,19]. Appropriate support should therefore be provided to the at-risk individuals alongside risk communication. This could be provided by HCPs involved in communicating risk information to patients though they would need to receive appropriate training and tools.

Many HCPs in the current study preferred the idea of prescribing lifestyle-related treatment or regular monitoring for those at risk of RA (to allow early intervention when RA developed) rather than pharmacological preventive interventions. In contrast, participants in studies around preventive treatments for CVD expressed a preference for pharmacological treatment compared to lifestyle-based interventions [36].

Strengths and limitations

Using qualitative interviews combined with an inductive thematic analytical approach provided the opportunity for new concepts to be explored in depth with a different group of stakeholders [27, 31], generating rich and informative data. This study further benefits from extensive research partner involvement in the design of the interview schedule. Furthermore, the results from the study represent the perceptions of HCPs with varying degrees of experiences and a variety of relevant healthcare roles.

However, there are some limitations. Firstly, all interviews were conducted with HCPs who worked in a healthcare setting within the Midlands, UK. Their views may not be representative of HCPs working in other regions. Further research is needed to understand the views of healthcare professionals in other regions within the UK and in other nations with different healthcare systems. Further studies are also needed to understand the views of other types of HCPs involved in management of RA who were not represented here, such as physiotherapists and occupational therapists. Secondly, the use of a convenience sample in this study could have led to potential bias in the types of participants recruited. As such, their views and motivations may not reflect all relevant HCPs. Furthermore, it is possible that some participants were more exposed to research in the area of interest than is typical, as they were recruited through clinical members of research team. However, this is less likely to be the case amongst GP participants, who comprise over half of the sample and are less likely to be familiar with the research area. Thirdly, the predominant use of telephone interviews within this study may have impacted on the data received due to the lack of non-verbal cues and rapport generally gained through face-to-face contact. However, telephone interviews can still provide rich, detailed and high-quality data [37]. Finally, we did not collect data on how many RA patients were in contact with the HCPs in our study in a typical week, or the type of hospital/primary care setting that participants worked in. Further studies are needed to explore how these, and other contextual variables are associated with perceptual variation.

Conclusions

To ensure the successful implementation of predictive and preventative approaches for RA, HCPs across primary and secondary care services need appropriate training around predictive tools, interpretation of results, communication of results to at-risk individuals, and options for preventive interventions. There is a further need for evidence of cost-efficiency of preventive approaches. Appropriate resource allocation and development of national guidelines are also needed, along with the development of risk communication tools and psychosocial support resources. In designing preventive services for RA, much can be learned from other chronic disease areas such as CVD. Implementation studies that take into account the needs identified by HCPs in the current study are required to inform the development of effective future strategies that will be widely accepted and applied within healthcare services.