Appropriating and asserting power on inflammatory arthritis teams: A social network perspective

Abstract Background Therapeutic interventions for people with inflammatory arthritis (IA) increasingly involve multidisciplinary teams and strive to foster patient‐centred care and shared decision making. Participation in health‐care decisions requires patients to assert themselves and negotiate power in encounters with clinicians; however, clinical contexts often afford less authority for patients than clinicians. This disadvantage may inhibit patients' involvement in their own health care. Objective To identify communication attributes, IA patients use to influence and negotiate their treatment with members of their health‐care network. Method A qualitative social network approach was used to analyse data from a larger study that investigated IA patients' overall experiences of multidisciplinary care. Fourteen patients with IA attended individual semi‐structured interviews. Researchers used thematic analysis to identify patterns of assertiveness and influence in the data. Results Participants experienced loss of identity, control and agency in addition to the physical symptoms of IA. However, they had a sense of personal responsibility for managing their health care. Perceptions of health‐care team support enhanced patients' influence in treatment negotiations. Notably, there appeared to be an underlying tension between being empowered or disempowered. Discussion and conclusions The findings have significant implications for treatment decision communication approaches to IA care. A social network perspective may provide a pathway for clinicians to better understand the complexities of communication with their patients. This approach may reduce unequal power dynamics that occur within clinician/patient interactions and afford people with IA agency, control and affirmation of identity within their health‐care network.


| INTRODUC TI ON
Therapeutic interventions for people with inflammatory arthritis (IA) increasingly involve multidisciplinary teams with a focus on patient-centred care and shared decision making. [1][2][3][4][5] A typical approach to IA care (Box 1) acknowledges patients' rights to be involved in their own health-care decisions, [8][9][10][11] appears to limit consequences of IA and contributes to better outcomes for patients. [2][3][4][5]9 However, studies illuminate how the ideals of patient-centred care 5,12,13 and shared decision making 11,14 do not guarantee patients' involvement in their health-care decisions. There is considerable variability in people's desire and ability to participate. 5,11,12,15,16 Furthermore, the level of desire and ability to participate may change or fluctuate over time as people's physical and psychological distress increases or decreases 17,18 ; they learn more about their disease 5,15,17 ; their disease progresses or stabilizes 11 ; or their priorities shift across the life course. 5 Individuals also need to exert control in their medical negotiations, 15 but this may be difficult for patients. 15,19,20 This is in part due to clinicians' institutionally legitimized expertize which affords them a position of power and control of the encounter and medical decision negotiations. 10,16,[19][20][21][22] Power is the ability of an individual to attain one's goals 22,23 and describes important relational dynamics that shape patient/clinician relationships and patient health outcomes. 8,10,[20][21][22] Power imbalances in the clinician/patient relationship may place patients at a disadvantage and inhibit patients' active involvement in their own care. 9,16,24 Patients who feel in control and actively participate in their health-care decisions experience improved health outcomes and quality of life, and optimized health/disease management. 5,9,25 Therefore, it is critical that patients are partners in their care and negotiate their treatment plans. To achieve this goal, power disparities between clinicians and patients need to be illuminated and addressed. 15,16,22 IA is the most common cause of disability in Canada 1 and requires long-term management and adherence to treatment to minimize joint damage and reduce physiological and psychological symptoms. 5 Variations in models of care 26,27 make IA an ideal condition to examine differences in patient experiences of team care. Research that illuminates how people with IA assert themselves in medical decision making is limited. In particular, there is a lack of empirical research that examines the concept of power in the IA patient/clinician medical encounter from the patient perspective.
We sought to specifically explore how patients, from their perspective, assert themselves in their care network, including whether they assert themselves similarly or differently with various healthcare team members. The goal of this research is to gain a broader and deeper understanding of how patients take control of their treatment negotiations within their care network and identify communication mechanisms patients use to assert themselves and influence their treatment decisions.

BOX 1. IA Multidisciplinary care pathway
The standard care pathway for inflammatory arthritis to reduce symptoms and limit joint damage and disability requires early identification and rheumatology referral, 2,3 and multidisciplinary patient-centred care with rehabilitation professional referral as appropriate for the patient. 3 For example, current treatment recommendations for RA emphasize: • Early, aggressive treatment and a 'treat-to-target' approach (T2T) 6 : the escalation of treatment until a target is reached which is modified when target is no longer being met 6,7 • Target is to achieve and maintain remission or low disease activity in cases of established long-standing disease 6,7 • Medical treatment with non-steroidal anti-inflammatory drugs, disease modifying anti-rheumatic drugs and biologics 2,3,6,7

| ME THODS
This research, set within the Canadian health-care context (Box 2), was part of a larger study that investigated patients' overall experiences of being diagnosed with inflammatory arthritis and their interactions with clinicians and other health-care supporters. The study was approved by a research ethics review board.

| Study design
Drawing on data from a larger qualitative study that explored IA teambased care, we sought to answer the research question 'How do patients with inflammatory arthritis experience influence, authority, and control in their health-care decision within their care network?' A qualitative social network perspective guided the larger study design and subanalysis of data that we report here. A social network paradigm focuses on relationships between actors in a social network. 30 Particular attention is given to individual characteristics and relationships (ties) that enable and/or constrain choice and agency. 30 For instance, communication pathways are ties by which actors in networks interact and exchange knowledge, ideas, views and resources. 30 Different social networks influence individuals' perspectives, values, ideas and knowledge. 30 People attempt to establish and stabilize themselves within a social network by forming an identity that is part of a discourse community of a social context. 31 A person's identity is a composite of identities formed by the various social contexts they have been, and are, exposed to. 30 Typically, networks refer to individuals, teams or organizations. 30 We specifically focused on the ego networks of people with IA. An ego network analysis involves exploring individuals' perceptions and attributes of their network relationships. 30 When people are diagnosed with IA, their lives change which destabilizes their identity. 32 This in-depth analysis was aimed towards a nuanced understanding of patients' identity development, and control and agency within a multidisciplinary team. This may provide novel insights into the flow of power among members of patients' multidisciplinary teams and help us better understand affordances which promote patient control over obtaining their goals.

| Recruitment
Data collection for the larger study employed purposeful sampling to recruit participants 18 years or older, living in a major western Adopting a patient-centred approach, 34 participants generated an ego network that depicted who they perceived to be on their multidisciplinary team. Participants were interviewed by phone or in person.
Demographic data and field observations were collected at the time of the interview. The research team contacted participants at a later date to request an optional follow-up interview to consolidate the data and explore questions that emerged from the primary interviews.
Primary and follow-up interviews were approximately 50 minutes in length. Interviews were audio-recorded, transcribed and anonymized. Participants were provided with the opportunity to review and comment on their interview transcripts.

| Data analysis
We coded transcripts to describe phenomena around how patients perceived their authority to lead their own care within their network.
This analysis followed a three-stage iterative and cyclical systematic process: item analysis, pattern analysis and structural analysis. Item analysis involved compiling groups of similar items of interest (such as understanding IA, taking responsibility and ignoring advice) which led to the identification of primary codes for organizing the data.

BOX 3. Research team background
Research design team.
PI: A social scientist.
Health professionals: a rheumatologist, occupational therapist and physiotherapist.
Public participation: three people with IA.
Research interviewers for this subanalysis. Pattern analysis involved a process of comparison, contrast and integration, organizing items together in higher-order patterns: for instance, preparing for a consultation, expressions of influence and barriers to gaining power. As relationships were the unit of analysis, we looked for ties and relationships between the sets of patterns to generate structures. This structural analysis stage brought together pieces of an analytic puzzle to create an overall conceptually integrated picture of the phenomena under exploration. 35 During structural analysis, we incorporated many initial codes and developed theoretically informed themes that drew on social network theory and were concerned with the interconnection between personal characteristics, network supports and social control. NVivo 11 software helped organize the data throughout analysis.

| RE SULTS
We recruited and interviewed 14 participants (Table 1)  (Jamie) Participants also described various communication attributes (Table 2) and nuanced ways (ties) which enabled them to position themselves within their health-care network and thus gain agency and influence over their treatment decisions. Taking personal responsibility for their health and well-being, and support from their health-care network were important factors for gaining agency and control within their network. Descriptions also provided insights into participants' identities as people with IA.

| Personal responsibility
Personal responsibility for disease management meant taking their health into their own hands, working hard to improve their health, showing an interest in their care, taking their medications and following clinician instructions.

Social network survey question
As a guide for participants we outlined a health-care team as including: (a) at least two health professionals working together (eg nurse, GP, specialist physicians, occupational therapist, social worker), and (b) the patient and (c) informal caregivers (eg family, friends, those who help around the home).
'Who do you perceive as being on your healthcare team?
List as many individuals as you like, they can be health professionals or informal caregivers.' Participants were also asked to identify which members of their network had the most or least influence on their treatment decisions.  (Cam)

Follow-up interview example question
Self-advocacy was perceived as part of participants' network role.
In itself, self-advocacy could be empowering and provide the impetus to challenge proposed medical therapies.

| Team support
The composition of participants' networks varied by number and professional/non-professional identity. Several participants identified ties which supported and empowered them in treatment   Participants described defying advice to take prescribed medications and to rest. However, these expressions of agency and control, at times unknown to other team members, did not appear to fracture relational ties between participants and their clinicians.

| The empowerment/disempowerment dynamic
Overall, there was an implicit dynamic between being empowered and disempowered which appeared to reflect participants' perception of their position and identity within their health-care network.  the ability to act, did not ensure that these participants had power:

| D ISCUSS I ON
they faced barriers to achieve their health-care goals and needs.
This analysis suggests people with IA seek control of their treatment negotiations through agency and establishing identity. An important contribution of this analysis is that acknowledgement of people's illness accounts may be necessary to promote the flow of power through communication pathways.

| Seeking control through agency and establishing identity
Findings revealed personal characteristics and communication attributes where participants attempted to gain agency, identity and exert control in their treatment negotiations. Chronic illness destabilizes a person's identity and sense of self. [37][38][39] People naturally seek to reconstruct their identity within their new social context by seeking control to establish their position within a network. 31 Our findings, similar to the literature, suggest patients with chronic inflammatory conditions invest time, work hard to meet their health-care needs, 38,40 show an interest in their care 17,38,40 and consider it their responsibility to actively participate in their treatment decisions. 5,25 Assertiveness promotes engagement in decision making 11 and may disrupt 16 and inhibit 24 clinicians' attempts of control.
Knowledge acquisition, a core component of an individual's illness identity, 16 enables patients to exert control in their treatment negotiations. 12,16 Knowledge and its relationship to power is intrinsic to patient/clinician encounters. 12 and be dismissive. 16 From a social network perspective, people's views and beliefs can influence the behaviour of other individuals 30 which our analysis suggests may constrain patient agency and the flow of power.
Patients who fear being perceived as difficult, potentially creating conflict and jeopardizing treatment, 15,20 may endeavour to be perceived as a 'good' (normative) patient who refrains from challenging medical authority. 15,16 Our findings imply that participants developed various strategies to be perceived as credible and/or good patients 38 and used different strategies at different points in time to achieve their goals. Socially internalized norms, a good or difficult patient for example, may not assign authority and can thus constrain patient agency. 39 However, conforming to these norms and deferring to clinician authority does not necessarily remove patient agency, identity and control. 16,23 Rather, it was sometimes a way of exerting control.
Agency can be communicated verbally and non-verbally. 31 As reported in our findings, patients may not adhere to their physician's prescribed treatment 8,12,16,21 and may discard prescriptions and change physicians. 8 However, as previously identified in the literature, 8

| Unheard and dismissed illness accounts
Stories, or accounts, can be considered ties through which individuals integrate their experiences from their various social networks. 31 Recounting illness experiences helps individuals understand 21,39,44 and promotes discussion of their illness experiences, enables identity development 31,37,44 and provides a way for individuals to express power and be recognized as an autonomous person. 45 Significantly, our analysis suggests that not being heard or believed rendered some participants powerless to gain agency and control in their treatment decisions. Not being heard, 21,46 listened to, 11,15 believed 40,46 or taken seriously 11,40,46 are barriers to patient involvement in treatment negotiations. Such communication practices, which fracture communication ties, can lead to tension in the clinician/patient relationship, 21 challenge women patients' self-esteem and self-worth when accounts of pain are disbelieved 38,46 and enable clinicians to exert control in the decision-making process. 15,24 Decision practices that limit patients' involvement and autonomy fail to respect the patient's personhood 43 and disempower patients which may constrain disease self-management. 24 Conversely, listening to, honouring and respecting people's illness accounts acknowledges their experiential expertize and promotes letting go of expert authority, power and knowledge. 37,44,45 This analysis implies that for patients to engage equitably in treatment decisions, they need to be able to speak freely, without fear of reprisal, knowing that their voiced illness experiences will be acknowledged and respected.  43 However, professional and lay illness knowledge and expertize are not afforded equal recognition in medical decision processes. 9,11,15,20,43 This is important to address since the flow of resources, such as power, through network ties is related to the level of actor control which is, in part, determined by their structural network position. 30 Clinicians, however, may not be aware of the power differential and hierarchical assumptions that exist in their patient/ clinician relationships, 10,20,24 and the potential to promote or hinder patient/client agency in treatment negotiations. 22 Finally, ego networks are influenced by organizational structures which may control communication styles. 30 We suggest that future research to support agentic patient participation in treatment decision making consider engaging a social network lens that can integrate insights from individual (ego), organizational (health-care team) and institutional (policy) levels to refine understanding.

| Limitations
The study findings are context-specific and may not be transferable to other IA populations with different characteristics or lower levels of patient activation. Only two men participated which may reflect a selection bias and/or disease pathology. Of note, most forms of inflammatory arthritis affect more women than men. For instance, more than three of five people with rheumatoid arthritis are women. 48,49 Research also indicates that women may be more likely to self-select for health research. 50  Future research could determine if there are differences related to age, gender or other sociodemographic characteristics. There was considerable variability in participants' ages yet our analysis did not identify any age-related differences. Additionally, our analysis did not address perceptions of gender treatment inequality around experiences of pain which has been identified in the literature. 38,46 Furthermore, the data did not identify temporal changes in participants' perceptions of influencing treatment decisions.

| CON CLUS ION
We explored the experiences of people with IA to gain a broader to data collection. The authors also express their appreciation to the participants who were willing to share their time and insights.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared.

AUTH O R S CO NTR I B UTI O N S
Wendy Hartford collected, analysed and interpreted data, and drafted all components of the manuscript. Dr Laura Nimmon conceived of and designed the larger study and contributed to analysis and interpretation of data for the current paper. Dr Laura Nimmon was also involved in editing the manuscript and revising it critically for important intellectual content. Dr Catherine Backman, Dr Linda C Li and Ms Annette McKinnon contributed to analysis and interpretation of data and revising the manuscript critically for important intellectual content. All authors have given final approval of the version to be published and agreed to be accountable for all aspects of the work.

E TH I C A L A PPROVA L
Ethics approval was obtained from the Behavioural Research Ethics Board (BREB), University of British Columbia (reference # H15-01751).

CO N S E NT TO PA RTI CI PATE
All participants provided signed consent prior to participating in the study.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.