Interactional practices in person‐centred care: Conversation analysis of nurse‐patient disagreement during self‐management support

Abstract Background Person‐centred care implies a change in interaction between care professionals and patients where patients are not passive recipients but co‐producers of care. The interactional practices of person‐centred care remain largely unexplored. Objective This study focuses on the analysis of disagreements, which are described as an important part in the co‐production of knowledge in interaction. Design A qualitative exploratory study using conversation analysis. Setting and participants Data were collected from a nurse‐led person‐centred intervention in a hospital outpatient setting. Interactions between adult patients with irritable bowel syndrome (n = 17) and a registered nurse were audio‐recorded. COREQ guidelines were applied. Results Disagreements were found after demonstration of the nurse's or patients’ respective professional or personal knowledge. Disagreements were also evident when deciding on strategies for self‐management. Although negotiations between opposing views of the nurse and patient were seen as important, the patient generally claimed final authority both in knowing how IBS is perceived and in the right to choose self‐management strategies. The nurse generally oriented towards patient authority, but instances of demonstration of nurse authority despite patient resistance were also found. Discussion and conclusions This study provides information on how co‐production of knowledge and decisions occur in the context of a person‐centred care intervention. Negotiations between nurse and patient views require a flexible approach to communication, adapting interaction to each context while bearing in mind the patients having the final authority. To facilitate co‐production, the patient's role and responsibilities in interaction should be explicitly stated.


| BACKG ROU N D
Person-centred care (PCC), compared to usual care, has been associated with shorter hospital stays, 1 improved self-efficacy 2 and quality of life. 3 In PCC, patients serve as experts of their own health and unique living conditions, and health-care professionals (HCP) represent generic knowledge on health, illness and care. 4,5 When power and responsibility are shared, patients become part of the health-care team. 6 Although PCC's efficacy has been proved in various settings, the actual interactional practices involved need more study. This paper addresses this knowledge gap by analysing interactional data involving a registered nurse (RN) and patients with irritable bowel syndrome (IBS). The data were collected in a pilot PCC intervention for IBS that has shown potential to decrease symptom severity. 7 Since no curative medical treatment exists and IBS is long-term or chronic, the intervention focuses on supporting patients in their self-management regarding diet, stress and physical activity. The intervention is based on the principles of PCC -patient narrative, partnership and documentation -as described by Ekman et al 8

| Communication in person-centred care
Health-care services should prioritize supporting self-management for patients with chronic or long-term disorders and co-production of care and health is vital to that endeavour. 10 Although different definitions of PCC exist, all represent a shift from paternalistic to egalitarian relationships between HCP and patients -that is, a partnership in which health and care is co-produced. [4][5][6]11 This partnership is created when HCP are engaged listeners, non-judgemental and respect their patients' personal wishes and needs. 6,12 However, HCP tend to overestimate the extent care is adapted to each person and care is often task-oriented rather than person-centred. 13,14 Because the scarce literature on person-centred communication focuses mainly on dementia care (i.e. involving people with speech-language disorders), 15 its findings might not be transferable to other contexts and patient groups. However, more literature has focused on the related concepts of patient-centred communication and shared decision making (SDM), which involve patients in treatment decisions. Patient-centred communication and SDM can improve patient satisfaction 16,17 and adherence to treatment. 18 The so-called 'high participation patients' (i.e. those who express opinions) force their physicians to use a patient-centred communication style. 19 However, Kunneman et al 20 report that SDM is often used in an instrumental way, ignoring humanistic aspects of interaction.

| Conversation analysis and PCC
A valuable method to investigate how people structure their interaction is Conversation Analysis (CA). 23 CA, rooted in sociology and ethnomethodology, was developed in the early 1960s by Harvey Sacks and collaborators. 24 CA, which is data and participant driven, relies on sequential context. 25,26 That is, the analysis takes a bottomup approach that examines how participants handle utterances in specific interactional contexts.
Related to co-production in PCC, CA research explores disagreement between speakers. Disagreement has been discussed largely in relation to Pomerantz's 27 observation that in a friendly conversation agreement is the preferred response due to the preference for interactional consensus. 27 Therefore, disagreement could generally be seen as dispreferred, even if in some sequential positions (e.g. speaker self-deprecation) disagreement is the preferred response.
Nevertheless, recent research argues that disagreement cannot be seen as positive or negative (or preferred or dispreferred), but as a natural part of interaction when expression opposing views. 28 Disagreement can entail intimacy and companionship between speakers. 29,30 In business meetings, disagreement can bring together opposing viewpoints. 31 Hence, disagreement is more expected and appreciated in some contexts, which affects the form of disagreements. 28 Pomerantz describes disagreement as a delay combined with an initial partial agreement. However, Kotthoff 32 shows that if disagreement escalates to heated dispute or conflict, these mitigating signals gradually disappear.
Another aspect to consider is how disagreements are enacted in interaction. Opposing views or disagreements have been discussed with respect to the epistemic and deontic status of participants. [33][34][35] The epistemic status of a speaker is a condition of the person's right to knowledge within a specific domain, 36 whereas the deontic status is a condition of the participant's right to decide how something should proceed. 37 Superior status or authority implies both that someone claims authority and, perhaps more importantly, someone accepts that authority. 38 Epistemic and deontic status is managed, for example, through the use of modality in interaction. 39 In PCC, the interaction generally strives for shared status between the professional and patient as both have the right to express their respective knowledge as well as to collaborate as partners when making decisions. 33 However, interactional equality is difficult and the professional domain is often given authority. 33,35 Using CA, this study builds on the few studies that unpack interactional practices used in PCC. The study attempts to increase the knowledge of how PCC is managed in clinical encounters. As previous research has discussed the importance of disagreement in co-production, this study's analytical focus is on the sequential positioning and enactment of disagreement through epistemic and deontic status.

| ME THODS
Data were collected in a pilot project that evaluated a personcentred intervention for patients with IBS at a hospital outpatient setting specialized in functional bowel disorders. 7 The intervention consisted of 4 parts: individual support sessions with an RN (second author, IB) two to 4 times every second week, with additional contact by phone and/or e-mail; health diaries; written information; and patient-held medical records. The RN was trained in the ethics and principles for PCC but has no specific training in communication apart from what is included in a general nursing degree. The patients were referred to the clinic by their general practitioner or by selfreferral and were on a waiting list for a group education programme.
Of the 105 patients on the waiting list, 36 were purposefully sampled to obtain a variation regarding age and gender. Patients were sent a letter that asked them to participate in the pilot study rather than the educational programme. Of these, 20 agreed, but two were found ineligible because of serious psychiatric disorders or insufficient language skills and one did not show up for the first session.
Therefore, 17 patients were included in the study. 2 patients had mild IBS, 9 moderate and 6 severe. 2 participants were retired, 2 on full-time sick leave, 2 unemployed, 2 students and 9 employed.
All were Swedish citizens, and 2 were children of immigrants. The intervention was evaluated using interviews and questionnaires. 7 The present analysis covered 27 interactions (audio-recorded) between the RN and the 17 patients (4 male and 13 female; mean age 39 years).
The audio-recordings were transcribed verbatim with sequences of interest transcribed using established CA guidelines (Appendix 1). 40 The initial viewing of the dataset revealed an overarching pattern. First, the patients, guided by the RN's questions, described how IBS affected their life. Second, in the following 1 or 2 encounters, the patients focused on self-management strategies; for some patients, these discussions resulted in a collaboratively compiled health plan.
After the initial viewing, the analysis was focused on disagreement. We used Sifanous 41 definition of disagreement: 'the expression of a view that differs from that expressed by another speaker'.
More specifically, we focused our analysis on sequences demonstrating verbal disagreements. 27 The analysis more specifically contained selecting target instances of disagreement. In total, 52 target instances, involving 12 participants, were identified. Each selected target instance consisted of an extract encompassing an utterance demonstrating disagreement and its preceding and following turns related to that disagreement. We conducted a turn-by-turn analysis of both how participants displayed and handled disagreement in sequences. Hence, the analysis included both an exploration of the situations in which disagreements appeared as well as how disagreements were negotiated in the following turns.
Collections of disagreements were then made, organized around the sequential positioning, that is the turn upon which disagreement was the response. These collections were further sorted into 2 overarching areas connected to the negotiation of disagreements in terms of demonstration of epistemic/deontic authority. The data were analysed separately by the 2 authors as well as collaboratively during monthly data sessions. To increase the reliability of the analysis, three separate data sessions were held with researchers and 2 graduate students familiar with CA. 24 The project was approved by the regional ethics review board in Gothenburg (application no. 434-15). All participants gave written informed consent to participate. The Consolidated criteria for reporting qualitative research (COREQ) was applied (see Data S1). 42

| Portrayal of knowledge about IBS symptoms
Both the RN and patients initiated sequences of disagreement after demonstrating their knowledge about IBS symptoms. Patients initiated disagreement following the RN's explanation of generic information about symptoms or models, which did not correspond with the personal experience. The RN initiated disagreements when patients were inclined to place the generic or professional explanation above their own perception.
In Extract 1, the RN and the patient (P17) discuss why the patient experiences gastric gas (Table 1).
In a display of epistemic status, the RN tells the patient that IBS symptoms occur because the patient has problems passing gas (lines 1-11). After listening to the RN's description with minimal responses, potentially signalling disagreement, 27 the patient responds by saying 'that's not how I experience it' (line 14). The patient uses the primary tense (i.e. tenses that express present and future time) to reflect the high truth value (or being the reality) of the statement, a strategy that claims epistemic authority over the RN. 39 However, the patient also signals the delicacy of the situation by laughing 43 and quickly modifies his view by stating she might be right even though this is not how he experiences IBS. The low modal operator 'might' downgrades the truth value of his previous statement. 39 His simultaneously agrees and disagrees, resulting in weak disagreement. 27 Then, the patient posits that the gas could have been collected for a long time, which the RN is quick to accept (line 18). The RN then modifies the patient's explanation to encompass both of their accounts and the patient signals agreement (lines 18-26).
In Extract 2, the RN and the patient (P8) are talking about the patient's experiences with stress. The patient notes that several HCP have asked her about her stress ( Table 2).
In line 1, the patient states that she has felt shocked when HCP asked if she experiences stress, as this proposition does not align with her personal experience (or epistemic domain). The patient orients partly towards this professional view (lines 3-6) when she suggests that 'perhaps' she has some inner stress. However, the use of the low modal adjunct 'perhaps' simultaneously downgrades the truth value of this professional account. In the following turn, the RN expresses disagreement, which, as in Extract 1, is vague. 27 At first, the RN says 'yes', but she then states that 'or it might not be', revealing that the RN wants the patient to trust her own personal knowledge (i.e. physical and emotional experience). The RN goes on to clarify that the patient's symptoms could be connected to something

| Advice on strategies for self-management
The data reveal that patients often disagreed with the RN's advice about self-management. In Extract 4, the RN and a male patient (P6) discuss the use of antidepressants as a strategy for self-management ( in 'if you could try' (line 5). By constructing her proposition this way, the RN signals a contingency -that is, the patient can either accept or reject the suggestion. 45 Hence, this request can be seen as a way to share the deontic right with the patient. 38 The patient replies that she feels a direct connection between her IBS and pasta -in fact, she frames her experience using a high modal adjunct 'always'. 39  P2: yeah but yeah now that you mention it I haven't even reflected over that before but e:: (.) yeah response suggests she agrees her work conditions should be adapted (lines [16][17][18][19][20][21]. In the last line (line 24), she states this is the first time she has considered that her job should be adapted to her situation. In this interaction, the participants find a possible strategy that considers both the professional domain (stress should be reduced) and the personal domain (work conditions are difficult to change).
Extract 7 (Table 7)  That is, the RN claims the deontic authority using her physical positioning in front of the computer and simultaneously writing the health plan. The patient clarifies that her exercise is yoga and uses a primary tense to invoke her epistemic and deontic status as she is the one engaged in the exercise (lines [10][11][12][13][14]. 39 This linguistic move resists the RN's deontic right to determine the amount of training.

| D ISCUSS I ON
In the context of a person-centred self-management intervention, As with previous research, we found that disagreement produced a back and forth movement between participants, demonstrating their knowledge while inviting the other person's account. 28 These negotiations produced new and specific knowledge or resulted in decisions. In this process, the RN often established a weaker epistemic stance or sharing of status and the patients claimed their authority via their personal experience. The patients not only claimed epistemic and deontic authority but also accepted the RN as an authority and invited a shared status.

TA B L E 7 Extract 7
However, our data also demonstrate that the RN invoked her deontic right to decide what should be written in a patient's health plan by establishing her professional experience and knowledge as well as by using her physical positioning and power over documentation. This particular extract, however, can be seen as a breach of PCC as the RN ignores the patient's resistance and avoids shared status. This extract also reflects the fact that even if the patient had been given the possibility to portray her epistemic domain, the professional can 'stumble on the final step' -that is, produce the final version of the health plan. In PCC, a health plan should be written by the professional and patient together and either can take initiative in the actual writing. 2

| CON CLUS I ON S AND IMPLI C ATI ON S
This study provides important information on how co-production is constructed in the context of a nurse-led self-management intervention, but it does not formulate a general method or technique for PCC. PCC considers the context and the persons participating in an interaction. 22 We encourage HCP to 'dance with their patients', which necessitates listening to the music and not simply following supporting patients decision making. 21 Our study confirms that HCP need to challenge their patients' views while acknowledging that their patients have the ultimate epistemic and deontic authority.
Although our results stress interactional flexibility, some general implications can be derived. To provide opportunity for coproduction, professionals should explicitly tell their patients that their knowledge is valued and that professional knowledge should be seen as complementary (i.e. make sure that patients are aware of their rights as well as responsibilities). In addition, neither the professional nor the patient should be afraid of disagreeing with the other's account as it is in this back and forth movement that new and specific knowledge is produced. If the patients do not disagree with anything that the professional says, questions should be asked regarding how proposed strategies could work.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

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 from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.