The potential and pitfalls of narrative elicitation in person‐centred care

Abstract Background Revitalized interest in narrative has informed some recent models of patient and person‐centred care. Yet, scarce attention has been paid to how narrative elicitation is actually used in person‐centred care practice and in which ways it is incorporated into clinical routine. Aim We aimed to identify facilitators and barriers for narrative elicitation and setting goals in a particular example of person‐centred care practice (University of Gothenburg Centre for Person‐centred Care, GPCC) where narrative elicitation is considered as a method of setting goals for the patient. Methods Observation of 14 admission interviews including narrative elicitation on an internal medicine ward in Sweden where person‐centred care was implemented. Five focus group vignette‐based interviews with nurses (n = 53) were conducted to assess confirmation of the emerging themes. Results The inductive analysis resulted in three themes about the strategies to elicit patients' narratives: (a) Preparing for narrative elicitation, (b) Lingering in the patient's narrative, and (c) Co‐creating, that is, the practitioner's and third parties' engagement in the patient's narration. Even though there were obstacles to eliciting narratives and setting lifeworld goals in a medical setting, narrative elicitation was often useful to turn general and medical goals into more specific and personal goals. Conclusions Narrative elicitation is neither a simple transition from traditional medical history taking nor a type of structured interview. It entails skills and strategies to be practiced. On the one hand, it revitalizes ethical considerations about clinical relationship building. On the other hand, it can help patients articulate lifeworld goals that are meaningful and important for themselves.


| INTRODUC TI ON
Revitalized interest in narrative in health care has envisaged a paradigmatic change that would reinvigorate the art of medicine 1,2 and encourage health-care professionals to think beyond the biomedical realm. Narrative-based approaches aim to embrace idiosyncratic experiences of illness, 3,4 enhance clinical dialogue 5,6 and empower people with medical conditions vis-à-vis the dominance of biomedical knowledge and language. [3][4][5][6][7][8][9][10][11][12] Some recent models of patient and person-centred care have considered narrative as a way to attend to the person behind the patient. [12][13][14] In other words, the patient is more than a diagnosis and a passive recipient of health care. Alongside the efforts to improve health-care professionals' narrative skills and competence, 15 some recent frameworks of person-centred care (hereafter PCC) 13,14,16,17 have therefore promulgated narrative elicitation as a means to gain better understanding of the person behind the patient.
There is a wide range of definitions and practices claiming to be person-centred. [18][19][20] It is mainly construed as an overall change in health-care organization and ethics, with less consideration for practice implications. 21 The selected case in this study (University of Gothenburg Centre for Person-Centred Care, GPCC) combines an ethical approach which acknowledges the capacities of the person and three routines guiding health professionals, that is, eliciting the patient's narrative, partnership and documentation. It is a particular example that interweaves a narrative-based approach with person-centredness. It encourages health professionals to attend to the patient as a person with capabilities, resources and a narrative to relate. 13,14,17 Narrative elicitation consists of asking questions beyond the diagnostic workup, guiding the person to grasp and relate their wishes and capabilities, and probing their accounts into shape so that the patients set their own goals. It is suggested as a method for health professionals to acknowledge patients' experiences of illness and give patients space to bring their resources and goals. The underlying understanding is that narrative elicitation leads these goals to be less biomedical and technical, yet more meaningful for the persons and their lifeworld, that is, the ways in which they perceive and make sense of their illness in the context of their everyday lives. 22,23 GPCC's emphasis on narrative elicitation as a new 'routine' has aimed to interrupt the dominance of biomedical language 24 and empower patients as important actors in health-care delivery.
Yet, narrative elicitation requires organizational, technical and attitudinal changes. 6,11,12 Narrative elicitation is more dynamic and unpredictable than structured ways of taking a history. There are no strict guidelines that always work, hence professionals need to develop individual and collective strategies to perform it. How narrative elicitation is used in person-centred care practice and how it can reshape clinical routines should be further observed, examined and documented to elucidate facilitators and barriers in the process. We also need to trace whether and in which ways narrative elicitation helps patients to bring their capabilities and set goals in clinical communication. It is against this backdrop that we propose close examination of narrative elicitation as it unfolds in one ethically driven and evidence-based practice of PCC.
The aim of this study is to identify strategies and barriers in narrative elicitation and to discuss the relationship between narrative elicitation and setting lifeworld goals. We will argue that narrative elicitation is not always straightforward in practice, and thus entails strategies to overcome the practical and professional challenges.

| Research design
Our previous study about the implementation of the GPCC has shown that translating PCC into practice requires contextual, sometimes contested, and often creative adaptation. 16

| Data collection
The observation study took 13 days of ethnographic fieldwork 29 over 3 months in 2017. Two ward managers and two RNs experienced in PCC facilitated access to the ward for ÖN. ÖN, trained in ethnographic methods, was on site during one whole shift each day of fieldwork. Each day of fieldwork, ÖN participated in the staff meeting before the start of the shift, and informed people about the study and his presence. As all RNs were expected to conduct admission interviews, they were first asked if they agreed to be observed by the researcher. Then, when a new patient was admitted, ÖN approached each RN in charge of the interview if he could observe it.
All participating RNs consented to be observed. Then both ÖN and the RN informed the patient and asked for consent for ÖN's presence and observation during the interview. Two patients did not want to participate in the study and the researcher did not observe these admission interviews. The study was approved by the Regional Ethical Board of Gothenburg, Sweden. All participant nurses and patients were provided written and verbal information about the study and gave their consent to participate. ÖN observed 14 admission interviews. He also observed the preparation for two admission interviews without participating in the interviews themselves because one was an emergency case and patient consent could not be requested, and the second was an infection case where the patient was isolated. He wrote down notes during observation periods and elaborated these notes after the observation. He had short conversations about the admission interview with the nurse and the patient afterwards when this was possible.
Vignettes were developed to stimulate focus group discussions and to refine the findings of the observational study. The RNs were eager to discuss his observations with the researcher. ÖN wrote eight vignettes illustrating some challenges and facilitators to narrative elicitation; these were hypothetical but inspired by his observations. The research team and two RNs commented and refined the vignettes, which AW checked for medical accuracy. These were used for five focus group interviews with registered and assistant nurses (n = 53), lasting 75 minutes on average. Three vignettes were selected and introduced in each group by ÖN and AW. Nurses were asked what they thought was successful in terms of narrative elicitation in the vignette and what they would do if they were in the place of the nurse. These focus-group discussions were audio-recorded and transcribed. As these focus group discussions took place during staff meetings, some nurses had to leave in the middle of the discussion for other duties.

| Data analysis
The research design led to two different sets of data. The first set consisted of the observation notes 29 of admission interviews. These included the minutiae of admission interviews, the details of nurses' attitudes, formulation of questions, turn-taking in conversation as well as patients' reactions, how and what type of goals were recorded, the length of the interview, and challenges during the interaction. The notes for each case were written down by ÖN, and circulated to the other team members. Team members asked ÖN questions to clarify certain details and to interrogate his interpretations. Then, ÖN coded and analysed thematically the field notes.
Three themes emerged from inductive analysis. 30 Subsequently, patients' documented goals were categorized depending on their content and nature. The research team met regularly face-to-face and online to discuss ÖN's ongoing analysis.
The second set of data was generated in the focus group discussions. The emerging themes from analysis informed the preparation of vignettes, where the researchers incorporated the identified strategies into scenarios. The use of vignettes aimed both to integrate multiple methods and to validate the themes through informant feedback. This feedback is a pivotal triangulation technique in qualitative research 31 and was used to minimize the single observer's biases. For instance, the researcher was not a health professional himself and observed the professionals' use of personal information and included this theme in one scenario. This theme was elaborated after the participants' feedback.
The two sets of data were used in data analysis. For the themes for which informant feedback was positive, the analysis of field notes was used. For the presentation of contested themes, data from the focus groups were included to offset the researcher's subjectivity and to give voice to informants' objections.

| RE SULTS
In this section, first, we will present three main strategies for eliciting the patient's narrative. The first theme was preparing the interview, that is, what nurses do before the admission interview. The second concerns what happens during the admission interview to linger in the patient's narrative. The third was about the co-creation of narrative via joint interviews or self-disclosure Table 1. Second, we will elaborate on the nature of goals in relation to strategies Table 2.

| Preparing for narrative elicitation
Narrative elicitation is a difficult task on the specialized wards where patients are generally admitted for specific and relatively short-term interventions. The turnover of patients is also high. Traditionally, it [Field note, case 9]

| Lingering in the patient's narrative
Active listening is about skilfully navigating conversation and silence without interruptions during face-to-face communication. 12,32 Narrative elicitation required nurses to be more attentive to the interaction and to provide patients with the time and attention they needed. Taking a seat, having a calm posture and keeping eye contact were some common strategies. One way to linger in the patient's narrative was asking questions exploring the same topic further with follow-up questions. 11 This included repeating some details of the story, asking follow-up questions if necessary, alluding to significant events in the narrative (such as a recent heart attack or loss of appetite), helping patients connect these to potential wishes, plans and goals (such as attending a dinner, travelling, starting or commencing a new hobby).

| Co-creating narrative
There were two ways in which narratives were created in dialogue: firstly through self-disclosure and secondly through joint interviews, that is, interviews including third parties as facilitators or information resources in communication. 17 (N9, Focus group 1) As some patients had long-standing illnesses, they were also used to talking and asking questions in medical language. This posed an obstacle to address other concerns and talk about lifeworld goals.
Secondly, the relationship between narrative elicitation and setting lifeworld goals was not self-evident for many nurses. In some cases, patients were happy and eager to tell their stories, yet there was need for interpretation of these accounts to articulate a life-

| Generic vs specific goals
The patients' goals in this study can be classified into two groups.
The first group consisted of general goals like 'being healthy' (cases 6 and 9) or specific medical goals like 'getting help for breathlessness' (case 13 Is it still important for you?' (Case 11)). These specific questions led to more specific lifeworld goals such as 'restart his model ship building'.
As the Table 2  Patients are often allocated short consultations and health-care professionals feel more and more pressure to be time effective and pose questions addressing medical issues. 6,8,11 Asking questions beyond the medical realm not only requires time and well-organized division of labour, but also skills of formulating narrative-inducing questions and active listening 32 as the patient's narrative during the admission interview emerges 'in the context of requests, acknowledgements, expansions, and elaborations'. 36,p. 368 Some nurses managed to elicit narratives by drawing upon their professional and personal experience but training programmes could endorse and sustain these skills. 27 Narrative elicitation has revitalized some ethical dilemmas. Selfdisclosure 34  The ambivalence about joint interviews is another issue. This can hark back to some criticisms to the narrative turn that was considered to emphasize 'the isolated actor who experiences and narrates as a matter of private and privileged experience'. 38 While for some family members or close friends were considered helpful partners in elicitation, others considered them harmful to the dialogue between the nurse and the person. This is an understandable concern in the case of family members who take control of the conversation and attempt to impose their own agendas. Having joint interviews in PCC appears somehow contradictory as if one's narrative would not be genuine if third parties were involved in narrative elicitation, but models of PCC generally need to have a more relational vision of personhood and acknowledge that the uniqueness of the person is always shaped and expressed through a web of relations. 17 Given the potential contribution of third parties in both narrative elicitation and goal setting in some observed cases, it is not a clear-cut question. One way to acknowledge potential benefits but also harms of the third parties in narrative elicitation can be to go beyond this individualistic and dyadic understanding of narrative elicitation 17 by taking the person's web of relations critically and informatively into consideration.
This study also calls into question the relationship between eliciting narratives and setting lifeworld goals. Some patients persevere in expressing goals like 'being healthy'. This points to some challenges that narrative-based approaches face: many patients are not familiar with narrative elicitation in medical settings and may even prefer to focus on medical issues.
Many people have long been inclined to focus on their medical conditions during medical history taking, since their resources and capabilities are rarely taken into account in clinical communication, not only in admission interviews. Narrative elicitation is therefore a way to open this space for acknowledging their resources and capabilities. While 'being healthy' is a more generic goal, often renegotiated by people with chronic illnesses, 'being able to do gardening' is more precise and arguably more motivating to take part in care planning. To decide whether these specific goals are genuine and attainable 16  This study focused on observations of narrative elicitation on a specific ward. The themes generated by the situated observation of the researcher were triangulated with continuous feedback from the nurses and focus group interviews. Different settings may present other strategies and realities depending on the context. It is also difficult to point to a particular set of strategies that always work. However, it is possible to highlight some common patterns in eliciting narratives.
As is the case with all observation studies, research participants might have paid more attention to what they did and how in the presence of a researcher. They might have attempted to demonstrate best practice, but this was equally valuable for the aim of this study.

ACK N OWLED G EM ENTS
The authors are grateful to the ward managers and nurses who gave access to the ward, and the nurses and patients for their willingness to participate in the study.

CO N FLI C T O F I NTE R E S T
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.