Knowledge sharing to support long‐term condition self‐management—Patient and health‐care professional perspectives

Abstract Background Increased self‐management is a suggested solution to the burden on health‐care services of long‐term conditions (LTCs). This requires effective sharing of knowledge between health‐care professionals and patients, and is an underexplored area. Objective To understand how patients and health‐care professionals (HCPs) share and utilize knowledge in the social context of health‐care interactions within long‐term condition management. Methods Thematic analysis of 93 hours of observations of health‐care interactions and 33 semi‐structured interviews involving patients, carers and HCPs. Results 3 themes were identified: normative social roles, differing professional roles and the value of knowledge. Knowledge sharing was a complex process heavily influenced by social and cultural norms within the health‐care context. Not all knowledge was easily shared within routine health‐care interactions. Discussion The social context in which health‐care is practised influences what knowledge is shared and how this is achieved. It favours sharing of clinical knowledge from HCPs to patients and disadvantages patients in their ability to share their unique knowledge based on lived experience of illness. The opportunities for patients to be supported in their knowledge, skills and confidence within routine health‐care interactions are limited. Conclusion Both patients and HCPs need support to recognize the characteristics of the social context of health care and their understandings of their roles within this in order for them to move beyond accepted behaviours to develop more effective partnership working. Patient or Public Contribution Patients were involved in initial design of the study, particularly ethics of ethnographic observation.


| BACKG ROU N D
In the UK, effective management of long-term conditions (LTCs) is seen to be one of the greatest challenges facing health services. 1 Individuals' health-related behaviours strongly influence the prevalence and course of disease and illness, and the associated requirement for health-care interventions. 2,3 One of the proposed solutions for the challenge posed by increasing prevalence of LTCs is to support patients to better manage their conditions themselves, thereby reducing uptake of health-care services and progression of severity of disease. 2,4,5 This requires both patients and health-care professionals (HCPs) to share knowledge and utilize it in order for supportive self-management to take place. This paper seeks to explore how knowledge sharing takes place within the context of LTC management by: • Problematizing knowledge itself • Taking an alternative perspective to view how knowledge is utilized by social actors within health-care decision making • Illuminating the influence of the social context in which knowledge sharing takes place Self-management is broadly understood to encompass actions taken by individuals to recognize, treat and manage their own health and is associated with improved health outcomes. 6,7 Selfmanagement programmes aim to build individuals' knowledge, skills and confidence both in their condition and the management activities associated with it. [8][9][10][11] Self-management and the associated engagement of patients in their health care are associated with increased uptake and sustainability of healthy behaviours and improved clinical outcomes. [9][10][11] The assumption is that knowledge can be 'given' to patients; the main issue being ensuring that they then use this knowledge to inform their behaviours. 8,11,12 This assumption is contrary to much literature concerned with knowledge within health-care contexts which highlight the difficulties of knowledge sharing associated with organizational barriers, resource constraints, multi-professional working and the constant production of new knowledge. 13,14 There is also an assumption that knowledge sharing is primarily a unidirectional activity with HCPs 'transferring' knowledge to patients. This is in contrast not only to the theory associated with knowledge sharing 15,16 but also to personalized care models in which the lived experience of illness by patients is purported to be equally valued. 17 It has been suggested that an increased focus on self-management introduces a new paradigm centred around the patient-professional partnership 9 which necessarily involves multidirection knowledge sharing.
Shared decision making is implicit in the personalized care agenda. 17 It has been shown to have benefits for patients, carers and HCPs, but also challenges in its practical implementation. [18][19][20][21] The shared decision-making literature tends to focus on communication and process; the knowledge involved is often unconsidered.
By foregrounding and problematizing knowledge itself, its influence on sharing between social actors can be explored.
Health literacy is a related concept which has been widely discussed [22][23][24] and relates to the ability of patients to obtain, understand and use information in order to make appropriate health-care decisions. 25 Lower health literacy has been associated with poorer health outcomes and poorer utilization of health services. 26 This results in less appropriate use of health services overall-increased use of emergency services, reduced engagement with preventative care and reduced ability to self-manage routine health needs. 26 Rather than considering how patients are able to obtain and understand health-care information, this paper explores how easy it is for patients and HCPs to share this knowledge as social actors within interactions. This alternative lens through which to view health-care interactions allows elucidation of some of the cultural and social issues at play within the health-care context.
There are a range of social factors which can make knowledge sharing complex. 27,28 These factors relate to how society is organized, such as the social constructs of patient, nurse and doctor along with their associated activities and expectations. Social actors sharing knowledge come from different worlds, occupy different social roles associated with differential status and communicate in unique ways. [29][30][31] It is these social factors in relation to knowledge sharing which the current paper seeks to explore. This paper aimed to illuminate the social context and the cultural norms associated with it, in which knowledge sharing takes place. Cultural norms encompass the social behaviours accepted and perpetuated by society, and can be exemplified by the patient and HCP roles. 32 This context and accepted norms associated with it can be so embedded as to be imperceptible to the social actors operating within it, but nevertheless imperative to acknowledge in order to understand the setting in which health care is enacted.

| Design
A qualitative approach underpinned by ethnographic principles [33][34][35] was utilized to explore the behaviours associated with, and perceptions of, patients, carers and HCPs in relation to knowledge sharing. The main theme used within this sampling approach was knowledge and expertise. To self-manage, patients must have some level of knowledge and expertise in the practices required to manage their conditions. In addition, relationships with primary HCP were considered as it was conjectured that the character of these relationships may have an influence on how knowledge was shared as a socially situated activity. Carers were recruited to the study if they attended routine health-care appointments with a patient, and their participation was therefore driven by the patient participants. Sampling considerations and demographics are shown in Table 1. HCP participants were those who were involved in the care of the patient participants and who were observed interacting with them.

| Participants
HCP participant characteristics are shown in Table 2. Demographic data were not collected on HCP participants.

| Ethical considerations
All participants were recruited to the study in compliance with ICH-GCP requirements. This, and the process of ethical approval, ensured that processes were in place to protect participants from potential harm. In addition, all data were protected in line with General Data Protection Regulations and information governance requirements of both the academic sponsor of the study and the health-care environment were chosen as study sites. None of the participants were known to the researcher prior to their participation. Initial general observation was undertaken within public areas such as outpatient clinics or patient information sessions. The aim of this was to allow the researcher to familiarize themselves with the general environment, key actors and processes in order to determine how more focused observation should be undertaken, and with whom. The researcher was identifiable and information regarding the conduct of a research study was provided to individuals located in the area.
During general observation, the researcher conversed with patients who were in the environment and if appropriate discussed further participation in the research study which would involve observation of confidential consultations and interviews. Written information was provided to potential participants at this stage. Other potential participants were approached by their usual care provider and provided with written study information. Written informed consent was obtained prior to any further observation or interview. Four potential participants who were approached declined to participate.

| Data collection
Data collection involved observations and ethnographic interviews. In addition, 33 interviews were undertaken with patients, carers and HCPs. All patients and carers were interviewed save for one patient who became too unwell, and selected HCPs were interviewed based on how much contribution they had made to the observational data as shown in Table 2. The interviews were semi-structured, and whilst topics for interviews were common to both patients and HCPs, the questions asked varied between the two groups. A topic guide was developed by the researcher who also conducted the interviews. Examples of questions are shown in Table 3. Interviews were participant-led, so the topic guide was used as an indicative tool, but the course of the interview was contingent on the participant.   taken to minimize influence however as the researcher was not in uniform at any time, and data collected in organizations in which they were unknown in their professional role. Ultimately, it was felt that the position of the researcher enriched the data rather than restricted it, but there is no doubt that it had an influence.

| Data analysis
Data were managed using NVivo software. Thematic analysis took place concurrently with data collection, and initial analysis informed subsequent data collection using the constant comparison method. 36 This is commensurate with the ethnographic approach where analysis is an ongoing iterative process with constant movement between the field, generated data and emerging ideas. 34 The units of analysis were the two groups of patients and health-care professionals. Analysis was inductive, with open coding generating numerous categories which then was followed by focused coding involving synthesis of initial codes into categories. 37 From 159 initial codes, 13 broad categories were formulated. A more abstract review of the data was then undertaken, considering the larger narrative and its relation to current theoretical understanding of knowledge sharing. 38 The constant comparative method ensured that existing data were constantly compared with new data, resulting in categories being elaborated or integrated within the process of analysis. 36 The initial analysis and coding was undertaken by one researcher, but then reviewed with two other researchers who were part of the study team. Broader themes and subsequent findings were discussed with all three researchers.
Data saturation was not an aim within this study; rather an attempt was made to gain a breadth and depth of experience and perspective.
The limits to data collection were mainly pragmatic in terms of the number of participants a single researcher could adequately follow simultaneously, and the need for data collection to be concluded in the timeframe required. Utilizing the constant comparison method did show, however, that there were increasingly fewer amendments to existing codes as a result of new data generated and that data could be explained using existing codes and therefore data collection was discontinued.

| RE SULTS
Reported here are three main themes emergent from the data which illuminated how knowledge sharing was achieved between patients and HCPs within routine interactions.
• Normative social roles which focus on how social actors' understandings of their identity and what is expected of them influences how they behave within interactions.
• How differing professional roles within the health-care sphere influence what knowledge is shared.
• Different types of knowledge prevalent in health care and the value assigned to these by social actors. Just sort of assume that they know nothing. SL007

(Renal doctor) Interview
The patient role was also associated with a responsibility for knowledge sharing, but this was primarily to allow HCPs to make diagnosis and treatment decisions.
I think I tell them everything…it's always better to tell them as much as you know, to help them diagnose you or treat you.

PN009 (Gastro patient) Interview
It was apparent that part of the patient role was not only to share knowledge, but also to follow the advice of the HCP. This was not articulated, but was apparent in observed behaviours.   comprehensive knowledge sharing to take place however, particularly from patients to the HCP team.

PN002 (Gastro Patient) Observation
The patient tries to contribute her experiential knowledge of how she feels and this is ignored in favour of a clinically focused treatment plan. In fact, the knowledge which she contributes is not

| Discussion
Despite being a knowledge-rich environment, knowledge sharing is complex within health care and there are difficulties associated with it. 13,14 This paper has highlighted the influence of social and cultural aspects of the health-care environment in this process. Asymmetry within doctor/patient interactions has been recognized previously, 39 but this has often focused on communication skills and processes. 40 One of the advantages of foregrounding knowledge itself is that its influence on how it is shared has been elucidated. It is clear that there are different types of knowledge, these types of knowledge are assigned differential value, and specific types of knowledge are associated with particular social roles. This is exemplified by higher value biomedical knowledge associated with HCPs in comparison with less valued experiential knowledge which patients contribute. The knowledge itself is therefore related to, and contributes to, the normative social roles of HCP and patient.
Furthermore, violating these social norms in relation to knowledge can be problematic in health-care interactions.
A deficit model of knowledge is common in health care which assumes that patients have insufficient knowledge in order to selfmanage. 41 The goal of health care is therefore to furnish patients with the knowledge which they require to manage their conditions.
In contrast, this study would suggest that it is not necessarily a lack of knowledge which is the issue, but rather an undervaluing of the unique experiential knowledge which patients have and a corresponding lack of opportunity to effectively utilize it within the decision-making process.
Understanding the challenges of knowledge sharing is crucial to support multi-directional knowledge sharing between patients and HCPs. This ensures that not only do patients have access to knowledge which allows them to self-manage, but that they are able to feedback the success or otherwise of a programme of selfmanagement based on experiential knowledge. Only then can a personalized plan of care be developed which allows for individuals' health to be effectively managed within their unique set of circumstances. 17 Most discussions regarding self-management focus almost exclusively on 'increasing' the knowledge which patients have. 8 Little consideration is given to whether patients feel able to utilize this knowledge within the routine interactions concerned with health-care management and planning. Moreover, it has been highlighted that knowledge and skills, and the confidence to utilize them, decline over time unless supported in the longer term. 36 Developing knowledge and skills and supporting in routine interactions the confidence to utilize them would be a cost-effective approach, thereby removing the need for additional ongoing support mechanisms for patients. Whilst recent policy appears to endorse the value of patients' unique knowledge gained from lived experience, this is not always reflected in practice. 17 Embedded practices of both HCPs and patients mean that the accepted value of different types of knowledge is often unchallenged thereby making it more difficult for sharing of patients' experiential knowledge due to its perceived lower value.

| LI M ITATI O N S
This study has some limitations in its design. Comprehensive demographics including educational level, employment status and duration of illness were not collected on participants. As these characteristics can have a bearing on individuals' ability to participate in health-care interactions, this information would have allowed additional elements of analysis to have been undertaken.
This does not influence the findings in as much as the themes identified were common to all participants regardless of these or any other attributes, but the extent to which particular characteristics influenced practices was not able to be explored. In addition, the data were collected by a single researcher who had familiarity with the health-care setting. Whilst this may have had advantages and was accounted for in the study design and data analysis, a larger study with more opportunity for inter-researcher discussion may have strengthened the study.

| Potential for further study
There are further avenues of research suggested by these results, one being to explore whether the issues identified are the same in primary care or general practice where HCPs are less specialist in particular areas of health-care practice. Power dynamics within the relationships between HCP and patient may be different in this context for a variety of reasons, as well as the value of knowledge being shared within these interactions. In addition, there may be some value in exploring whether carers mediate knowledge sharing within health care. This was not possible in this study as there was only one participant who had carers present at their consultations, but exploring this further may be an interesting avenue of further study.

| CON CLUS ION
Knowledge sharing is a socially situated activity. Understanding the social context of knowledge sharing is crucial, as is understanding the roles and perspectives of the actors involved. As we try to encourage patients to take on more self-management activities, it is essential to understand how this relates to their current understandings of their roles within health care and their relationships with other actors. It is not enough for policy to endorse the value of patients' unique knowledge unless there is support in practice to make this a reality.
Accepted social roles of both HCPs and patients are part of the culture of health care and exceptionally challenging to alter.
Recognition of these roles-their benefits and challenges-along with an acknowledgement of practices which perpetuate them is the first step in addressing some of the issues which they present.
Changes to practice can contribute to alteration of culture incrementally over time. This requires work with both HCPs and patients simultaneously as changes to the practices of one group will have little impact if met by a lack of understanding of on the part of other social actors.
The tension for patients in self-management is that within the health-care context of the hospital, their unique knowledge is undervalued, but once out of that environment, they rely on this to manage their health conditions. This is a paradox which requires addressing if patients are to feel empowered to manage their conditions based not only on knowledge and advice from HCPs, but also their own experiential knowledge. This could, and should, be reinforced through interactions in routine health-care consultations.

ACK N OWLED G EM ENTS
This study/project is funded by/supported by the National Institute

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.

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.