Nurses’ application of the components of family nursing conversations in home health care: a qualitative content analysis

Aim The purpose of this study was to describe how nurses apply the components of family nursing conversations in their home healthcare practice. Method A qualitative content analysis with a deductive approach was conducted. Home healthcare nurses conducted family nursing conversations with families from their practice. Families were selected based on three nursing diagnoses: risk of caregiver role strain, caregiver role strain or interrupted family processes. Nurses audio‐recorded each conversation and completed a written reflection form afterwards. Transcripts of the audio‐recorded conversations were analysed in Atlas.ti 8.0 to come to descriptions of how nurses applied each component. Nurses’ reflections on their application were integrated in the descriptions. Results A total of 17 conversations were audio‐recorded. The application of each component was described as well as nurses’ reflections on their application. Nurses altered or omitted components due to their clinical judgment of families’ needs in specific situations, due to needs for adjustment of components in the transfer from theory to practice or due to limited skill or self‐confidence. Conclusion All of the components were applied in a cohesive manner. Nurses’ application of the components demonstrates that clinical judgment is important in applying them. Further training or experience may be required to optimise nurses’ skill and self‐confidence in applying the components. This study demonstrates the applicability of the family nursing conversations components in home health care, allowing exploration of the working mechanisms and benefits of family nursing conversations for families involved in long‐term caregiving in future studies.


INTRODUCTION
Illness of a family member impacts the entire family. Family relationships, roles, and tasks as well as activities and communication may change (Årestedt et al., 2014). In situations of severe stress such as illness, resilient families will find effective ways for positive adaptation whereas the confrontation with an illness could result in crisis for those that are less resilient (Walsh, 2003(Walsh, , 2012. Caring for a family member might bring about positive experiences (Yu et al., 2018) and have favorable effects on the health and wellbeing of family members (Brown & Brown, 2014;Roth et al., 2015). However, those family members that provide intensive care are especially at risk for caregiver burden with negative consequences for their health and work participation (Colombo et al., 2011;Heger, 2017;Luttik et al., 2007;Naef et al., 2017;Schmitz & Westphal, 2017) as well as for the quality of the care they provide (Beach & Schulz, 2017). Support from professional caregivers seems to be important for preventing or decreasing family caregiver burden (Van Houtven et al., 2011). A recent integrative review found that, in addition to such supportive care needs, family caregivers also consider collaboration with home care nurses important in caring for the patient (Ris et al., 2018;Wittenberg et al., 2018).
The theory of Family Systems Nursing emphasizes that nurses should approach families rather than only the patients as the unit of care as families are always impacted by illness (Wright & Leahey, 2013). An important intervention within Family Systems Nursing is the Family Health Conversation that was developed in Sweden (Benzein et al., 2008). These conversations have been described in terms of 12 well-defined core components (Östlund et al., 2015). Such a clear and specific description of an intervention is beneficial for educating professionals and is likely to increase intervention integrity (Carroll et al., 2007). The Family Health Conversation model is an intervention that typically consists of three conversations over a period of six to ten weeks that are intended to solve problems that negatively affect family health (Benzein et al., 2008). The intervention is concluded by sending the family a closing letter with nurses' reflections. From studies regarding Family Health Conversations, it appears that the intervention was delivered by nurses that were not involved in daily care for the patient (Benzein et al., 2015;Östlund et al., 2015;Persson & Benzein, 2014).
Within the current study, a family nursing intervention that is intended to be conducted on a regular basis as part of routine nursing care is described. Incorporation into routine nursing care is considered important to facilitate family-professional collaboration in long-term care situations. Therefore, this intervention, the family nursing conversation, is to be conducted by the nurse 63 that also provides and coordinates the regular care for the patient. Family nursing conversations are aimed at fostering family resilience, facilitating collaboration between family members and professional caregivers, and preventing or decreasing caregiver burden. The Family Resilience Framework identifies three domains of key family processes that professionals can focus on in order to foster family resilience (Walsh, 2003(Walsh, , 2016b(Walsh, , 2016a. First, resilient families hold beliefs that are optimistic and hopeful and that allow them to give meaning and purpose to the adverse situation. Second, in terms of organization, resilient families are flexible and able to adapt to a changed situation; they are connected, support each other, and can tolerate differences; and they have access to resources. Third, communication of information and emotions in resilient families is clear and open, and families collaboratively solve problems and make shared decisions. The family Health Conversation components (Östlund et al., 2015) have been adapted to allow incorporation into routine nursing care and achievement of the aims of family nursing conversations, especially family resilience processes, as shown in Table 1. All of the components are to be applied in relation to the care situation. More widespread or complex individual or family issues may be identified but are only discussed further when they are relevant to the care situation and are within the nurse's expertise. When this is not the case, nurses will refer the family to appropriate professionals and focus the family nursing conversation on the care situation.

Aim
This study is part of a larger project in which family nursing conversations are developed, implemented and tested in nursing practice in the Netherlands, using the Knowledge-to-Action framework (Graham et al., 2006). As part of the implementation process, this framework emphasizes the need to monitor how knowledge is actually used when it is applied in practice, in order to further adapt it to the local context and ultimately evaluate its effects. As this is the first time the family nursing conversation components are transferred from theory to home health care practice, the aim of this study is to describe how nurses' apply each of the described components in their family nursing conversations in home health care. Components of family nursing conversations (adapted from Östlund and colleagues (Östlund et al., 2015)) Family resilience processes and family functioning domain* (Walsh, 2003(Walsh, , 2016b 1 Jointly reflecting with the family on expectations of the conversation, and jointly setting the goal for the conversation.
Clarity 1 Collaborative problem-solving 1 Positive outlook 3 2 Getting to know each other; who is present and who is absent.
Connectedness 2 Social and economic resources 2 3 Exploring the family structure and finding out who is part of the family by making and discussing the genogram with the family.

Design
This study was conducted using a qualitative content analysis (Elo & Kyngäs, 2008) with a deductive approach. The units of analysis were the transcripts of audiorecorded family nursing conversations and nurses' reflection forms about these conversations.

Participants
Ten home health care nurses from three home health care organizations in the northern part of the Netherlands conducted the family nursing conversations. These nurses coordinated and participated in the routine care for the patients. Nurses were all female with a mean age of 47 (± 9) years and, on average, 13.5 (± 12) years of work experience. All of the nurses had recently received a six-day educational intervention on Family Systems Nursing and family nursing conversations as described elsewhere (Broekema et al., 2018). The family nursing conversation components were part of the educational intervention.

Data Collection
In the three months following the educational intervention (January -April 2017), participating nurses were asked to organize and conduct family nursing conversations with three families from their daily practice. In accordance with the aims of these conversations, nurses selected families with challenged family functioning or with family caregivers at risk for overburden. The selection, therefore, was based on the following NANDA-I nursing diagnoses (Herdman & Kamitsuru, 2014): 1) Risk of caregiver role strain; 2) Caregiver role strain; and 3) Interrupted family processes. The nurses individually conducted the family nursing conversations. When the family agreed, the nurse audio-recorded the conversation and subsequently completed a written reflection form on which she reflected on her application of each of the components.

Ethical Considerations
Approval of the research project of which this study was a part was waived by the medical ethical committee of the university, as the study does not fall under the Dutch Medical Research Involving Human Subjects Acts. All of the participants to the family nursing conversations received verbal and written information about the study's purpose and procedures. Participants were informed that their data would be treated confidentially, and all names, addresses, and other identifiable personal details would be removed from the transcripts and in the analysis. The conversations were audio-recorded when all of the participants in the conversations provided their 66 written informed consent. Participants could refuse or withdraw their consent at any time without consequences for the received nursing care. The audio recordings and transcripts were stored without identifiable information.

Data Analysis
All of the audio-recorded family nursing conversations were transcribed verbatim and analyzed using Atlas.ti 8.0. Before analyzing a transcript, the researcher read through the transcript while listening to the audio-recording, both to check the transcript's accuracy and to gain understanding of the complete conversation. The analysis focused on the manifest content of the transcripts and the reflection forms.
Analysis occurred in three phases. During the first deductive phase, all occurrences of the 12 components in nurses' contributions to the conversations were coded as such. First, four research assistants (fourth-year bachelor nursing students) coded each conversation in pairs and then discussed their coding to reach consensus. Subsequently, the first author independently coded the conversations and compared the coding to that of the research assistants. Discrepancies were discussed and easily resolved. In the second phase, all text fragments within each component were read through multiple times in order to allow creating a description of the application of the components. Fragments were also read in the context of the complete conversation to describe how components were integrated. The first author formulated the descriptions and subsequently discussed and refined these with the other authors. In the final analysis phase, the outcomes of the first two phases were compared with the written reflections that nurses provided on their use of the components. Attention was paid to any clarifications or explanations that nurses provided in their reflections regarding their application of the components. Comments from the reflection forms were integrated into the descriptions when relevant to explain the application of the components.

RESULTS
Within the preset time period, a total of 17 family nursing conversations with 17 families were successfully audio-recorded. An additional 15 conversations were conducted but were not audio-recorded due to failing recorders (n=7), families that declined permission for audio recording (n=5), and nurses who felt uncomfortable asking permission for audio recording (n=3). Each nurse delivered at least one audiorecorded conversation. Saturation was reached and quotes from all 17 conversations were considered for use in the descriptions. The 17 audio-recorded conversations lasted 41 (± 16) minutes on average. For 13 of them, the nurse completed the reflection 67 form. An overview of conversation participants is provided in Table 2. Component 3: Exploring the family structure and finding out who is part of the family by making and discussing the genogram with the family. Two variants of this component occurred. In some conversations, the nurse explored the family structure and made a genogram during the conversation, in others, the family structure was explored without making a genogram. In the reflection forms, these nurses explained that they did not feel comfortable drawing during the conversation. Some nurses made the genogram afterwards. The exploration of the family structure was usually led by the nurse with specific questions about family members. In cases in which a genogram was made, this was usually introduced informally: Differences existed in the extensiveness of the exploration, e.g., the number of family members and individual characteristics that were included. When the family structure was not explored, nurses reported that they either already had sufficient 69 knowledge about the family structure or that they prioritized other topics.
Component 4: Exploring relationships within the family and relationships between the family and other people and organizations by making and discussing the ecomap with the family.
For this component, again, two variants occurred: an exploration of the family's relationships by making an ecomap or a verbal exploration of the family's relationships with an ecomap sometimes being made afterwards. The exploration of relationships within the family was usually combined with component 3, exploring the family structure. Nurses decided what aspects of family relationships were discussed, such as the quality of the relationship, i.e., "Good contact with her?", or the support that is or could be provided, i.e., "You get a lot of support from them?". Family members were subsequently openly invited to share others who were important to them which afforded an exploration of relationships with people outside the family and with organizations.

Nurse: And do you have friends or other networks that you would say are very important to you? [Conversation 3]
Component 5: Inviting each family member to share their story and narrate expectations, needs, and emotions related to the care situation.
Nurses primarily asked open questions in order to elicit stories and the subsequent emotions, needs, and/or expectations. As family members narrated their stories, nurses encouraged them and suggested new topics. In some conversations, all family members were invited individually to share their story. In others, the nurses invited the family as a whole and sometimes the two approaches were combined. Family members that were less visible in the conversation were usually actively involved by the nurse. In the reflection forms, nurses indeed stressed their efforts to provide every family member the opportunity to share their story. This is evident in the following successive actions from the nurse; family members' responses are left out. Some nurses explained in the reflection forms that they did not acknowledge a painful topic when it had been previously discussed multiple times during the conversation or when the nurse knew from prior experience that it would have overly upset the participants. Nurses reacted to beliefs that family members spontaneously mentioned, and sometimes actively elicited family members' beliefs. Support of facilitative beliefs occurred with a brief confirmation (e.g., "That's true"), by reformulating or repeating the belief, or by sharing opinions or experiences that affirmed the belief. This usually occurred in the flow of the conversation. Challenging constraining beliefs involved explicit discussion of beliefs and offering alternative beliefs: In the reflection forms, some nurses mentioned that they signaled constraining beliefs but either did not want to confront the family at this stage of the contact or did not feel capable of challenging these beliefs. Supporting facilitating beliefs was considered to be easier.

Component 11: Summarizing the central issues that have been raised and pursued in the conversation.
Summaries occurred throughout the conversation. At the beginning, some nurses summarized previous conversations. During the conversation, brief summaries of a topic that had just been discussed were offered in order to mark the transition to a next topic. At the end of the conversation, sometimes all of the issues that were raised during the conversation were summarized, sometimes only the last topic, and sometimes only the agreements that were made. In some situations, nurses actively invited the family to respond or contribute to the summary, as in the following quote. After each question or statement by the nurse, family members confirmed what was said or added information; these responses are omitted:

DISCUSSION
This study aimed to describe how home health care nurses applied family nursing conversation components during the first transfer of these components to nursing practice, as part of a larger implementation project. All components occurred in the conversations. Nurses typically introduced components implicitly by applying them, e.g., asking a question about the family structure or inviting family members to share their story. Some components seemed more easy to apply than others, for example 'inviting family members to share their story' and 'giving commendations about family strengths, competencies and resources'. Nurses described other components as more difficult, for example 'jointly reflecting with the family on expectations for the conversation and setting a joint goal', and 'signaling and challenging family members' constraining beliefs related to the care situation'. The reflection forms revealed some lack of self-confidence or skill. This will need to be resolved, possibly through gaining more experience, since in the Swedish study into the components of Family Health Conversations (Östlund et al., 2015) the occurrence of some components already increased in the second and third conversation. In addition, in our earlier study evaluating the educational intervention (Broekema et al., 2018), nurses indeed recommended additional experience rather than more 74 education to improve their feelings of competence. Nevertheless, it may also be that additional educational needs will arise.
Components were applied in connection with each other. Nurses for example typically extracted the 'shared question or problem regarding the care situation' from 'family members' shared stories, expectations, needs and emotions'. Subsequently the shared question was formulated as a goal to which 'joint goals and agreements for the care situation' were related. Moreover, nurses tended to apply a certain structure in the components: during the first four components, the nurse gathered specific information, using closed-ended questions in accordance with communication theories (Ilie, 2015;Williams, 2005). Then starting with component 5, the family was encouraged to share and participate, through a larger number of open-ended questions. In terms of family resilience processes, the focus was first on organizational patterns and shifted gradually to communication/collaborative problem-solving and belief systems (2,22).
Nurses adapted their application of the components in order to optimally serve the needs of the care situation and the family. An example is the decision to postpone the exploration of the family structure when a family is obviously preoccupied with an urgent issue in the care situation. These adaptations were generally well substantiated and based on nurses' clinical judgment (Manetti, 2018). A flexible approach rather than a strict protocol in applying the components may, therefore, be argued for. In future education, explicit attention to clinical reasoning and decision making in the context of family nursing conversations would be valuable for optimizing the fit between the intervention and the family situation (Johansen & O'Brien, 2016). Such a need for adaptation to the context in order to have optimal effect is one of the factors that defines an intervention as complex (Craig et al., 2019). Fidelity is traditionally defined as the degree to which intervention components are conducted as planned (Schoenwald, 2011). In complex interventions it may be more feasible to define fidelity as the degree to which the underlying function of components is achieved in practice (Hawe, 2015). To assess and optimize nurses' fidelity to the function of the components, it will be necessary to come to understand the working mechanisms of the family nursing conversations components in practice (Moore et al., 2015). This study provides an overview of the way nurses apply the components in family nursing conversations in their every-day home health care nursing with a heterogeneous sample of patients and families. As such it demonstrates the real-world applicability of the components in home health care nursing, by regular nurses that participated in a six-day educational intervention (Broekema et al., 2018).
This study does not allow statements about the effectiveness of the conversations with regard to the aims of decreasing family caregiver burden and achieving family resilience and family-nurse collaboration. However, the joint goals and agreements that were developed during the conversations were generally related to these aims. In addition, the components were applied in a way that family resilience processes (Walsh, 2003(Walsh, , 2016b were encouraged: family structure and social resources were discussed, a positive outlook was encouraged, and open communication and collaborative goal-setting and problem-solving occurred. Family resilience processes could be further encouraged through meta-communication with the family about the purpose and expectations of the family nursing conversation at the beginning of the conversation. This first component was hardly present in the conversations that were analyzed in this study. The component is however important for immediately alerting families that this conversation will be different from other contacts with health care professionals in that collaboration with the family and therefore the family's contribution is crucial. By applying the component, family resilience processes could be encouraged from the beginning of the family nursing conversation.

Limitations
First, visual recordings of the conversations in addition to the audio recordings would have allowed inclusion of the nonverbal strategies that nurses used to apply the components. Second, this study only provides insight into the application of the components by nurses who were relatively inexperienced in family nursing conversations. It is, therefore, not possible to disentangle needs for more experience from needs for further educational interventions. Thirdly, this qualitative study cannot be used to assess the overall quality of the conversations that were conducted, as it only describes the ways in which nurses applied the components and nurses' reflections on their application. The results indicate that the needs of each specific care situation should be taken into account in order to evaluate the quality of a family nursing conversation; only assessing the degree to which a conversation includes all theoretical components does not suffice. Finally, despite and partly due to the heterogeneous sample, the relatively small sample size did not allow for exploration of the application of components in subgroups of patients, families or nurses. The study therefore provides limited insight in the reasons behind variation in nurses' application of the components; insight is solely based on nurses' reflection forms.

Conclusion
Nurses applied the family nursing conversation components in a cohesive manner tailored to the care situation. Nurses' application of the components demonstrates 76 that the components can be applied in daily home health care nursing. It will be important to assess their applicability in other settings including hospital care, residential care, and mental health care. Nurses' clinical judgment was important to tailor the components to the needs of individual families. Future research is necessary to assess the effectiveness and working mechanisms of family nursing conversations according to the described components in fostering family resilience, preventing family caregiver burden, and optimizing collaboration between the family and professional caregivers.