Implementation of communication routines facilitating person‐centred care in long‐term residential care: A pilot study

Abstract Background Specific routines such as the development of personal communication plans can improve the interaction between people with communication disorders and healthcare staff. Objective: This pilot study explores a model for implementing communication routines including personal communication plans in long‐term residential care. Design, Setting and Participants This multiple case study includes two residential care facilities that differ in size and the number of languages spoken. Intervention or main variables studied implementation strategies involved workshops, individual coaching and follow‐up visits. Main outcome measure implementation was assessed using goal attainment measurements, and the staff's views about facilitators and barriers to implementation were explored through structured interviews using the Theoretical Domains Framework. Results The overall implementation success rate for the facilities was moderate, and one of the facilities showed promising results related to personal communication plans. Both facilities experienced barriers to implementation, including management problems and a lack of reminders. However, the two facilities seem to have different motivations for change. Discussion and Conclusions Regardless of the features of the facility, successful implementation requires stable and committed leadership. Moreover, experiences with language diversity may motivate staff to implement communication routines. Patient or Public Contribution The content of the implementation model used (i.e., choice of specific routines and implementation strategies) was coproduced by staff, managers and the researchers involved in this project. The staff and managers were responsible for the implementation of the new routines under the supervision of the researchers.


| INTRODUCTION
In long-term residential care, meaningful interaction is crucial for the well-being of residents. 1 However, long-term care facilities provide limited possibilities for communication outside of care routines. [2][3][4] In Sweden, only the most dependent elderly people receive residential care. Residential care residents often have hearing loss, cognitive decline due to ageing and communication disorders due to neurological disease or injury. These conditions may severely affect a person's ability both to understand and to convey their wishes and feelings. In addition, language diversity among residents as well as staff is common and this creates additional challenges for everyday interaction. 5,6 Person-centred care (PCC) entails that the beliefs, needs and preferences of the person in need of care are respected and a partnership is established between patients and staff. [7][8][9] Patients are not passive recipients of care but play an active role in planning their own care with their care providers. Studies exploring PCC in various contexts have shown that PCC has a positive effect on the cost of care, the duration of hospital stays, 10,11 patient selfefficacy 12 and the level of stress in healthcare staff. 13 In PCC, the formation of personhood is seen as an activity that requires interaction with others. 14 Therefore, a conversation partner plays a crucial role in supporting communicatively vulnerable patients. 15,16 Several researchers have assessed programmes that create routines that facilitate communication targeting healthcare staff and people with communication disorders. [17][18][19][20][21][22][23] For example, Simmons-Mackie et al. 17 have found promising results from training in supported communication and the setting of specific goals that enhance communication access and participation for residents with aphasia. The success of communication plans has also been reported. 18,19,22,24,25 A communication plan, which is drawn up by staff and residents together, is a summary of a resident's communicative ability and preferred communication strategies. In a study of nurses who work in a long-term care facility for people suffering from stroke, a workshop and the development of personal communication plans resulted in nurses increasing their use of communication strategies in their interactions with residents. 19 The staff also reported greater awareness of the importance of adapting their communication to different individuals, that is, communication became more person-centred. Implementation science has shown that successful and sustainable implementation in health services depends on several factors. 26 For example, change is more likely if the staff (the intended adopters/ users) want it (i.e., there is tension for a change). 27 Implementation also has to adapt to the local context, 27,28 and some note the importance of engaging key individuals or champions who support the project in the organization's social networks. 29 Moreover, the support of active and empowering managers has been found to be essential, 28,30 and giving feedback on the impact of the implementation increases the chance of long-term success. 27 However, there is no gold standard for combining implementation strategies concerning behavioural changes in staff working in residential care facilities. 31 In addition, strategies used for implementation are sensitive to context. 31,32 Hence, before embarking on a large-scale implementation effort, baseline facilitators and barriers to implementation need thorough exploration. 31 Although communication plans have proven helpful, more research is needed on how to implement this kind of routine in the long-term residential care context where residents and staff experience a variety of barriers to communication. Therefore, this pilot study explores a model for implementing communication routines, including personal communication plans for all residents, to facilitate PCC in two long-term residential care facilities. This study has two aims: to explore the level of success in implementing personal communication plans as well as other routines supporting interaction and to present a preliminary exploration of facilitators and barriers to implementation.

| METHODS
Using an exploratory and multiple-case design, this study examines two municipal long-term residential care facilities for older people in the same district located in western Sweden. Before the present study, both facilities had taken part in a project to implement PCC that focused on incontinence. The participating staff in this project described difficulties providing PCC for residents with communication disorders, which led them to contact the authors of the present paper. That is, the intervention was initiated by the staff from the two participating facilities and the content of the intervention was planned and coproduced with the participants in this study.
In Sweden, long-term care is decentralized and highly subsidized as municipalities are responsible for institutional care such as residential care facilities, and approximately 95% of the cost is covered by county councils and municipalities as well as national taxes. 33 Residents pay the remaining 5% of the cost, which includes care, rent and meals. Since the early 2000s, municipal institutional care has been seriously downsized resulting in only the most dependent older people receiving residential care (i.e., people who require more assistance and medical attention than is possible to provide in their homes). Professional care staff working in residential care facilities generally include management, registered nurses (RNs), physiotherapists and occupational therapists. However, the largest staff group is enrolled nurses (ENs) and/or nursing assistants (NAs).
ENs require a secondary education in nursing (at high school) or approximately 60 weeks of postsecondary education, but NAs do not generally require any formal education. Both ENs and NAs manage daily care under the supervision of an RN. For simplicity, all the nursing staff are referred to as ENs in the following text. Typically, speech-language pathologists are not on staff but can be consulted as part of specialist care.

| Residential care facilities
The first facility (F1) has 20 one-room apartments divided into three units. Two of the units offer dementia care and one offers care for FORSGREN AND SALDERT | 2983 those with physical disabilities. At F1, all available care staff working on the units (28 ENs) agreed to participate, and four of them became key ENs (i.e., ENs who had specific responsibility for the project; see Section 2.2 below). However, only 24 ENs provided background data (e.g., age, language background and employment) and 13 of the ENs participated in workshops included in the implementation project. All ENs who returned background data were female (20-66 years old; medium: 46 years). The ENs had 2-41 (medium: 22) years of work experience in healthcare and Swedish as their native/best language (Supporting Information: Appendix 1). One of the key ENs dropped out early in the implementation process due to sickness, and at the end of the project, none of the key ENs was still working on the unit. one UM was present initially, and another manager was employed just after the implementation process started. During implementation, the first manager retired, and a third manager was employed. By the end of the project, UM2 was on sick leave and her job had been taken over by a fourth temporary UM. At that time, the third UM had changed jobs. Four of 43 residents were more actively involved in the study (e.g., participated in Phase 3 and provided background information). As in F1, the residents experienced a variety of barriers to communication spanning from hearing loss to dementia. All four participants had Swedish as their native/best language (Supporting Information: Appendix 4).

| The implementation model
This section describes the four phases of the implementation process and the activities and strategies adopted ( Figure 1).

| Phase 1-Initiation of the project
The first phase, which explored the tension for change, began with a meeting with the two UMs and two female speech-language pathologists (SLPs; the authors of this paper, E.

| Phase 4-Monitor and feedback
In Phase 4, four follow-up visits were conducted at both facilitiesimmediately after the workshops and goal setting and 4, 7 and 12 months later. The first author E. F. was on site approximately 8-15 h a week over 1-2 weeks at each follow-up and was available if the staff requested assistance. Another important task was to monitor and give feedback on goal attainment.

| Method for exploring success in implementation
To assess the level of success in implementing communication routines, a measure of attained unit specific goals was collected based on an adapted version of the Goal Attainment Scaling. 34 This was done after the workshops and unit goal setting at follow-up 1 and at three additional follow-ups (4, 7 and 12 months after goal setting).
The UMs in collaboration with the first author E. F. specified the steps needed to attain each goal. These steps were converted into a three-graded scale of goal attainment. For example, for goal 5 at F1

| Method for exploring facilitators and barriers to implementation
To explore the facilitators and barriers to implementation, formalized interviews were conducted by the first author E. F. at follow-up 4, a year after the workshops and goal setting. At F1, one EN (who had participated in the initial planning meeting and had taken some responsibility for the unit-specific goals) and the planner were interviewed. At F2, two key ENs (key EN1 and key EN3) were interviewed. The first author E. F. had previous experience in interviewing and had continuous contact with all the respondents during the implementation. All respondents were involved during the whole process of implementation (more or less) and agreed to an interview. The respondents have approached face-to-face and all accepted participation.
An interview guide was constructed using the Theoretical Domains Framework (TDF). 35,36 TDF is based on theories relevant to behaviour change in healthcare providers and it is divided into 14 categories or domains. 36 The TDF domains were used to select 16 questions for the interview (Supporting Information: Appendix 6).
The questions were not pilot tested. The interviews were conducted at the units in a private room where only the interviewer and the respondent were present. Interviews were audio recorded and lasted between 25 and 60 min. No field notes were made during and/or after the interview.
Because the interviews were guided by the TDF, a theory-led thematic analysis was performed on the data. 37 That is, the data were sorted into two predetermined themes-barriers to implementation and facilitators to implementation. All interviews were transcribed, and each item of information related to the theme was extracted and condensed by the first author E. F. Interrater reliability was calculated on 20% of the condensed items using blinded assessments made by a second rater (C. S.), and the two raters reached a 100% agreement on the coding of the items as either barriers or facilitators. The Consolidated Criteria for Reporting Qualitative Research (COREQ) 38 were applied (Supporting Information: Appendix 7).

| ETHICAL CONSIDERATIONS
Ethical approval for the study was obtained from the Regional Ethical Review Board of Västra Götaland in Sweden [reg no: 1016-13].
Participants' anonymity was preserved, and they all gave their written informed consent. Participants were informed that their participation was voluntary and that they could withdraw their consent at any time. Extra care was taken in informing participating residents. They were initially approached about participating by one of the ENs working in the unit and were then asked again by the first author using picture support. The residents' significant others, family FORSGREN AND SALDERT | 2985 members or legal guardians were also informed about the study and gave written consent on behalf of the residents when required.

| Implementation success
Analysis of unit goal attainment showed that F1 had attained 40% of the main goal of creating personal communication plans for all residents at follow-up 1, 75% at follow-up 3 and 60% at follow-up 4.
Three of the other seven unit-specific goals were partially attained and four were completely attained at follow-up 4 ( Table 1).
At F2, the goal of creating personal communication plans was not attained for any of the residents directly after the workshop and goal setting (follow-up 1), but 16% were attained at follow-up 3 and 14% remained at follow-up 4. Four of the seven unit-specific goals were partially attained and three were completely attained at follow-up 4 ( Table 2).

| Attitude, motivation and feelings
The respondents saw the goals set as important but not demanding.

| Environmental resources
Factors in the environment such as particular tools, staff availabil-

| Follow-up and reminders
Follow-up visits were conducted by the first author on four occasions. These reminders of the goals were seen as facilitating.
However, barriers included the lack of additional reminders and checks.
At F1, although it was said that an outsider can 'push to get  30 propose that if the management is unable to actively help empower staff, then this task should be delegated to other formal leaders. Although the key ENs are not management or formal leaders, their status as informal leaders was one reason that they were selected for our study. Nevertheless, for various reasons, including staff turnover, these key ENs were not active throughout the project, a shortcoming that negatively impacted on-site guidance. Regrettably, the problem of high turnover of staff and management is common in long-term residential care. 40 High staff turnover is associated with larger facilities and problems such as lower staffing levels and poorer quality of care. Furthermore, this high rate of turnover means heavier workloads for the remaining staff, which can lead to carers becoming burned out and ultimately the depersonalization of staff and residents. 41 Hence, the rapid turnover of the management and staff is a serious threat to PCC.
Growing workloads can also result from colleagues taking on responsibilities outside of their care routines; at both facilities, time management in relation to colleagues was identified as a barrier to goal attainment.
A difference between the two facilities was found concerning the motivation for change. At  In relation to the broader field of implementation research, our study should be seen as one of the initial steps in a larger implementation effort as it contributes to knowledge of barriers and facilitators for implementation in the context of long-term residential care. In a larger-scale implementation effort, an explicit theory guiding implementation should be stated and a complete evaluation of the particular process of implementation conducted. 31 Research should focus on the process and process evaluations. 45 The entire implementation process should be continually monitored to assess whether it is following the plan, to explore how a particular event/activity/strategy affected participants and to identify how contextual factors interplayed with implementation. 46