The development and process evaluation of the ACTION study: A person-centred communication intervention targeting nurse assistants in home care for older persons


 Background A rapidly ageing population challenges the health care system in general and home care services in particular. Communication is a cornerstone of person-centred care. However, little research has been conducted on how to improve communication between health care professionals and older persons in home care contexts, despite research showing the importance of such interactions. The increasing demands on how to best and efficiently improve competence in health professionals is the reason why the ACTION intervention was conducted. This paper aims to describe the development and process evaluation of an educational intervention for nurse assistants (NAs) in home care and highlights the potential of self-directed web-based learning as well as the pitfalls of conducting complex interventions in home care. Methods A web-based educational intervention focusing on person-centred communication was developed that targeted NAs in home care for older persons. Twenty-seven NAs from two units in Sweden were recruited, and 23 NAs were offered the educational intervention. Data were collected from multiple sources before, during and after the intervention and were analysed using quantitative and qualitative methods. Data were extracted from the web platform and analysed to determine the NAs´ engagement in the intervention. Additionally, interviews, evaluation forms and field notes were used to evaluate the feasibility of the intervention. Results The main findings suggest that web-based education seems to be an appropriate strategy in the home care context. The majority of the NAs (91%) participated in full or in part in the intervention. During the implementation process, some adaptions were required to fit the local circumstances regarding technical support, scheduling, and the design of the lectures. The NAs perceived the format to be easy to use and flexible and appreciated the stepwise modules. The content was perceived as valuable. Conclusions Our findings show that the benefits of the web-based educational intervention included the short and focused lectures as well as its accessibility. Challenges with the implementation process included gaining access to the NAs and motivating and involving the NAs. This study emphasizes the environmental support needed to successfully conduct complex interventions, including physical, organizational and cultural aspects.

older persons (25). Working in home care services is characterized by mobile, dynamic and sometimes unpredictable working conditions. NAs move between home care units and the homes of older persons according to a combination of set schedules and the emergence of care needs. This poses challenges for NAs´ competence development and learning. Previous work (26,27,28,29) has proposed flexible learning approaches such as web-based or blended approaches that combine elearning and face-to-face interactions. These types of approaches have been effective in a wide range of health professional contexts and purposes, e.g., for providing advice and support to cancer survivors (28), conducting motivational interviewing (27), and changing clinician behaviour (26). A meta-analysis of 201 articles performed by Cook et al. (29) described positive effects of web-based training for health care professionals, compared to no intervention, on knowledge outcomes, skills, learner behaviour and patient effects. Research on how to best and most efficiently improve PCC and communication competency in home care contexts needs further exploration.

Aim
The aim of this paper was to describe the development and process evaluation of the ACTION intervention. The following research questions were addressed:

1.
What were the NAs levels of adherence to the intervention?

2.
What were the NAs expectations for the intervention? 3.
What were the NAs experiences of the implementation of the intervention? 4. What experiences were described by the NAs following the intervention?

Design and development of the intervention
In the ACTION study, a web-based educational intervention for person-centred communication targeted at NAs in home care for older persons was developed. This paper reports the development and process evaluation of the intervention. During the development of the intervention, previously implemented complex interventions in applied clinical research were considered with the aim of succeeding in applying an effective and feasible intervention (30). A team consisting of researchers and teachers with expertise in the area developed the intervention. During the planning and development of the intervention, users and stakeholders in home care were involved.
The intervention was designed to enhance learners´ communication competency through flexible web-based education that combined theoretical knowledge with reflections, supervision and course leader feedback. The backbone of the intervention was a web-based learning platform where several types of content and learning strategies were combined, including video lectures, movie clips, quizzes and reflection assignments, which were complemented with in-person group sessions with supervision. The content was structured in eight stepwise modules (see Table 1) focusing on communication and PCC from a caring science perspective; the content was intended to be distributed over eight weeks, with approximately one module administered each week. Two of the modules were organized as live group sessions to be conducted at the workplace. The first four modules were pre-recorded when developing the intervention. The last three modules were designed based on the participants input and requests regarding the content and structure and consequently, were recorded during the early stages of the intervention. To provide flexibility, the NAs themselves decided when to carry out each module. Each module was planned to take a maximum of 15-30 minutes, and the movie clips and web-based lectures used were between 5-25 minutes in length. The quiz or reflection assignment for the current module had to be completed before the NA could proceed to the next module. The intention of the assignments was to enhance self-directed learning and to support deep learning and reflections and prevent the participants from "clicking through" the course pages without engaging with the content. The educational content was tailored to NAs and built upon the previously described framework of personcentred care and communication with older persons, in addition to previous research from the COMHOME project (31).

Study setting and recruitment procedures
The study was planned and prepared in 2017, and the intervention was performed in spring 2018. A convenience sample of NAs was derived from a middle-sized community in western Sweden. The eligibility criterion for participants was having a permanent job position enabling engagement in the intervention and data collection. All eligible NAs from two geographical districts at the home care service organization were recruited, resulting in 27 NAs. Of these NAs, one declined to participate, one was on long-term sick leave, and three terminated their employment. After one new employee was hired, 23 NAs were eventually offered the intervention. The research team planned how the NAs would participate in the educational intervention, but the manager of the home care units, the team leaders and the NAs themselves were responsible for organizing the time and resources for the intervention at their home care units. For an overview of the study process, see Figure 1.

Data collection
Data were collected from multiple sources before, during and after the intervention, and consisted of user analytics data and data from pre-and post-interviews, evaluation forms, and field notes (see Figure 1).
Before the start of the intervention, data were collected via participant demographic questions, group interviews and field notes. Twenty-two of the NAs participated in four group interviews, which focused on their expectations and worries regarding the upcoming educational intervention. Towards the end of these group interviews, the NAs were asked to write personal notes about their thoughts on anonymous Post-It notes.
During the course of the intervention, user data were extracted from the web-based platform on the Field notes were written continuously during the process; the contents of field notes included reports from visits; notes from discussions with management, team leaders and NAs; reflections; e-mail communication; and questions to discuss with the project team. The field notes were used to gather input on the process, design and content of the intervention.

Data analysis
User statistics from the web-based platform were analysed using IBM SPSS Statistics version 25 to analyse the participant flow and identify patterns in adherence and use. To analyse participants' adherence and activity on the web-based system, time and number of visits in each module were analysed. Pearson correlation tests were performed to analyse the potential impact of background variables, e.g., age, gender, and language, on adherence. Interviews, evaluation forms and field notes were analysed qualitatively to explore the NAs´ experiences of the intervention. A qualitative content analysis by Elo and Kyngäs (32) was performed. In this step, the data were read several times and then coded and grouped, resulting in three categories: expectations and worries before the intervention; the implementation process; and experiences from the intervention. In the results, quotations from the interviews and evaluation forms were used to illuminate the NAs´ experiences.
The analysis was mainly performed by two members of the research team, and further discussed and validated by all authors. All authors read and commented on the final version of the analysis to ensure the rigour of the categories described.

Results
The results are presented in five sections: (I) demographic characteristics of the participants, (II) participants´ adherence to the intervention, (III) expectations and worries before the intervention, (IV) the implementation process, and (V) experiences from the intervention.

Demographic characteristics of the participants
The demographic characteristics of the 23 NAs who were offered the educational intervention are presented in Table 2. The group had a median age of 39. The gender distribution (male=3, female=20) was in line with the current overall characteristics of NAs, as the NA profession is female dominated. NAs´years of experiences working in home care varied. One-third (n=7) of the participants had a native language other than Swedish. 1.

Participants´ adherence to the intervention
The majority of the NAs had participated to a large extent in the educational intervention (see Figure   2). Most NAs, 91% (n=21), completely or partially participated in five or more modules. A few of the NAs, 9% (n=2), participated in four or fewer modules. After the fifth module, some of the participants dropped out. The last module was completed by half of the NAs (n=12, 52%).

Expectations and worries before the intervention
Overall, the NAs´ expectations before the education included both curiosity and concern. Not knowing what was to come was exciting and led to interest in what the content could contribute to their knowledge and competence development. At this point, the NAs expressed that the information they had received on the design and content of the educational intervention were limited. They did not express any concern regarding the web-based format, and they looked forward to being introduced and beginning it. 1.

The implementation process
Important aspects during the implementation of the intervention were found to be related to reminders, time constraints, when to actually complete the educational intervention, the equipment at the workplaces, and modification of the lectures. Questions also arose regarding the physical location where the NAs were expected to work with the course material. The manager's recommendation was to primarily do this at the workplace during their working hours. If necessary, e.g., due to a tight schedule or heavy workload, completion of the modules from home and being compensated for this as over-time was allowed. This resulted in some NAs preferring to undertake the education at home to not be disturbed.
There were benefits with the web-based format being flexible for NAs and allowing the research team to follow their progression during the education. The web-based format allowed for early detection of those who fell behind or did not follow the modules according to the plan. These NAs were contacted through e-mail or during one of the visits by the researcher, and they were offered help if there were any technical or other problems. Some difficulties in contacting NAs arose, as some did not respond to e-mails or rarely were at the workplace during visits. The team leaders helped contact those individuals. A few NAs mentioned technical problems logging on to the web platform. After they contacted the responsible researcher, help was received from the IT support at the university.
Another challenge noted was the limited number of desktop computers, with two per home care unit, allocated for the intervention and their everyday work. Some NAs expressed worries concerning noise and privacy when listening to the web-based lectures and when writing assignments. To establish a sense of empowerment and secure resource allocation, the research team decided to provide the home care units with tablets and headphones.
During the implementation of the intervention, some NAs indicated that the first lectures were too long and disengaging in some parts. This feedback was taken into account when developing the last three modules, resulting in shorter lectures and more engaging approaches, such as conversationbased lectures including several presenters.

Experiences from the intervention
Overall, the NAs found web-based education to be easy to use, and appreciated the stepwise modules. Some NAs found the web format challenging due to a lack of digital experience.
"Too much using the computer, especially when you're not that technical." Compared to traditional education, the web-based education format included benefits such as accessibility and flexibility to independently decide when to undergo the modules. The tablets and headphones offered were appreciated by the NAs, as they allowed for additional flexibility and privacy. The overall structure and variation was appreciated, but some NAs expressed that the intervention consisted of too many modules. Another weakness with the intervention being web-based was the lack of interaction with the lecturers, and some NAs asked for more engaging forms of lectures. Some lectures perceived as too long, resulting in a loss of interest or the module being interrupted. The opportunity to discuss difficult matters with each other was valuable and provided the NAs with useful insights into their daily work. There was a desire for more opportunities similar to the group discussion and to meet in groups and to reflect together on different aspects of home care. Others thought that one session was sufficient.

"I think that it [the education] was great! It was great that you could log in and then
The short movie clips were found to be particularly interesting and stimulating. These could provide nuance and deepen the participants´ understandings of the education content.

"Some of the movie clips have been good. I remember one in particular that was about end-of-life
care. The nurse assistants in the movie did it all so beautifully, and it consisted of the most important parts." All the modules except the group supervision were followed by either a quiz or a reflection assignment. The NAs reported that the quizzes were fun and that the reflection assignments were meaningful parts of their learning but that they were time consuming. Participants completing the reflection assignment in the final module received written feedback from the researchers. This feedback was expressed as positive and valuable since it provided confirmation on how the NAs reflected on their current communication ability.
"I think that the reflective assignments gives you more… They feel better… but they are more burdensome and time consuming. /…/ The first two weeks, I felt that this is something I must do at home so you can think more; you need more time for them." Some NAs perceived the intervention as burdensome, and a few expressed that they felt forced to participate. They also expressed frustration with the lack of involvement in the decision phase and felt that the intervention imposed an additional workload, which influenced their motivation to complete the modules. Other aspects that reduced the NAs´ motivation during the intervention were

Discussion
To our knowledge, this is the first study to use a web-based approach to improve person-centred communication skills among nurse assistants in home care. The intervention seems to be an appropriate strategy in this context compared to traditional classroom-or group-based education, but further improvements are required. Our findings emphasize the benefits of the web-based learning format, the challenges of implementation particular to the home care context, and the importance of organizational support and resources.

Web-based learning in home care settings
The feasibility of the ACTION intervention was confirmed by the relatively high level of adherence among the NAs and because it was generally well received with respect to both format and content.
The main advantage of the intervention was the web-based format, with learning modules designed to be short and focused. The course content was easy to access and required limited time commitment from the NAs. Today, there is an increased demand for e-learning in nursing overall (26), especially in the area of communication (25,33). Evidence suggests that web-based education on communication can improve healthcare professionals' communication self-efficacy (34), which in turn is a strong predictor for learning and future performance (35).

Implementation challenges
During the implementation process, challenges with gaining access to the NAs arose. Initially, the manager facilitated researchers' full access to the NAs at the home care units. In practice, access was difficult, mostly due to the NAs´ working conditions. During the onsite visits by the researchers, contact was primarily made with the team leaders and occasionally with the manager but seldom with the NAs. Most of the information and support offered by the researchers was addressed to the team leaders, making them informal local facilitators who were important in supporting the implementation and participant involvement. Consequently, the importance of local facilitators was underestimated.
As emphasized by Schön et al. (37) and Rogers (38), the importance of local champions when implementing innovative tools and technologies in organizations must be addressed. Hence, as this study shows, identifying, planning for and clarifying the different roles of those involved in the intervention, and the functions and expectations related to these roles, is crucial early in the process.
Thus, teamwork with local stakeholders, facilitators and participants is important to succeed in the implementation of clinical interventions.

Organizational support and resources
When participating in this type of self-directed learning intervention, high levels of participant involvement, motivation and responsibility are required for successful implementation. Hence, both the physical environment, e.g., equipment and locations, as well as the workplace culture need to be supportive to enhance NAs' participation in the intervention. This is often referred to as facilitating conditions and has been identified as a crucial factor for the successful implementation of new technologies and tools (38, 39). In this study, the computers at the home care units were insufficient and were therefore completed with headphones and tablets. This improvement was not only necessary but also appreciated by the NAs. Another practical problem that emerged in the initial phase of the implementation was the lack of time assigned for the NAs. The education was something that the NAs were supposed to engage in independently "when they found the time to do so" during their heavy workdays. This approach failed almost immediately, resulting in specifically allocated time for the intervention. The participants expressed frustration regarding their lack of involvement in the initial decision phase.
During the early phases of the implementation, attempts were made to involve the NAs in the planning of the content, and suggestions were collected and implemented in the final modules. The NAs' frustration indicates the important role of management regarding attitudes and commitment not only in the decision phase but also during the whole process to retain the motivation needed to complete the education. This finding is in line with those of Venkatesh and Bala (39), who emphasized the importance of managerial efforts both pre-and post-implementation, as well as those of MacCormack, Dewing and McCance (21), who described the importance of creating a culture that can sustain person-centred care in both health care teams and organizations when developing PCC. To effect such changes, these authors underlined that organizations cannot rely on the motivation of individual participants. Instead, commitment is required from all levels to support active learning and achieve a potential culture change.

Strengths and limitations
Conducting, implementing and evaluating complex interventions is challenging, as emphasized by Craig et al. (30). This small-scale study attempted to address some of these challenges in the context of home care. Although analysis of learning outcomes or effects of the intervention was not included in this study, the findings show some promise with respect to how this type of intervention can be conducted. Nevertheless, a number of limitations need to be recognized, including the number of included participants, interpretation of the data, and involvement of researchers during the process.
The number of participants and the specifics of the local context in this study may limit its generalizability and impact, as in any small-scale intervention. Nonetheless, as recommended by the MRC guidelines (30), these types of studies are important since they examine key uncertainties identified during development and implementation. The experiences reported from this intervention contribute knowledge of the factors that appear to influence the applicability and use of this type of intervention.
In this study, we chose a mixed-method approach. Qualitative data from interviews, evaluation forms and field notes gave valuable information on the implementation process and the NAs' experiences of the intervention, while quantitative user analytics data contributed to the understanding of the NAs' engagement with the course content. Both methods have their benefits and their limitations with respect to data collection and interpretation: the data from the web-based platform showed a high degree of participation, and the details of every NA´s activity in each module (e.g., time spent and clicks) could be extracted. However, care should be taken when interpreting these variables in terms of learning outcomes because, as in any indirect observation, it is impossible to know the exact reasons behind the amount of time spent or actions taken at a particular course module page. As such, the post-interviews provided valuable complementary insights. However, there were some potential limitations regarding the data collection. Throughout the intervention, one member of the research team (the first author), had the main responsibility for the implementation and evaluation processes (e.g., evaluation forms and interviews) and was also responsible for developing some of the course modules. This involvement generated both benefits, in terms of establishing trust, and risks, e.g., whether participants felt that they could be honest during interviews. To address this issue, the evaluation also included anonymous written individual evaluation forms. As reported, these forms included both positive and negative feedback.

Conclusions
This process evaluation identifies important challenges and considerations for the successful conduct of complex interventions in home care settings. We found that the benefits of the web-based format were the short and focused lectures, as well as its accessibility and efficiency. Implementing the intervention in the home care context was challenging, due to the NAs´ working conditions. Gaining access to the NAs and motivating and involving the NAs were identified as important aspects of the implementation process. This study emphasizes the importance of a supportive environment, including physical, organizational and cultural surroundings, when conducting a complex intervention in a home care setting. If designed and implemented appropriately, web-based self-directed approaches such as the one studied in this paper can be useful tools in the quality improvement of home care services. Similar educational interventions can be developed for not only nurse assistants but also other health care professionals and for different care contexts.

Ethics approval and consent to participate
The study has been approved by the regional ethical review board in Gothenburg (Dnr. 260/17) and conforms to the Declaration of Helsinki. Written and verbal information about the study was given to all participants, and informed consent was obtained. The sampling of participants and data storage, flow and access are outlined in accordance with legislation and safety routines to safeguard the security, privacy and confidentiality of participants. Handling of personal information is accordance with the General Data Protection Regulation (GDPR).

Consent for publication
Not applicable.

Availability of data and materials
The dataset generated and analysed during the current study is available from the corresponding author on reasonable request.  Figure 1 Flow chart of the ACTION study process, including the recruitment of participants and data collection.

Figure 2
Overview of the participants' adherence to the eight modules of the education.