Process evaluation of the response of nursing homes to the implementation of the dementia-specific case conference concept WELCOME-IdA: A qualitative study

A mixed longitudinal and retrospective qualitative process evaluation design was used to answer the research question [22]. The study was part of a larger process evaluation of the FallDem trial [20, 23] and was conducted alongside the FallDem trial [14]. This report is based on the recommendations of the Consolidated Criteria for Reporting Qualitative Research (COREQ) (Additional file 1) [24].

WELCOME-IdA is defined as a structured, goal-directed, intra-professional method in which nursing staff are guided through the process of searching for potential triggers and causes of challenging behaviour with the help of a structured assessment instrument called the IdA (Innovative dementia-oriented Assessment System) (Additional file 2) [12]. The IdA was developed based on a comprehensive literature review [25] and expert consultation. A detailed description of the development and evaluation of IdA has been published elsewhere [26]. We chose an intra-professional nursing approach to WELCOME-IdA due to the results of a prior feasibility study [10]. Additional disciplines were not directly included in the DSCCs. The main reason was that physicians and therapists are not employed by nursing homes in the national health care system. Due to this structure, the direct participation of other professions in regular case conferences is scarcely feasible. At the same time, we decided not to involve residents or residents’ families in the DSCCs. The results of the feasibility study indicated that nursing staff did not feel competent to include residents’ relatives directly into the case conference. The shared decision-making process took place among the nursing staff. However, interdisciplinary consultation as well as the involvement of residents’ relatives in WELCOME-IdA took place before or after the DSCC.

WELCOME-IdA is embedded within the general theory of hermeneutics and the need-driven dementia-compromised behaviour (NDB) model [27, 28]. It includes a predefined role structure and a process structure (Fig. 1). Both help systemize reflection on the case (the resident with challenging behaviour) and prevent nursing staff from digressing in everyday conversations and from drawing hasty conclusions or developing poorly considered action plans during the DSCC.

The role structure includes four roles. (A) Moderator: The moderator is a person trained in moderating the case conference. She/he ensures adherence to method, time frame and roles. Adhering to the method involves process control, working through the different phases of WELCOME-IdA, and summary and visualization (e.g., using a flipchart) of the results. Adhering to the timeframe requires the moderator to ensure that sufficient time is available for the individual phases of WELCOME-IdA so that the DSCC can be completed on time. The moderator also ensures that the participants adhere to their assigned roles and tasks during the DSCC. (B) Case reporter: The case reporter prepares the necessary information for the DSCC. Based on this information, she/he presents the initial problem to the group and defines his or her expectations for the DSCC. The case reporter should be the primary caregiver of the discussed resident. She/he also ensures that the care interventions agreed upon in the DSCC are subsequently integrated into daily nursing practice. (C) Keeper of the minutes: The keeper of the minutes completes the IdA and writes down the central results in a protocol. (D) Reflection partner: A total of 2–5 people (core team members) should take on the role of reflection partners, who answer the IdA questions by gathering and supplementing information. They provide critical but supportive feedback, form hypotheses and develop hypothesis-driven care interventions.

WELCOME-IdA suggests that all roles in the DSCC except the moderator should be performed by people from the specific nursing ward where the case resident lives.

The process structure of WELCOME-IdA covers six phases: (1) introduction of the case, (2) description of the problem, (3) analysis of the situation, (4) formulation of hypotheses, (5) development and definition of hypothesis-driven care interventions and (6) closing. The IdA supports phases 2–4. For the description of the case, the IdA provides 14 guiding questions for assessing the challenging behaviour; these questions refer to the nature, frequency, context and consequences of the behaviour. For the analysis of the challenging situation, the IdA contains five different domains (state of health and independence in everyday life, communication, personality and lifestyle prior to the onset of dementia, moods and emotions, and environmental influences). Each domain ends with a summary of the results of the behaviour analysis, which supports the formulation of hypotheses about the possible triggers of the challenging behaviour (Fig. 1).

The DSCC should last between 60 and 90 min and should be conducted at least once a month, preferably in a room where there will be no disturbances.

Each DSCC should be preceded by a preparatory phase, which encompasses the formal preparation for the DSCC, the selection of the case and the content-related preparation of the case. Formal preparation primarily includes inviting the participants of the DSCC to the meeting and recording the times in the duty roster. Any team member can propose a resident with challenging behaviour as a case; the team jointly decides whether a DSCC is to be conducted. The proposing person then arranges the DSCC together with the leading ward nurse. A room must be available for the undisturbed completion of the DSCC. IdA sheets must also be prepared for the moderator and the keeper of the minutes.

The case reporter, which is usually the primary caregiver of the selected resident, conducts content-related preparation. Content-related preparation requires the case reporter to outline, at minimum, the problem to be discussed and his/her expectations for the DSCC. To increase the effectiveness of the DSCC, it is further recommended that background information on the resident concerned be collected and prepared.

The DSCC ends with a post-processing phase in which the entire nursing team is informed about the results of the DSCC. Furthermore, the results are documented in the nursing record. The agreed-upon care interventions are delivered to the resident and evaluated at the beginning of the next DSCC.

In each of the 4 nursing homes in the FallDem trial, the intervention started with 2 days of in-service training on WELCOME-IdA and was followed by four facilitated DSCCs per participating nursing team (on-the-job training). Subsequently, each nursing team was expected to conduct a minimum of four case conferences without any external assistance (off-the-job training). Additional training in moderation techniques was offered to at least 4 selected people (a group of moderators) per nursing home. Each nursing home was expected to establish a steering group. This group was responsible for the implementation process (such as the designation of responsibilities and the provision of structural requirements). The steering group was also in charge of conducting an assessment (at the beginning of the study) of the strengths and weaknesses of its organization in relation to the context in which the case conferences were conducted. Based on the results of this assessment, a tailored implementation plan was expected to be developed by the steering groups for each nursing home.

A total of 4 nursing homes with 7 nursing teams were recruited. Nursing homes were included if they were located in North Rhine-Westphalia due to the foundation’s purpose. The leading managers of the nursing homes had to select at least two nursing wards with two nursing teams to participate in the intervention study (Table 1). Three nursing homes belonged to a non-profit care provider; 1 nursing home belonged to a public provider. The nursing homes had a mean size of 78.3 residents (min. 54; max. 100), which is above the national average of 63 residents [29]. The nursing homes had a minimum of 2 units. Most residents had low to moderate levels of care dependency according to the Medical Service of the Health Insurance Funds (MDK) assessment. The levels of care dependency are consistent with the average national distribution of care levels in nursing homes [29]. On average, there were 51.8 care staff at each of the four nursing homes, but nursing home E82 had the lowest number of nursing staff (n = 38). Fifty percent of the nursing staff were registered nurses (qualified by a 3-year vocational training programme), which complies with the legal regulations for German nursing homes [30].

None of the nursing homes had experience with WELCOME-IdA. All participants in the qualitative study were recruited via the stepped-wedge cluster randomized controlled trial (RCT) studying the effectiveness of WELCOME-IdA [14]. Each nursing home had a study coordinator who was responsible for recruiting the interview partners. All interview partners were provided with written material in advance of the interviews. For the telephone interviews, the participants gave verbal informed consent prior to each interview, which was then audiotaped. Written informed consent was obtained for all group interviews prior to data collection. None of the participants received financial incentives or gifts.

During the stepped-wedge cluster RCT conducted between September 2013 and March 2015, an average of 13.8 people (5–22) participated in the 1st WELCOME-IdA in-service training session, and 14.8 people (6–25) participated in the 2nd WELCOME-IdA in-service training session. They represented a skill mix of registered nurses and nursing assistants. The registered nurses comprised the largest group of participants in all nursing homes. Almost all people selected to be trained as moderators of DSCCs also attended the WELCOME-IdA in-service training sessions (Table 2) [23].

WELCOME-IdA recommends that core nursing teams of 5–8 people participate in DSCCs, and this recommendation was partially observed during the WELCOME-IdA on-the job training and off-the-job training phases. Comparing the absolute number of participants (N) with the number who participated continuously (core team) in the intervention, only cluster E29 demonstrated the continuous participation of at least five core members. The people who continuously participated in all components of the intervention were mostly skilled moderators [23] (Table 2). The four nursing homes completed 47 DSCCs in total. Each nursing team performed 7 DSCCs on average (min 5; max 8) during the 7-month intervention phase (Table 2).

The data collection process consisted of longitudinal semi-structured telephone interviews and retrospective focus group interviews, all of which complemented one another to present a more comprehensive picture of the different objects under investigation. The telephone interviews aimed to obtain individual interviewees’ perspective on the ongoing implementation of the intervention, while the focus groups were conducted to learn how the different groups experienced the implementation process. Within both formats, the interviewees were addressed both as representatives of their organization as a group and as individuals. However, the dynamics of the group interviews resulted in data that can be understood as representing the participants’ individual perspectives on the organization.

Telephone interviews: A total of 34 structured telephone interviews with open questions [31] were conducted in parallel to the implementation of WELCOME-IdA (Table 3). The interviews aimed to assess whether the structured preparation and post-processing of the DSCCs had taken place and whether any adaptations of WELCOME-IdA occurred during its application in practice [20] (Table 4).

The interviews were conducted promptly after the DSCCs were carried out, while the nurses’ memories of the previous DSCC and their preparations for it were still very present and information could be gathered regarding preparation for the upcoming DSCC. The telephone interviews were also a means to stay in close contact with the nursing teams.

The research team developed an interview guideline. RG, TQ, DH, UR, or SR performed all the phone interviews with 1 registered nurse per nursing team. For 6 of the 7 nursing teams, the selected nurse was the head nurse of the nursing team. For one nursing team, the interviews were conducted with the nurse manager (Table 4). The interviews lasted 15 min on average (min. 7; max. 24) and were audio recorded.

Focus group interviews: Fifteen semi-structured focus group interviews were conducted with the steering groups (n = 4), the moderators (n = 4) and the nursing teams (n = 7) at the end of the intervention phase (Table 1) to gain insight into the overall response of each nursing home to the intervention (Table 5). The research team created the semi-structured interview guideline. MH and MR performed all the interviews in the nursing homes. Field notes were taken during the interviews. The interviews lasted 40 min on average (min 25, max 61) and were audio recorded.

All interviews were transcribed verbatim and analysed using primarily deductive content analysis, although inductive categories have been developed [32]. This method focuses on the identification and categorical structuring of themes that emerge from the interviews. The main categories were derived deductively from the interview guidelines. The WELCOME-IdA manual [33] was used to develop a coding tree and to define the main categories in a code book (Additional file 3). The categories were further developed and differentiated inductively with reference to the data [32]. The coding was performed in sections (i.e., mostly based on a question and the participant’s answer) to make textual context available for the summary of the results. The coding was performed by one researcher (RG or DH) and checked by a second researcher (DH or ST) to improve inter-subjectivity and comprehensibility. All interviews were conducted and analysed in German. For the purpose of reporting the findings, individual German quotations were translated into English and proofread by a native English speaker to ensure the transparency and trustworthiness of the results [24]. The software MAXQDA 2018 was used for data analysis and data management.

As indicated earlier, there was no continuous “core nursing team”, as recommended in WELCOME-IdA, in any of the four participating nursing homes. Below are the reasons why continuity of participants was not achieved within the nursing teams:

Thus, in one nursing home, the DSCC was always planned for the day on which most of the staff of the respective nursing ward was present (E79).

Altogether, two reasons influenced the decision to not pursue the establishment of core nursing teams: (1) organizational factors; (2) a decision by some of the steering groups to involve as much of the staff as possible.

The results show that the selection of staff members for the predefined roles should not be based solely on competence and time resources and that possible role conflicts should also be considered.

Based on our data, we suggest filling the more formal role of the moderator with staff members without a leadership role to prevent other participants from feeling inhibited. The exertion of structural power is still very much alive in nursing, and staff members experience mixed feelings when a manager becomes deeply involved in the preparation of a case and in reflection on practice [34]. Moreover, the role of the moderator should be performed by staff members who are not the primary caregiver of the resident and who are external to the resident’s ward to prevent the moderator from stepping out of this role and engaging in the discussion about the resident, which is in line with previous research on DSCCs [10].

The role of and the tasks undertaken by the case reporter proved particularly difficult, and the nursing homes reported substantial problems preparing the case for a DSCC, which might be partly because the nursing homes did not establish core nursing teams and thus disrupted continuous learning. Moreover, the role of the case reporter switched between the different primary care nurses from one DSCC to another. Thus, each primary care nurse requires time to establish a routine to perform as a case reporter. The different strategies applied by the nursing homes to overcome the difficulties of the case reporter in preparing the case (e.g., creating pairs, gaining support from moderators) could be included in the future implementation of DSCCs as exemplary methods. The results also reveal a gap, however, because specific training for case reporters is not yet provided. Contrary to the assumption, a structured knowledge circulation process across the educational and preparation phases with different intensities of facilitation cannot be taken for granted. Because the case reporter presents the case and therefore introduces the perspective by which the case is viewed, it is important to ensure hermeneutically trained consciousness from the start and to train the case reporter how to express an understanding of the case [35]. Thus, training of the case reporter should be required in the future curriculum.

Selecting the reflection partners solely from staff members within the resident’s ward or mixing staff members from different wards can have advantages as well as disadvantages. Cooperation and mutual understanding within the care teams can be strengthened but can also lead to rivalry over who has better knowledge of the resident and/or greater ability to reflect. This experience may stimulate withdrawal from engagement because this feeling can lead to the assumption that humiliation may occur, which undermines the conditions for constructive learning while reflecting on a case [36]. A decisive factor might be how different actors have already worked together in the nursing homes before the implementation of WELCOME-IdA. This is similar to the results of a previous study [10] in which the participation of external staff was also perceived as a hurdle when participants did not work together in daily practice. Therefore, in future implementations, the existing working structure should be addressed at the beginning of the introduction of DSCCs with the steering group. The involvement of external parties should only take place when a sufficient routine for performing DSCCs has been established in the resident’s own nursing team. This ensures sufficient competence of the moderator to address a greater heterogeneity of participants.

The results indicate that WELCOME-IdA’s requirement of building core teams was not practicable for various reasons, and positive effects of mixing participants were observed. The selection of individuals for a case conference should therefore be guided by the need to achieve a match between the case and the potential participants in terms of reflection potential and solution-finding potential. Such adjustments to the intervention would provide nursing homes with greater flexibility in dealing with situations such as employee sick leave or vacation. Nevertheless, participants need to have a certain level of knowledge about WELCOME-IdA and the selected case. The particular advantages of the variability versus continuity of the participants must be weighed carefully. The moderators could have an intervening function in this process because they act as consistent key individuals who would potentially be able to counterbalance knowledge differences within the groups.

The results indicate that the preparation and post-processing phases of WELCOME-IdA are particularly important for the success of DSCCs. At the same time, these phases show the greatest need for adaptation within the process structure of WELCOME-IdA.

In the preparation phase, the adaptation of the completion of the IdA sheets beforehand accentuates the need for knowledge circulation between staff members and third parties, such as relatives and physicians. Knowledge about the case is collected and consolidated by the case reporter, whose role is strengthened through this adaption. However, the underlying idea of the WELCOME-IdA concept is to gain a comprehensive picture by integrating the diverse perspectives of participants. The prepared analysis should provide the starting point for thorough on-site discussion and reflection on the case. Furthermore, participants should have the opportunity to propose IdA domains other than those proposed by the case reporter for an in-depth behaviour analysis.

In the post-processing phase, a crucial step for ensuring that the DSCC is successful is the circulation of results (hypotheses and identified interventions) to the entire team. The results indicate that the nursing homes chose communication modes that corresponded with the usual practices in their organization. On the one hand, adjusting modes of knowledge circulation to the social system is necessary to promote implementation [37]. On the other hand, regardless of peculiarities, research shows that some formalization of communication is also necessary [38]. In particular, the promptness and extent of information transfer are of central importance because every delay and every gap during this phase jeopardizes the implementation of the developed interventions.

Digital communication media could support the rapid dissemination of information, and visualization in the wardroom has proven to be promising. Both variants would also enable part-time employees and nursing staff on night shift to access information promptly. In addition, absences of nursing staff or other external people (e.g., physician) should be taken into account when allocating responsibilities for interventions. Moreover, the time interval between DSCCs could be scheduled differently to ensure that the responsible person has enough time to integrate the hypotheses and intervention into care planning or to consciously delegate this task.

The case-specific knowledge circulation process ended when the case was evaluated. The concept’s target of performing the evaluation at the beginning of the next DSCC did not prove viable because of incompatibility with the social system and the care routines [39]. Indeed, mixed-participant groups are not suitable for evaluating the outcomes of a case in which they were not involved. Therefore, future evaluations should take place internally, within the ward and with case-relevant staff. Team meetings or handovers have proven to be a way to evaluate the DSCC within the team. The visualization of the interventions in the wardroom combined with the request to indicate the success or failure of an intervention on the flipchart can also increase awareness of an evaluation and serve as a basis for discussion during the final evaluation of a DSCC within the team.