Establishing the Geriatric Emergency Department Intervention in Queensland emergency departments: a qualitative implementation study using the i-PARIHS model

GEDI is a nurse-led, physician-championed innovative model of care that aims to improve outcomes for frail older persons presenting to the ED. The GEDI nurses are advanced practitioners who have additional experience and education in gerontology and care of frail, older people. They work with the primary care ED nurses and ED physicians providing targeted geriatric assessment, multi-disciplinary shared decision-making and coordination of care to facilitate rapid access and coordination of care through ED, hospital and community services. The details of this model of care are presented elsewhere [10, 12, 22]. Critical to the integration of the GEDI model into the ED is the role the ED physician plays in driving acceptance, policy change, and clinical support to overcome barriers to implementation. An extensive toolkit was developed, as part of this project, to assist in setting up and successfully employing this approach to care [23].

The implementation of the GEDI model in two Queensland EDs was based on the i-PARIHS approach to facilitation [16] and a Cochrane review that provides evidence of the importance of tailoring interventions to the context [24]. The i-PARIHS framework rationalises implementation of research into the clinical practice settings through implementation context, facilitation as integral to success and the recipients of change [16]. As a first step the project team developed a GEDI Implementation Toolkit [23] which was tailored to the state government policies and procedures. In line with the i-PARIHS model, the implementation process was then managed procedurally by two layers of facilitators, external and internal. The external facilitators (EFs) consisted of the ED physician and senior GEDI nurse who had been involved in the initial GEDI trial. According to i-PARIHS, the EF is a knowledge broker, linking the knowledge producers (i.e., clinical and research team) to the recipients or knowledge users (i.e., hospital ED staff) [16]. In this project these EF clinicians had skills in knowledge translation, change management, negotiation and influencing, and their activities included mentoring, coaching and guiding the internal facilitators. They also developed resources related to facilitation. These resources included a web-based toolkit, and short video vignettes for use by the internal facilitators where ‘tricks of the trade’ were shared.

It was planned that these two EF clinicians would host visits, for the internal facilitators (IFs), at the ED where the GEDI model was successfully implemented and visit both new implementation sites a number of times before and during the implementation. They would then have regular telephone or videoconference calls with the implementation sites (see Fig. 1). Funding from the Corporate Division of Queensland Health, the state-wide public hospital and health service provider, was given to their public hospital to release them from other duties and allow them to engage fully in supporting the implementation sites. At each implementation site there were two IFs who were the local champions. The IFs (ED physicians and senior nurses at each site) reflect the role of boundary spanners in i-PARIHS [16]. In this approach to implementation the IFs interact and connect with local staff and the EFs (See Fig. 1).

The IFs bring content expertise, related to care of older adults in the ED, contextual knowledge of the hospital system and how to navigate local hospital processes. It was planned that the IFs would receive guidance from the EFs to develop and apply skills in knowledge translation and change management. The IFs were the expert clinicians who would manage the program. It was planned that the IFs would present the toolkit to staff and establish a local support program, enabling local staff to share their learning about what worked and what did not work in the local context. They would also work with other GEDI and ED staff to undertake an environmental scan and then develop an action plan to maximise enablers and overcome barriers to implementing GEDI [25].

Semi-structured, audio recorded, interviews were conducted with a range of staff at the two implementation sites, by author AC, who has PhD and post-doctoral training in qualitative interviewing. The author AC was not known to the interviewees, prior to the interviews but had worked as the Research Fellow/Project Manager on the original study evaluating the GEDI model.

A purposive sample of relevant medical and nursing staff was contacted via email. Emails inviting staff to participate in the study were sent to the IFs and EFs, middle managers in ED involved in the implementation, individuals who were appointed to GEDI roles and frontline ED clinicians. Study information was sent out by the administrative assistant of the ED. Once individuals had indicated a willingness to be interviewed, they were contacted by the interviewer. Informed consent was obtained from interested staff and interviews were scheduled at a time and place suitable to that staff member. All staff members elected to be interviewed in a private space in their workplace and all interviews took between 15 and 60 minutes.

The interviewer employed a range of different questions and prompts for each of the groups of participants. Each participant was asked to explain their role and how they were involved in or interacted with the GEDI model. Then participants from each group were asked to reflect on the implementation process, how the GEDI model impacted their role and how it influenced workplace practices in the ED. Please see Additional File 1 (Supplementary Materials) for a copy of the interview schedule.

Transcribed interviews were read and re-read, and an initial label (code) was assigned to sections of text relating to the adoption or adaptation of different aspects of the GEDI model by the sites; and the factors influencing implementation at each site. Codes were then amalgamated into categories and themes [26]. This initial coding was undertaken by MW. This researcher is a very experienced researcher who has both doctoral and post-doctoral training and experience in qualitative research methods and has taught these methods for over 30 years.

The i-PARIHS framework conceptualises successful implementation (SI) as involving facilitation (Facⁿ) that addresses the innovation in practice (I), the recipients of the innovation (R), and the quality of the context (C) [SI = Facⁿ(I + R + C)] [27]. Consequently, in addition to the coding and theme development described above, the qualitative data were explored in terms of innovation (and evidence), recipients and context and how these elements were impacted by facilitation.

Once the initial analysis had been undertaken by MW and illustrative quotes had been presented in support of the sub-themes and themes, all authors discussed the findings. As EM and AT were the external facilitators and thus heavily invested in the project, the other authors MW and AC ensured that any emergent themes that were challenged by EM and AT could be supported by interview data from study participants. The study findings were not able to be returned to participants as most participants had left their positions and were not contactable once analysis was undertaken.

There were five sub-themes, each with a number of categories, identified that related to the elements of the GEDI model that were adopted and/or adapted for use in the study EDs. These categories were: Team Structure, Service Focus, Organisation and Funding of the GEDI model, Staff Education and Data Collection for Service Evaluation. Table 1 provides these sub-themes and categories along with exemplar quotes from the study participants.

The GEDI model specified a team approach, in which nurses with expertise in the management of frail older adults, especially in community settings, were upskilled in ED nursing and assisted the ED teams to make disposition decisions. These GEDI nurses were supported by a Physician Champion. This senior medical officer had additional training in the care of older adults and medical management of geriatric syndromes. This championing role extended beyond the change management process and meant that the GEDI model had a champion in senior medical forums and in management decision making and budget meetings. While the study sites, initially, understood why these roles were set up this way they all adapted the model. Most of the nurses were ED nurses who expressed an interest in caring for older adults but were expected to upgrade their knowledge and skills in their own time, using their own resources. Sometimes this meant they reverted to the role of providing nursing care in the ED and were diverted from assisting with disposition decision making. In addition, the implementation sites added geriatricians, physiotherapists and/or pharmacists to the team.

The Implementation project was structured around the i-PARIHS model with EFs and IFs, a structured program of visits/teleconferences by the staff from implementation sites to the development (host) site and vice versa and funding to support these activities. As can be seen from Table 2, what was planned (see Fig. 1) did not eventuate as the senior ED management in the host hospital felt they could not support releasing the EFs to allow them to facilitate at the implementation sites. The differences between what was adopted from the model and what was adapted were sometimes seen to be really useful e.g., adding a pharmacist or physiotherapist to the team. Other adaptations were less successful e.g. employing ED nurses with no experience of community/gerontology nursing as GEDI nurses. The breakdown in the implementation plan meant that adaptions were not challenged and the evidence for changes to the model was not established. However, as one participant put it, “I think actually since this program started it has highlighted to other (ED) staff, (…) that we should be learning more about dementia. (…) There was a dementia/delirium workshop recently and a lot of them applied for it because they’re interested in learning more. (…) It’s a good thing. It’s really highlighted how we wish we could and should be doing a lot better for geriatrics in emergency departments.” (ED NUM: Hosp A/1).

For the qualitative interviews, all undertaken by one person, all the relevant people involved in the implementation phase were included, until data saturation occurred, in line with qualitative methods. Comments in the Results, however, indicated that following the exit of some staff the implementation process changed for those remaining and for those taking over from them. This was only a small number of people over the 12 months of the study but may have influenced the results. This is a known issue in the sustainability of change [41]. In addition, while every effort was made to ensure the data analysis was a true representation of the implementation, the inclusion of the facilitators in data analysis may have had unknown impact on the overall outcomes.