The larger project employed a multi-method, multi-phase research design within the pragmatic paradigm. This philosophical paradigm allows the researcher to focus on “what works” and provides solutions for problems utilising methods that best meet their needs and purposes [17]. The part of the project reported here employed a descriptive qualitative method [18] and, as such, the Standards for Reporting Qualitative Research (SRQR) checklist [19] was used to ensure accurate and complete reporting of the study. As this was an implementation study the StaRI checklist [20] was also used to ensure accurate reporting of the implementation elements of the study. Staff from two hospitals were interviewed, nine to 15 months after implementation commenced. Hospital A was a large hospital located in the tropical north of Australia. Hospital B was a medium sized hospital west of Brisbane, Queensland Australia. The management of both hospitals had agreed to trial the implementation of the GEDI model in their respective EDs.
Innovation to be Translated
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 [12, 21]. 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 to assist in setting up and successfully employing this approach to care [22].
The Approach to Knowledge Translation Used in This Study
The implementation of the GEDI model in two Queensland EDs was based on the i-PARIHS approach to facilitation [23] and a Cochrane review that provides evidence of the importance of tailoring interventions to the context [24]. As a first step the project team developed a GEDI Implementation Toolkit [22] 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) [23]. In this project these 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 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. Funding from the Clinical Excellence Division of Queensland Health was provided to their employer 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 [23]. In this approach to implementation the IFs interact and connect with local staff and the EFs. 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].
Evaluation of GEDI Implementation
Semi-structured, audio recorded, interviews were conducted with a range of staff at the two implementation sites, by author (inserted initials removed for blinding), who has PhD and post-doctoral training in qualitative interviewing. The author (inserted initials removed for blinding) 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 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.
Data Analysis
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 (inserted initials removed for blinding – a different person to the interviewer). 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 implementation 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 XX (inserted initials removed for blinding) and illustrative quotes had been presented in support of the sub-themes and themes, all authors discussed the findings. As XX (inserted initials removed for blinding) and XX (inserted initials removed for blinding) were the external facilitators and thus heavily invested in the project, the other authors XX and XX (inserted initials removed for blinding) ensured that any emergent themes that were challenged by (inserted initials removed for blinding) XX and XX 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.
Findings
In total 17 interviews were recorded with staff in Hospital A and 12 in Hospital B, in addition the two external facilitators were also interviewed. The roles of the people interviewed included middle managers and senior GEDI staff involved in setting up the model of care, middle managers involved in the on-going management of the model of care, GEDI clinicians (medical, nursing and pharmacy), ED clinical staff and the external and internal facilitators involved in the implementation process.
Generally, all interviewees expressed support for the GEDI model, and the toolkit developed during the initial research. A key function of the GEDI team is to support the ED primary care team to make decisions about whether a frail, older adult needs to be admitted to the hospital, or whether they can be discharged safely home, and access care in the community. As an ED nurse said,
“I think (the GEDI model) helps flow... Either it’s, “You’re probably going to be discharged. GEDI have already been in and worked that out. Your daughter is on the way.” That happens quickly, or what I’ll see is GEDI have come in and found a problem, and had it looked at and realised (going) home’s not going to work. They’re going to need a referral and (…) that happens earlier as a result of more investigation or more history taking on their part. And I think that definitely (improves) flow because we’ll arrive at that decision much earlier to refer rather than discharge.” (ED Nurse: Hospital A)
Analysis of the data identified three major themes. There were elements of the GEDI model (as detailed in the Toolkit) that had either been: (i) adopted or had been (ii) adapted by each site and altered in ways not suggested by the Toolkit. As one GEDI CNC said, ‘Yep, great we’ve got the GEDI; but you could adapt it.’ (GEDI Nurse: Hospital B). In addition, there were data that related to the (iii) factors that affected the implementation of the GEDI model.
Adoption and adaptation of the GEDI model
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 got 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 External Facilitators (EFs) and Internal Facilitators (IFs) and a structured program of visits by the staff from implementation sites to the development (host) site and vice versa. Then a range of formal teleconferenced meetings were organised. Attendees for these meeting were to be all EFs, IFs, ED managers and all GEDI staff. The EFs were also to provide support by phone to IFs, as required for up to 12 months. Funding was provided to the original hospital that developed the GEDI model, by the State Health Department, to cover backfilling the EFs, all travel and accommodation costs and support for data collection and analysis of the implementation sites. As can be seen from Table 2, what was planned 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. 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)