Barriers and facilitators of complex interventions in health services
Several factors have been identified in the literature as barriers and facilitators for implementing complex interventions as shown in Table 1 [39-51]. Common barriers across the studies included: education and training needs of staff, lack of leadership support, time constraints, complexity of intervention, lack of staff engagement and poor management and communication. Facilitators for implementing interventions included: sufficient resources, engagement of stakeholders, staff involvement and support from leaders and staff. Most of the barriers and facilitators cited in the literature were related to human factors and staffing issues rather than the type of intervention. The complexity of intervention was only cited in a few studies [39-51].
Human factors are by far the most complex to address when implementing interventions in health services as opposed to organisational factors [52]. Several key factors were identified from the literature. Understanding human behaviour, decision-making during critical situations and identifying sources of errors are key considerations and have the potential to influence the success or failure of project implementation. Moreover, effective teamwork requires cooperation, coordination and communication between the various team members [53]. Effective communication between and within teams enables cooperation and coordination. To ensure success every member of the team requires an understanding of the purpose, team roles, responsibilities, task requirements and the project plan. Trust in other team members and sharing information are also essential to enable cooperation between teams [52-56].
The synthesised model of implementation (the Triple C model) builds on these key human factors to enable successful implementation of complex interventions in health services as described in the next section [57].
The development of the Triple C model
The Triple C model proposes that to achieve successful implementation in health services requires attention to the social relations and processes that will result in outcomes. The three main elements of the Triple C model have been explained fully elsewhere [57].
It emphasises the processes by which complex interventions can be made practicable and embedded into daily clinical care by underscoring the significance of human factors. A three-stage implementation model was devised based on the findings of the literature and named the Triple C model (Consultation, Collaboration and Consolidation) as shown in Figure 1. The three stages of the model are in interconnected and overlap. As opposed to other models, this model considers the sustainability factor of any implementation project from the project outset and at each stage of the implementation process.
Sustainability of interventions are a key factor for continuous and successful change. Examples of sustainability strategies include: long-term action plans, tracking of program adoption, financial planning and mapping of the community settings where interventions take place. Sustainability has been defined in the literature as routinization, institutionalisation, durability, maintenance and long-term follow-up of an implementation [10,34,52,53]. Stirman and colleagues suggested that for an intervention to be sustainable certain core elements of the initial intervention must be displayed and maintained after the initial implementation of the intervention [58]. Moreover, most researchers have conceptualised implementation to be the last step of any implementation process. The Triple C model conceptualises sustainability as a set of processes that occur throughout the life cycle of any implementation process. The continuous consultation, collaboration and consolidation ensures that sustainability is not an end point but is a continuous process whereby the three stages are interconnected and overlap with each other to achieve sustainment [57].
Furthermore, health service research studies have increasingly recognised the value of adaptation in light of the everchanging context of health care services. Adaptability of an intervention to local context is necessary to ensure the usability and relevance of such interventions [59]. The relationship between adaptability and sustainability has been discussed in depth by Shelton et al., 2018. The authors highlight the importance of identifying barriers and facilitators to adaptability of an intervention to ensure its sustainability [10]. This is in addition to ensuring the core elements of the initial intervention are still maintained. The Triple C model allows for adaptability and sustainability through the continuous engagement of project stakeholders.
The consultation stage
The consultation stage is typically the initial stage of any implementation model and this is where all stakeholders are able to prioritise their workflow and initiate ideas and suggest areas for improvement. This stage should capture all the stakeholders’ priorities and map the pathway that will be taken to ensure the successful implementation of the project. A process map of the key steps involved in the project ensures a clear pathway of the project trajectory, areas for improvement and monitoring as shown in Table 2. To date, process mapping has not been used extensively in health care [60]. A study by Antonacci et al., 2018 highlighted the advantages of using process mapping for planning projects in health care [11]. The authors highlighted five key factors for successful process mapping including: appropriate and easy visual representation of the project; information collected from stakeholders; the ability of the facilitator to gather ideas from those involved and capture them on the map; knowledge of software and equipment used if needed and the ability to follow-up any missing steps or information throughout the process.
The collaboration stage
The collaboration stage aims at identifying who should be involved in the project based on their skills, knowledge and contribution to the overall project. This stage requires a high level of communication and openness between team members. Nystrom et al., 2018 highlighted the importance of collaboration on health projects from an interdisciplinary perspective to ensure the success of the project [12]. Several collaborative approaches can be used for successful implementation. These methods range from including higher degree students in the projects, clinicians having dual roles in the project as researchers and clinicians and involving staff from various levels of healthcare [12]. The challenges of successful collaboration include lack of clarity around roles and responsibilities in the project plan, organisational changes such as staff turnover, changing of policies or priorities and cultural differences amongst the project team [41,49-51]. On the other hand, enablers of good collaboration include established relationships, alignment of goals and priorities, skilled team members, clear communication, mutual trust and honesty between team members as shown in Table 2 [39,44,50,51].
The consolidation stage
This stage is the most crucial step in the model as it ensures the sustainability of the project and its incorporation into routine clinical care [3,24]. This stage may involve refinement of the initial ideas to ensure their successful adaptability to the local context while still leaving the core elements of the initial project unchanged. This stage is also done in each of the other earlier stages with the refinement process, consultation and collaboration are employed to ensure agreement about the project steps and its success. Consolidating successful interventions in a dynamic health service is challenging as it requires the use of a number of strategies, adapted to local context, that need to all work in sync. This process needs a few steps as follows; firstly, standardising policies and protocols across the health care setting to ensure minimal variability across departments; secondly, eliminating variances between policies and practices to ensure that processes are understood and orderly. Thirdly, having the right staff mix, with the appropriate skills and experience at all times; fourthly, having an idea about expected patient numbers that will benefit from the proposed intervention and ensuring that resources are available to meet any possible increase in numbers [3,24]. Finally, having access to business intelligence tools to continue and refine outcomes based on real time numbers is crucial for the success of this step as shown in Table 2.
All the elements discussed above on each of the three stages of the Triple C model have been mentioned by Proctor et al., 2015 to ensure sustainability of interventions [35]. The authors suggested that for sustainability of evidence-based interventions to occur, various factors need to be included in implementation models and these are training and funding, context, definitions and conceptualization and measurement and analysis. These factors have been captured by the above three stages through having the right skills mix of staff and resources, clarity of responsibilities, process mapping and having access to business intelligence tools respectively as detailed in Table 2 for each of the above stages [35].
Examples of using the Triple C model in health services research
The Triple C model has been used in several projects to verify its fit for purpose and to ensure its applicability to implementation science. This is the final stage of building a theory as stated above. The Triple C model was used in the successful implementation of an electronic wound care program across several organisations to track would healing and costs in rural Victoria in Australia [61,62]. The authors were able to show a significant improvement of wound healing times and decreasing dressing [62]. The researchers used the consultation stage to identify priorities areas of research and a map of operations for the project delivery. Sufficient resources were made possible by engaging multiple stakeholders early in the project through both in kind and financial resources. The collaboration stage was crucial as this project involved several organisations and training programs to ensure the successful delivery of the project across the multiple sites. A train-the-trainer program was devised to ensure efficient implementation of the program. The consolidation stage involved several steps such as standardising policies and procedures across all sites on management of wounds across the rural region and the establishment of a regional wound consultant role to oversee the project after its expiry. This was in addition to continuous data collection across the sites to ensure quality monitoring of wound healing and costs and its consistency with the initial plan of the project [61].
The second project where the Triple C was used was in the implementation of a medication safety program in an Aboriginal Health organisation in a large regional area in Australia [63]. This project employed a two-stage methodology. The first stage was consultation where interviews where conducted with staff and Aboriginal Health professionals to identify problems with medication issues in the Aboriginal community. The results from these interviews have informed a process map about the intervention to address the needs identified in the consultation stage [64]. The collaboration stage consisted of identifying the staff mix to deliver the intervention. In this case, it was a medication safety program consisted of staff training and development of policies addressing medication safety to be made available for all staff through an online platform. The consolidation stage of the project involved collation of data regarding satisfaction with the program and medications incidents [63,64].
Another project where the Triple C model was employed was the development of a skills matrix to identify areas of need to upskill palliative care nurses [65]. The overall objective of the project was to design and deliver educational programs that are relevant to the needs of palliative care staff across a large rural region involving several organisations. The project started with several consultation sessions addressing the training needs of staff involved in several organisations. Once the needs were identified, a process map regarding the delivery and implementation of the intervention was designed which included the development of a skills matrix to be used by managers for individual staff appraisals and identify their training needs and their levels of progress throughout the year, as the training occurs. The consolidation stage involved the use of this matrix as a standard form for staff appraisals and discussion about opportunities for future improvements [65].
A final example of a project using the Triple C model in practice was the implementation of the validated Distress Thermometer to improve identification, assessment and management of distress in the cancer care inpatient wards. This project was the result of a clinical incident and a root cause analysis recommendation. The setting for the project was the private and public inpatient oncology wards at a large tertiary referral hospital. Initially, the project involved consultation with all identified stakeholders to ensure by-in and identify areas for improvement across the public and private settings. Next, a procedure for management of patient distress in the oncology inpatient setting was developed in collaboration with multidisciplinary teams across oncology including; nurses, doctors and social workers. Staff were involved in the development of strategies that would be used to change practice including initiatives such as regular education sessions on identifying and managing distress for patients admitted to the cancer care inpatient wards and debrief sessions for staff where at-risk patients could be identified. The consolidation stage involved modifications to existing clinical documentation as well as the availability of a patient information brochure for patients and families to ensure sustainability of the changes [66].
Limitation of the model
While this model is adaptable to many contexts as shown above by the variety of projects, its utilisation will be limited by the resources available and time needed for each project. In addition, capacity and capability of staff involved in the implementation will influence successful use of the model. The availability of the business intelligence tools is also a limitation if electronic medica records are not embedded in the hospital system or if there is no method of collecting data automatically as this can be a very laborious task.