The DIvERT is designed for health professionals to intervene at the first sign of MSD and mobilise a timely response to manage the patient's MSD. This early intervention strategy is intended to ensure that the patient receives an appropriate and timely care plan and for the team to manage deterioration and prevent associated risks from escalating. DIvERT will be piloted in two selected clinical settings of an Academic Health Science Centre of a teaching hospital: a trauma unit and an acute surgical ward specialising in burn services; plastics; reconstructive surgery; and ear, nose, and throat surgery. DIvERT was commenced end of 2019 however, due COVID the pilot was stopped. The trial recommenced in 2021. Following implementation, if successful, DIvERT will form part of the hospital escalation response process for MSD, as illustrated in Figure 1.

The first response involves ward staff responding and attempting to manage the patient's deterioration. DIvERT is a second level response utilised when the patient's condition and symptoms become unmanageable for the ward staff. DIvERT involves a multidisciplinary team that provides timely and targeted interventions, assists with patient management, and makes recommendations to the ward team to address the patient's needs. The third level of escalation and response is Code Grey and the fourth level is Code Black.

Realist evaluation is a theory-driven evaluation methodology rooted in a realist philosophy of science, which emphasises an understanding of generative causation and how causal mechanisms are shaped and constrained by context (34). Generative causation refers to the idea that underlying mechanisms (causal forces) are triggered in particular contexts, which may not always be observable, and generate or lead to outcomes (35). The realist methodology is well suited to complex interventions that require multi-level organisational changes and involve human decisions and actions. The realist approach is intended to answer the questions, “what works, for whom, under what circumstances and how” instead of “does it work” (36, 37). In order to achieve this, it is necessary to uncover the architecture of programs, services, formal and informal resources, and efforts (38). R Pawson (39) argues that programs are incarnate theories; therefore, an implemented program is constantly testing theories about what might lead to changes.

The realist philosophical paradigm, realism, positions itself between the positivist and constructivist paradigms. Positivism primarily aligns with the hypothetico-deductive method to verify a priori hypotheses using quantitative approaches, wherein functional relationships are determined between causal and explanatory factors (40). Conversely, constructivism is aligned with qualitative research approaches, which emphasise the fact that reality is constructed through human interaction and action (41). Realism, thus, is the “middle ground”, which is the search for the alignment between reality and our constructions of reality (42). The realist ontology holds that reality exists independently of our perception of it and can, therefore, neither be observed nor perceived (43). Therefore, the realist understanding of the world recognises it as an open, complex, and nested system with different levels of stratified reality, the real domain, actual domain, and empirical domain (42, 43). Real refers to structures with causal powers and liabilities that enable or constrain actions, the actual refers to events or non-events that are enabled or constrained by the real domain, and the empirical refers to the observations and experiences that are a result of the actions and events in the actual domain (43). As well as acknowledging the social construction of reality, realists acknowledge that reality stems from a complex interplay between social, geographic, and historical context; hence, what is considered to be factual would likely change over time (39).

In realist methodology, we are interested in identifying and understanding people's responses to different resources offered within complex interventions (44). Rather than focusing on what can be seen at the surface level (empirical and actual domain), the realist approach is focused on understanding the complex layers that exist beneath the surface level (real domain) and explaining the reasoning behind human actions (42, 43). To do so, researchers must unpack the underlying generative mechanisms that explain “how” outcomes arise and the influence of context (35, 37). The realist driving concepts are retroduction, the search to unearth causal mechanisms (underpinning causal forces), and ontological depth. Thus, reality is stratified, the observable is a surface layer, and the causal features transpire at a deeper level (38, 45).

A significant advantage of a realist evaluation in enhancing evidence-based practice is its emphasis on detailed contextual analysis to provide explanatory insights into how a program functions (37). Furthermore, realist evaluations provide a valuable framework for conceptualising the underlying theories of a program. By examining and identifying the Context, Mechanism, and Outcome (CMO) configurations, researchers can better understand the interplay between contextual factors, mechanisms, and the resulting outcomes (intended and unintended). This is generally summarised as context + mechanism = outcome (42). The causal explanation resulting from this analysis will help the construction of a deeper understanding of why specific regularities occur and how they contribute to program outcomes. Table 2 provides a definition of context, mechanisms, and outcomes.

This evaluation will be reported in line with the RAMESES (Realist And Meta-narrative Evidence Syntheses: Evolving Standards) reporting standards for realist evaluations (46). Accordingly, we will follow the realist evaluation procedural steps as illustrated in Table 3.

Initial program theories (IPT) are a fundamental prerequisite of the evaluation methodology. The essential concept is, “If we do x, y will happen, because…,”. The “because” is critical to the structure; without it, it would just be a theory of action (48). “If… then” statements are a way to begin theorising, classifying, and organising the experiences and assumptions of the program designers and can lead to hypothetical CMO configurations (38, 49). IPTs can remain as “if…. then” statements, which become unpacked in the data analysis using CMO configurations. It has been suggested that CMOs can be developed concurrently either during the IPT development (for example, during the initial round of an evaluation, for testing during later stages) or retrospectively as a product of the evaluation rather than a design tool (48). As the most fundamental claims about an intervention, theories are seen as the starting and end point of the realist evaluation cycle (37).

IPTs are outlined in abstract terms and are concerned with identifying and explaining regularities (42). As pointed out, interventions are theories put into practice (50). Developing theories involves determining the logic of an intervention, deriving underlying assumptions about the effectiveness of the intervention, and explaining how it is expected to work. To construct our theories on a rapid response team model for managing MSD in patients, a comprehensive search plan will be developed, and a thorough scoping of published research and grey literature will be conducted. This approach ensures that a wide range of relevant information and evidence is gathered to inform the construction of initial theories. Incorporating published research as well as grey literature allows us to capture diverse perspectives and valuable insights that are not readily available in traditional academic publications. The search plan and scope will allow us to consider a variety of sources and contribute to a comprehensive understanding of the rapid response team model in managing patient MSD. In addition to conducting a literature review to gather relevant information, the process will be supplemented with creative brainstorming to develop the best explanation for how DIvERT is intended to work. As part of the second step of the realist evaluation cycle, initial theories will be tested and refined through realist synthesis (51).

In Step 2, we will complete a realist synthesis of relevant literature to test, refute, and refine our IPT. The synthesis will be and reported in line with the RAMESES (Realist And Meta-narrative Evidence Syntheses: Evolving Standards) publication standards for the realist synthesis (51). A realist synthesis is a theory-driven approach to synthesise the literature for empirical evidence to generate insights into complex interventions or programs and how they work in specific contexts (52). In contrast to traditional systematic reviews, realist synthesis attempts to gain a deeper understanding of the underlying mechanisms, contextual factors, and interactions that shape outcomes (51). Realist synthesis aim to answer not only the questions of “what works” but also “how, for whom, and under what circumstances” (42). Based on the realist synthesis interpretative, theory-driven approach, we will iteratively search for qualitative, quantitative, and descriptive literature to understand and unpack the underlying generative mechanisms (M) that explain “how” outcomes (O) are caused and the influence of context (C). We will follow the methodological steps outlined in Figure 2 to complete and report our synthesis based on RAMESES publication standards (51).

Refined program theories are intended to highlight the underlying generative mechanisms that contribute to the effectiveness of the rapid response team model in effectively addressing MSD. In addition, theories will consider the contextual factors that influence the implementation and effectiveness of the response model (53). The refined program theories resulting from the synthesis will be shared with the program designers within the organisation as part of the theory gleaning process. This collaborative step involves engaging with the program designers to examine and evaluate the refined theories to determine and prioritise theories for further testing in the next stage of the evaluation (51, 53). Program designers will review and provide input on the refined program theories during the theory gleaning process. Based on their expertise and knowledge of the intervention, they will evaluate whether theories are aligned with their initial assumptions and expectations (54). The objective is to gain deeper insights from program designer perspectives and determine which theories are most plausible and worthy of further testing. Through this collaborative engagement, the program designers can contribute valuable insights and identify potential additional contextual factors, mechanisms, or outcomes (53, 54). Discussions and feedback gained from the theory gleaning process will assist in refining the theories further, identifying additional contextual factors, mechanisms, and outcomes, and planning the design of data collection methodologies for the next phase (52, 54). Refined program theories are intended to highlight the underlying generative mechanisms that contribute to the effectiveness of the rapid response team model in effectively addressing MSD. In addition, theories will consider the contextual factors that influence the implementation and effectiveness of the response model (53).

As part of the realist evaluation methodology, researchers are guided by initial program theories (IPT) when selecting the appropriate instruments and measurement tools (48). As a basis for testing and refining the program theories, IPT's serve as a framework for identifying and collecting relevant data, such as interviews, observations, surveys, and document analyses (46, 55). The realist approach offers the advantage of integrating the strengths of both positivist and constructivist research paradigms. The positivist approach facilitates the collection of empirical evidence on program effectiveness, identifying patterns and quantifying the relationships between inputs, processes, and outcomes (34, 40). On the other hand, the constructivist approach provides rich contextual information, examines the perspectives and experiences of individuals and groups, and uncovers the underlying mechanisms and causal processes (34, 41). By embracing the complexity inherent in healthcare research, the realist methodology offers a sophisticated and nuanced understanding of research problems and questions, leading to comprehensive and insightful findings (56).

The DIvERT (De-scalation, Intervention, Early, Response, Team) intervention will be trialled in specific clinical settings within a tertiary teaching hospital in Melbourne, Australia. The clinical settings offer unique contexts that encompass differences in patient characteristics, presentations of MSD, and staff's experiences in managing patient deterioration. As part of the healthcare organisation's MSD monitoring and reporting process, electronic medical records, risk management systems, incident reporting systems, and security reporting systems are all used to collect MSD data. We will carry out a quantitative analysis of patient MSD data to gain a better understanding of the “what is going on” phenomenon by examining a range of variables, including but not limited to the patients' demographic characteristics, medical histories, factors contributing to MSD, how the MSD episodes were managed, and any interventions or treatments provided during the DIvERT responses. The analysis is also aimed to assess and establish the outcomes of patients who experience MSD, such stabilisation of MSD, length of stay in hospital, recurrent MSD episodes, and long-term management plans such as behavioural management plans for patients with complex MSD.

It is noted that, a current challenge faced by healthcare organisations is the lack of reliable systems for collecting quality data to inform continuous monitoring and improvement (17). Reporting MSD incidents serves as a foundation for promoting patient safety, facilitating data-driven decision-making such as supporting the implementation of interventions like DIvERT, advancing research efforts, ensuring regulatory compliance, and fostering a culture of continuous improvement in clinical practice (4, 57, 58). By complying with reporting requirements, healthcare systems establish accountability and promote transparency, ultimately enhancing the overall quality of care provided.

The organisation uses multiple incidents reporting databases, including Code Grey, DIvERT, electronic incident reporting forms, and security reports. These incident reports serve as the basis for testing different theories, such as theories relating to the prompt intervention, multidisciplinary collaboration, early warning signs recognition, tailored interventions, and clinical skills development. By analysing incident reports within these theoretical frameworks, we hope to gain insights into the effectiveness of DIvERT and understand how different factors, such as timing of interventions, collaboration between response team and ward staff, recognition of warning signs, and person centred approaches to managing MSD, contribute to the effectiveness of DIvERT.

Incident analysis is undertaken to uncover the key factors influencing outcomes, including effective and timely intervention preventing MSD risks from escalating to Code Grey, positive patient and staff experience, and positive organisational culture. These incident reporting systems capture comprehensive data, including demographic details, incident specifics, contributing factors, and immediate actions taken. Additionally, they are part of the organisation's risk management and clinical governance efforts, providing data on interventions, response outcomes, and follow-up plans for continuous improvement in patient and staff safety.

We will use electronic medical records to collate data missing from DIvERT, Code Grey, and security data records. Electronic health records will provide valuable insights into how the response team addresses individual patient characteristics and needs in managing MSD and the outcomes of patients presenting with MSD. In addition, the records will provide insights into how the response team considers and manages social, environmental, and other factors that may affect the patient's mental state. The evaluation is also aimed to test theories related to shared decision-making and handover, including patient involvement, and the provision of resources to support staff in caring for patients following an episode of deterioration.

We intend to survey clinical staff who work in the medical units where DIvERT is being trialled. The survey will enable the collection of self-reported data on variables such as training, confidence levels, and factors affecting communication, assessment, and documentation, providing insight into the perspectives and experiences of nurses. Additionally, we are interested in assessing staff ability to recognise early warning signs of patient MSD and understanding through free text responses, the contextual factors of what should be in place for DIvERT to be effective. Using the survey as a theory confirmatory approach, the objective is to empirically test, refine, and validate the underlying assumptions and causal mechanisms for evidence that best explains how an intervention works (34, 59). This ensures that the theories are grounded in empirical evidence and can inform effective decision-making and program improvement. The survey methodology is preferred for its systematic and efficient nature, which allows for the collection of valuable data from the participants (60).

Based on IPT's we will develop a survey instrument in line with the principles and criteria for designing survey questionnaires to test and refine our theories (61). Survey data will be collected and managed using REDCap electronic data capture tools hosted by the healthcare organisation. Additionally, we will provide a printed copy of the survey instrument for staff to complete to increase uptake. We will complete a pilot test of the survey instrument with select nurses to ensure consistency, clarity, and readability, as well as to check for technical issues, evaluate response options, assess question relevance, and ensure clear instructions.

A Microsoft Excel spreadsheet will be designed to collect and analyse the variables that have been previously discussed to test and refine our theories. In line with the realist evaluation approach, the data collection methods employed will be comprehensive enough to capture both intended and unintended outcomes and the contextual factors and mechanisms that influenced them (46). Our objective is to audit retrospective data for a minimum period of 6 months which should allow for the thorough analysis of MSD trends, interventions, patterns, and outcomes. However, if the sample size is deemed insufficient, we will extend the data collection period beyond 6 months to ensure rich enough data and an adequate sample size for analysis.

Qualitative data collection methods, such as interviews and field observations, will offer valuable insights into the underlying mechanisms, contextual factors, and staff experiences that contribute to the effectiveness of DIvERT (37). By engaging with clinical staff in the selected settings and observing real-time interactions, qualitative data collection enables a deeper understanding of how DIvERT operates, how interventions are implemented, and how they impact patient outcomes. Qualitative data captures rich narratives, perspectives, and nuances that quantitative data alone may not fully capture. Qualitative data collection methods provide a comprehensive and holistic view of the complex dynamics involved, helping to uncover the “how” and “why” behind the outcomes observed, and informing recommendations for improving DIvERT (42).

The purpose of the theory refinement interviews is to facilitate direct engagement with clinical staff and gather firsthand accounts, thoughts, feelings, and motivations, uncovering underlying beliefs, assumptions, and values that shape their experiences and decision-making processes regarding DIvERT. Furthermore, interviews will allow participants to share their feedback, suggestions, and concerns regarding DIvERT. According to Manzano (54), the semi-structured interviews should gather the interviewees’ stories about the program because those experiences can provide insights into the program's varying processes (mechanisms, contexts, and outcomes). This teacher-learner cycle's approach should be adopted with theories placed before the respondents to comment on, confirm, deny, and refine the theories (49). An interview guide will be developed to guide the interviews. Table 4 includes examples of some of the questions we intend to ask staff to confirm, disconfirm, and refine our IPT.

Clinical staff members participating in the DIvERT pilot program, including nurses, nurse managers, and allied health staff such as social workers, psychologist, and DIvERT team members will be interviewed. Interviews will be conducted face-to-face or virtually via the Microsoft Teams platform and recorded using a recording device. Manzano (54) suggests that realist inquiries should strive to collect a substantial amount of data. Using the purposive sampling technique, we will select participants based on their potential to contribute to refining the initial program theories (47).

For this evaluation, we anticipate conducting between 20 and 30 interviews, as suggested by M Mason (62). However, we will use saturation as our guiding principle during data collection. Saturation is achieved when additional data no longer yield new insights or information relevant to the research themes. In our study, while interviews will test predefined program theories, we will remain open to new emerging themes. We will continually code and analyse the data, closely monitoring theme development as more interviews are conducted. Regular reviews of themes and codes will be performed to ensure both depth and breadth in the data collected. Saturation will be considered reached once the data begin to repeat, with no new information contributing to the data collection process.

Observational research methods are essential for understanding people's actions, roles and behaviours in their natural environments (63). Field observations will involve gathering rich contextual data that complements interviews, surveys, and document analysis. The field observations are aimed to understand the “what is going on” phenomena (64, 65). The researcher will attend DIvERT activations on the medical wards as an observer to see how DIvERT functions in real time and to note the interaction patterns among healthcare professionals, including how they communicate and coordinate to meet MSD needs. Additionally, we will track the timeliness of responses, from the initial call to the team's arrival and their subsequent interventions. We will also monitor immediate patient outcomes post-DIvERT to observe any changes in the patient's mental state. Furthermore, we will identify challenges and barriers that impact the effective functioning of DIvERT.

The researcher will take field notes to document the events and activities that occur during the response calls. The data gathered through field observations can provide valuable insights for refining the DIvERT model, identifying areas for improvement, and informing recommendations for enhancing the management of patient MSD incidents. We aim to complete sufficient, but not exhaustive, field observations to gather data to test and refine the theories. We will follow the principles of reasonable judgment and determine the point of reaching contextual richness (51). Once it is determined that attending additional field observations will not yield new information, we will conclude the field observations.

In realist methodology, the purpose of data analysis is to uncover the underlying mechanisms and contextual factors that influence program outcomes through a confirmatory theory building theory-building approach (42, 46). To draw inferences from the evidence, the primary approach of constructing confirmatory theories, particularly within realist research, is retroductive reasoning (or abductive reasoning) (66). Analysis involves cross-examining program theories with collected data and using theories to understand configurations in data (34). Data analysis is aimed not merely to summarise and report themes in the data, nor is the approach designed to aggregate data by strength in numbers philosophy. Rather, it involves data analysis to improve an ontologically deep understanding of the phenomena being studied (38).

Microsoft Excel and STATA (Statistics and Data) will be used to process, organise, and analyse quantitative data. The main objectives of the quantitative data analysis are to: 1.

Summarise and present data using descriptive statistics and display data as frequency tables, and diagrams.

The qualitative data will be managed using NVivo. NVivo is a software program used for qualitative and mixed-methods research. NVivo helps qualitative researchers to organise, analyse and find insights in unstructured or qualitative data where deep levels of analysis are required (67). For realist approaches, NVivo offers a framework to structure and organise the iterative process of generating, refining, and testing complex program theories, especially when drawing on multiple data sources simultaneously (68). This is particularly beneficial given the multifaceted and dynamic nature of realist data (68). Qualitative data will be transcribed, organised, and labelled using Microsoft Word.

There are different ways of coding and reporting data, holistic coding, referential coding, and axial coding (38). For holistic coding, a single code is applied to a large unit of data and is most applicable when the researcher has a general idea of what to investigate in the data. In referential coding, researchers identify and code specific phrases in the data, focusing on instances where participants explicitly or implicitly refer to certain concepts, themes, and ideas. Thirdly, axial coding involves the breakdown and categorisation of data into specific themes, concepts, or categories based on their relationships (69). By exploring Context-Mechanism-Outcome relationships allows for a more in-depth analysis of how various elements interrelate and enables the identification of patterns that contribute to improved conceptual clarity. This evaluation will be based on the latter approach.

Transcribing and coding will be piloted following a few interviews to clarify the approach. The analytical process in realist methodology is iterative, meaning that coding and analysis can be started as data are collected (38). For interview, data analysis will be based on thematic analysis and survey data, we will adopt utilise content analysis. Using data to support, refute, and refine program theories, we will primarily focus on extrapolating the lines of association (46). The results will support inferences about varied influences operating within a specific CMO configurations. Labelled CMO configurations will attempt to explain how and why various mechanisms are activated or not in different contexts to generate different outcomes (34). CMO configurations are about specific characteristics of the program, and when considered as a whole, they become the product of the complete review (70). As new theories supporting causal insights emerge during data analysis, they will be articulated and labelled accordingly. Regular fortnightly meetings with academic supervisors will be held throughout the research process to check and validate analysis process, review overall progress, and address challenges.

(Present configured contexts and mechanisms into refined program theories).

In this phase the refined and evidence-informed theories to support “what works, for whom, under what circumstances, how and what did not work” about the DIvERT intervention will be presented (42). Refined theories will be presented with corresponding CMO configurations. Refined theories will guide future DIvERT program design and implementation. As pointed out, propositions for policy and practice should be explained clearly, follow logically from the analysis, and recommendations should account for strategies for different contexts (46). It is prudent for realist evaluations to be modest about evaluation claims, as evaluation results cannot be universally applicable. Rather, results should provide plausible explanations for the complex processes underlying programs (42).

We will engage with key stakeholders in the final stages of the realist evaluation, an important step that serves dual objectives. Firstly, engagement provides a means to validate the research findings, aligning them with the practical experiences and perspectives of those directly involved. This validation not only improves the credibility of the evaluation but also ensures that the results align with the stakeholders’ intentions of DIvERT. Secondly, engaging stakeholders facilitates a dynamic exchange of insights, enriching the understanding of contextual nuances that may influence the success of DIvERT. The collaborative process improves the applicability of the evaluation's recommendations, fostering stakeholder buy-in and promoting the effective implementation of evidence-based strategies in other clinical settings. As per reporting standards, strengths and limitations of the methodology, utility, and analysis will be reported (46).