Making change last? Exploring the value of sustainability approaches in healthcare: a scoping review

Background Numerous models, tools and frameworks have been produced to improve the sustainability of evidence-based interventions. Due to the vast number available, choosing the most appropriate one is increasingly difficult for researchers and practitioners. To understand the value of such approaches, evidence warranting their use is needed. However, there is limited understanding of how sustainability approaches have been used and how they have impacted research or practice. This review aims to consolidate evidence on the application and impact of sustainability approaches in healthcare settings. Methods A systematic scoping review was designed to search for peer-reviewed publications detailing the use of sustainability approaches in practice. A 5-stage framework for scoping reviews directed the search strategy, and quality assessment was performed using the Mixed Method Appraisal Tool. Searches were performed through electronic citation tracking and snowballing of references. Articles were obtained through Web of Science, PubMed and Google Scholar. Six outcome variables for sustainability were explored to ascertain impact of approaches. Results This review includes 68 articles demonstrating the application of sustainability approaches in practice. Results show an increase in the use of sustainability approaches in peer-reviewed studies. Approaches have been applied across a range of healthcare settings, including primary, secondary, tertiary and community healthcare. Approaches are used for five main purposes, namely analysis, evaluation, guidance, assessment and planning. Results outline benefits (e.g. improved conceptualisation of sustainability constructs and improved ability to interpret sustainability data) and challenges (e.g. issues with approach constructs and difficulty in application) associated with using a sustainability approach in practice. Few articles (14/68) reported the sustainability outcome variables explored; therefore, the impact of approaches on sustainability remains unclear. Additional sustainability outcome variables reported in retrieved articles are discussed. Conclusions This review provides practitioners and researchers with a consolidated evidence base on sustainability approaches. Findings highlight the remaining gaps in the literature and emphasise the need for improved rigour and reporting of sustainability approaches in research studies. To guide future assessment and study of sustainability in healthcare settings an updated list of sustainability outcome variables is proposed. Trial Registration This review was registered on the PROSPERO database CRD 42016040081 in June 2016.


(Continued from previous page)
Trial Registration: This review was registered on the PROSPERO database CRD 42016040081 in June 2016.
Keywords: Sustainability, tools, frameworks, models, healthcare improvement, sustainability outcomes Background Healthcare research continuously produces new and innovative findings but there is evidence that these improvements are often forgotten, changed or replaced in practice long before they can reach wider populations [1][2][3][4][5][6]. Therefore, the sustainability of evidence-based interventions has been identified as one of the most critical gaps in implementation science [3,7,8].
In healthcare, the term sustainability often refers to broad notions of continuation or permanence over time but the concept of what will be sustained is diverse [1,9,10]. While there is currently no consensus in the field on which definition of sustainability to operationalise and advance, for this paper, the comprehensive definition developed by Moore et al. will be used, namely that: "after a defined period of time, a program, clinical intervention, and/or implementation strategies continue to be delivered and/or individual behaviour change is maintained; the program and individual behaviour change may evolve or adapt while continuing to produce benefits for individuals/systems" [11]. This definition allows sustainability to be viewed as both an 'outcome' where health benefits and intervention activities are maintained, as well as an ongoing 'process', recognising the need to respond and adapt to promote continuation of improved practices, benefits or outcomes [12].
Rising demands and competition for resources have highlighted the need to understand if and how the sustainability of implemented initiatives can be influenced to support the long-term impact of investments [3,4]. Sustaining improved outcomes and processes is a recognised challenge to healthcare staff and stakeholders. Numerous definitions and determinants of sustainability make it exceedingly complicated for researchers and practitioners to understand or influence sustainability in practice [3,5,[13][14][15][16]. To address the challenge of conceptualising, measuring and influencing sustainability, many researchers and healthcare practitioners have developed sustainability frameworks, models and tools to assess sustainability in their initiatives or settings [17][18][19][20][21][22][23][24][25][26]. (The term 'sustainability approaches' will be used throughout this paper to refer to sustainability models, frameworks, tools, checklists, processes, strategies and conceptualisations [27]).
A 2018 systematic review identified 62 sustainability approaches designed for use in healthcare settings [27]. The review found that sustainability approaches may offer a means of connecting findings across diverse sustainability studies by providing explicit definitions and foundations for measurement but it also concluded that further research was needed to understand the application of these approaches in practice to assess their potential impact on sustainability processes and outcomes [17,19,27]. Although there is evidence of mounting interest in using sustainability approaches, little work has been conducted to understand how approaches have been used and how they have impacted research or practice [28]. This has resulted in seemingly random selection of approaches across research studies driven by 'convenience or prior exposure' [29][30][31]. This issue highlights an ongoing gap in knowledge between the development of frameworks, models or tools and the understanding of how they are used in and impact research and practice [29].
In order to understand the value of approaches such as models and frameworks, researchers and practitioners require supporting evidence to warrant their use [32]. Clear descriptions of how approaches inform planning and conduct of sustainability research, along with the potential challenges of use, would provide insight into which approaches have demonstrated positive impact for different programmes and settings [3,7,19,27,33,34]. Combining this evidence across sustainability approaches may also support further theory development to create improved frameworks and models for sustainability [19,35,36].
In the available literature, it is also not currently known if or how sustainability approaches are contributing to sustainability outcomes or improving sustainability reporting. Proctor et al. defined 'sustainability outcomes' as the "subsequent impact (healthcare improvement or public health outcomes) of sustained intervention use" [37]. To understand the impact of the approaches, more rigorous measurement and conceptualisation of sustainability outcomes is needed for the 'methodological objectivity and interpretability' of findings [3,17]. Scheirer and Dearing suggest six sustainability outcome variables [17], namely (1) continuation of benefits or outcomes for consumers, clients or patients; (2) continuation of programme activities or components of the original intervention; (3) maintenance of partnerships developed during the funded programme; (4) maintenance of new organisational practices, procedures and policies; (5) sustaining attention to the issue or problem; and (6) programme diffusion and replication in other sites.
These variables provide a quantifiable way to understand the success of a project or intervention [17,36]. Assessing the impact of sustainability approaches on these variables will support a nuanced understanding of how sustainability approaches contribute to sustainability outcomes, aiding researchers and practitioners to critically appraise their value to research and implementation efforts.

Aims and research questions
This review aims to consolidate the current evidence on the application and impact of sustainability approaches in healthcare settings, including the benefits and challenges that arise through their use, and explore the sustainability outcome variables they have supported. Two research questions will be explored -(1) What sustainability approaches have demonstrated application in healthcare practice within peer-reviewed publications? (2) What is the value of using a sustainability approach? What are the documented benefits and challenges associated with the use of sustainability approaches? What impact have approaches shown on sustainability outcomes variables?

Methods
A previous systematic review identified 62 papers that described sustainability approaches created for use in healthcare settings [27]. The sustainability approaches come in a variety of forms, including 32 frameworks, 16 models, 8 tools, 4 guidance strategies, 1 checklist and 1 process [27]. Full details on the sustainability approaches, including their constructs, scoring mechanisms and development, can be found here (https://doi. org/10.1186/s13012-017-0707-4) [27]. This scoping review builds on these findings to explore the application and impact of these sustainability approaches in healthcare settings.
A systematic scoping review was designed to explore the available evidence on the application and impact of sustainability approaches in healthcare. To ensure overall quality and coverage of the literature, authors developed the search strategy in collaboration with a medical librarian at their academic institution. A scoping review was deemed to be the appropriate method as it allowed for examination of the "extent, range and nature of research activity" in the use and impact of sustainability approaches [38]. This method specifically supported authors to track the application of the 62 sustainability approaches through citation tracking and snowballing (the 62 approach articles which were tracked through the literature are listed in Additional file 1) [27].
To ensure rigorous documentation of methods, Arksey and O'Malley's five-stage framework for scoping reviews was followed and is described below [38]. PRISMA guidelines were also used to ensure clarity of reporting [39,40]. The PRISMA extension for Scoping Reviews Checklist can be found in Additional file 2.

Research question(s)
The research questions were established through discussion between authors after consolidating the results from the previous systematic review, which demonstrated gaps in the current literature regarding the application and impact of sustainability approaches in healthcare settings.

Relevant studies
Citation tracking and snowballing for each of the 62 approaches was used to determine their use in peerreviewed publications [41]. The search was carried out between May and August 2018, with citation tracking and snowballing taking place on Web of Science and Google Scholar. This was done by putting the title and author of the sustainability approach paper into the database and selecting the 'times cited' or 'cited by' icon. Once citations were tracked for all 62 papers, we moved onto the next stage of the methodology consisting of screening the titles and abstracts.

Data collection process and study selection
We sought peer-reviewed publications detailing the application of a sustainability approach within healthcare settings. The level of use and rationale for application were not specified but could include assessment, planning, evaluation, monitoring, prediction or testing of sustainability. Papers published in peer-reviewed journals introducing a description of the approach used as well as clear information about application were included. Papers with vague or unclear descriptions on the approach used or application were excluded. For example, papers mentioning or introducing a sustainability approach but not applying it within their studies were excluded. Papers published in languages other than English were excluded. Commentaries, research protocols, conference proceedings, editorials and perspectives were excluded. Two screening steps were undertaken. First, the selected papers went through a rapid title and abstract screening; if a sustainability approach was not mentioned in the abstract, a rapid full-text screening was conducted to confirm whether the approach was used or not within the study. Full-text screening was then conducted by two authors (LL and AL) for all eligible papers.
Data extraction, data items and data synthesis A data extraction table was designed to record the key components and information from the retrieved papers. This method allowed for the data to be sorted into themes and categories promoting structured analysis and interpretation [38]. Data extraction included author(s), journal, year of publication, country of study, study methodology, sustainability approach used, purpose of use, description of use, healthcare setting, timeframe, measurement, documented benefits and challenges, and reported sustainability outcome variables.
Data was independently extracted by one author (LL) on an excel spreadsheet. A second author (AL) then read through and verified the extracted information from the original full text articles.

Quality appraisal
Although scoping reviews do not consider appraisal tools mandatory, the lack of clarity and rigour of much of the sustainability research prompted the use of a quality assessment tool within this study [17,38,42]. To establish study quality, the Mixed Method Appraisal Tool was used [43], which is divided into two sections. The first screening step is to ensure non-empirical studies such as reviews and theoretical papers are discounted for the analysis [44]. The screening questions are then followed by specific questions based on study methodology [43].

Collating, summarising and reporting the results
Following the quality assessment, data extraction categories and sustainability outcomes were compared and consolidated across the included studies. Results have been summarised using descriptive statistics and narrative summaries to show the studies' characteristics, purpose of use, the benefits and challenges of use as well as impact on sustainability outcome variables.

Results
The electronic citation tracking of the previously identified sustainability approaches identified 3119 publications for potential inclusion. Additional file 1 includes full information on how each of the 62 approaches were tracked through the literature. It lists the number of citations for each sustainability approach along with the number of full-text articles retrieved and included in this review. This process identified 109 eligible articles which underwent full-text screening. Following quality assessment, 41 articles were excluded based on inclusion criteria. Details on inclusions and exclusion of full text papers are included in Additional file 3. This resulted in 68 articles that were included in this review (Fig. 1). Table 1 summarises the included papers. Full data extraction (including study design, sustainability approach applied, documented benefits and challenges of use, and reported sustainability outcome variables) for the 68 articles can be found in Additional file 4.
Results are presented in three sections. The first describes the application of the sustainability approaches in healthcare settings, the second discussed the reported

Slaghuis's Framework and Instrument for Sustainability
To identify the predictive role of short-and long-term improvements in quality of chronic care delivery on program sustainability    Level of Institutionalisation Scale To identify facilitators and barriers to the long-term sustainability of antiretroviral therapy (ART) programs at six health facilities Uganda Primary Care benefits and challenges of use, and the third describes the impact of approaches on sustainability outcome variables.

Sustainability approach application
Results demonstrate an increase in the use of sustainability approaches in peer-reviewed studies over time (Fig. 2). Approaches have been applied in a range of healthcare settings (Fig. 3). Most (53%, 36/68) were based in primary care settings (e.g. primary care trusts, GP surgeries, clinics, etc.). Fewer approaches were used within tertiary care (13%, 9/68) or secondary care settings (10%, 7/68). Five approaches specified that they were used in community healthcare initiatives and three approaches were used in nursing education initiatives. The approaches have been applied in over 19 countries ( Table 2). The highest proportion were conducted in the United Kingdom (34%, 23/68), followed by 15% (10/68) in the United States. Commonly, studies were undertaken within one country, but a small proportion (4%, 3/ 68) took place across multiple countries. While multiple study designs were found, qualitative methods were the most common, with 59% (40/68) of the studies employing methods such as interviews, focus groups and participant observation; this included four case studies [52,54,110,112], two grounded theory approaches [59,105] and one ethnography [85]. In total 28% (19/68) were mixed-method studies and 13% (9/68) were quantitative study designs using questionnaires, regression analyses or weighted scales. Most studies (79%, 54/68) explored the application of the sustainability approach, across multiple sites or organisations. For example, Campbell et al. applied Gruen's Model of health-programme sustainability to describe the sustainability of a hospital-based smoking cessation program across six Canadian hospitals [53]. Finally, 21% (14/68) were used within single sites (e.g. intensive care units, care homes, GP clinics, etc.).

Sustainability approaches applied in reviewed studies
Within the 68 reviewed articles, 17 different sustainability approaches were applied. Table 3 summarises the sustainability approaches demonstrating use within the reviewed publications along with the purpose and constructs included within each approach. The most commonly applied approach was the Normalisation Process Theory (NPT), which was used to in 30 articles. The Normalisation Process Model, Level of Institutionalisation (LoIn) Scale, Program Sustainability Assessment Tool, Program Sustainability Index and the NHS III Sustainability Model were all used within 3 or more studies. The remaining 11 approaches were used in 2 or fewer articles.

Purpose of application
The purpose of using a sustainability approach fell into five categoriesanalysis (to examine and interpret data in relation to sustainability), evaluation (to appraise sustainability of an existing programme), guidance (to direct study design, data collection, discussion, etc.), assessment (to measure constructs for sustainability) and planning (to design a programme to achieve sustainability). Twenty-six (38%) articles used a sustainability approach for a combination of these purposes such as planning and evaluation or guidance and analysis (Fig. 4). For example, Johnson et al. used NPT to both guide and assess sustainability of an intervention to improve the nutritional care of preterm infants [79]. Nineteen of the articles (28%) used a Fig. 2 Number of articles describing the application of a sustainability approach by year sustainability approach for assessment. For example, Cramm et al. used Slaghuis's Framework and Instrument for Sustainability to assess connections between partnership functioning and sustainability of innovative programmes in community care [57]. Fifteen studies (22%) used an approach to inform data analysis for the study. This was demonstrated in Toledo Romanib et al.'s study where the LoIn scale was used to guide content analysis [102]. No study used a sustainability approach exclusively for planning for sustainability, but planning was reported in combination with other purposes.

Benefits and challenges
Details on and documentation of the application of the sustainability approaches varied across the studies. While 65% (44/68) of articles identified benefits associated with the use of approaches, only 34% (23/68) reported challenges. Table 4 reports findings on the benefits and challenges for the individual approaches. This is followed by discussion of six common themes related to the benefits and challenges found across sustainability approaches.

Benefits of use
For the 65% (44/68) of articles describing benefits of use, three main benefits were found, namely (1) improved conceptualisation of sustainability, sustainability constructs and their impact on systems, (2) improved ability to organise, analyse and interpret sustainably data, and (3) increased focus on areas for improvement.
Improved conceptualisation of sustainability, sustainability constructs and their impact on systems: 37% (25/68) The most common benefit reported from the use of sustainability approaches was the improved understanding of sustainability and sustainability constructs. Twenty-five articles stated they were able to improve conceptualisation of sustainability of interventions [47, 48, 50, 53, 54, 58, 61, 71-73, 75, 78-80, 86, 88, 92, 95, [113] To explore the social organisation of the work (implementation), of making practices routine elements of everyday life (embedding) and of sustaining embedded practices in their social contexts (integration) Coherence (or sense-making), cognitive participation (or engagement), collective action (work done to enable the intervention to happen), and reflexive monitoring (formal and informal appraisal of the benefits and costs of the intervention) 30 2. Normalisation Process Model [114] To assist in explaining the processes by which complex interventions become routinely embedded in healthcare practice Interactional workability, relational integration, skill-set workability, contextual integration 8 3. Level of Institutionalisation Scale [115] To measure the extent of programme integration into an organisation To analyse and map the nature and extent of integration in different settings, along with the factors that influence the integration process Nature of the problem, the intervention, the adoption system, health system characteristics, context 1   [58,120]. Some articles also explicitly reported gaining a better understanding of barriers and facilitators to sustainability of their initiatives. The Program Sustainability Assessment Tool was used to help predict an initiatives' integration in Scudder et al.'s study which found barriers in financial support and implementation funding [96,116]. Improved ability to organise, analyse and interpret sustainably data: 26% (18/68) Many of the approaches were used to guide study design and analysis. In this process, authors noted that the approaches aided analysis and interpretation of data [52,55,60,63,64,68,70,76,77,81,89,94,98,105,107,108,111]. For example, within Sanders et al.'s article, the NPT was used to organise their semi-structured interview data [94]. Authors noted that the NPT not only aided in the design of the interview but also guided thematic analysis on improving uptake of the new system for general practitioners. Similarly, Zakumumpa et al. found that the LoIn Scale aided them to analyse antiretroviral therapy institutionalisation scores across multiple cases [111].
Increased focus on areas for improvement: 9% (6/68) Finally, the use of sustainability approaches was beneficial to improve focus and highlight areas needing further work [56,62,66,69,82,85]. This was highlighted within Ford et al.'s study using the NHS III Sustainability Model. This study found that the model provided a measurement indicating where a component in the organisation needs improvement for sustainability [66]. Similarly, Blanchet et al. noted that the Sustainability Analysis Process aided them to better understand the need for participation of major stakeholders to promote sustainability within the rehabilitation sector [50].

Challenges of use
Key challenges were reported in 23 articles. These included (1) issues with approach constructs, (2) need for adaptation or improvement and (3) difficulty in application.
Issues with approach constructs: 16% (11/68) The most common challenge associated with the use of sustainability approaches was associated with approach constructs. Issues such as applicability, relevance, overlap or missing constructs were encountered across a number of studies [47,62,70,71,82,95,96,100,104,107,109]. For example, users found difficulty in discerning the differences between the constructs within the NPT, which were often reported to overlap in content and concept. Both Drew et al. and Gillespie et al. found that the overlapping nature of the four constructs within the NPT meant the data could be coded into more than one construct [62,71]. Another noted issue was the inability of constructs within approaches to cover all the relevant areas needed for studies [82,104,109]. This was identified within Underwood et al.'s study, which found that Leffer's Conceptual Framework for Partnership and Sustainability did not adequately explore resources or partner factors [104].
Need for adaptation and improvement: 10% (7/68) Seven studies noted that the approaches were a challenge to be used as originally designed and suggested areas for improvement [24,46,48,68,79,85,93]. For example, the Program Sustainability Index was used by Ahmad and Abu Talib to assess community capacity. During the application, authors stated that the index had to be adapted to suit the study needs [46]. A number of studies also recommended improvements for approaches to work more effectively. Redman and Barab suggested the LoIn scale required further testing with larger samples and different health programmes to improve its applicability [93]. Bamford et al. suggested that the NPT may require further development to explore specific behaviour change techniques to increase the practical value of the theory [48].
Difficulty in application: 9% (6/68) It was noted within six papers that authors encountered some difficulty in applying the approaches [24,52,77,91,97,110]. In Winterton and Chambers' study, they found that Shediac-Rizkallah and Bone's conceptual framework was difficult to apply to initiatives in certain ethnic populations [1,110]. Doyle et al. reported similar findings with the NHS III Sustainability Model. In this study, users reported concerns over aspects of the model design and only 12 of the 19 teams found it acceptable for routine use [24,125].

Examining impact on sustainability outcome variables
To understand potential links between the use of a sustainability approach and enhanced reporting of or impact on sustainability, six sustainability outcome variables were investigated. The majority (79%, 54/68) did not report sustainability outcome variables in their studies. Low reporting was attributed to how the approaches were used in practice as many articles applied approaches to aid conceptualisation of sustainability constructs; therefore, sustainability outcomes were not explicitly assessed. Only 21% (14/68) reported one or more sustainability outcome variable. Table 5 summarises the articles that reported sustainability outcomes. Below each outcome variable is discussed with specific examples of how each was reported within the studies.

Continued programme activities or components of the original intervention (9/14)
The most common outcome reported across the studies was continuation of programme activities or components, with nine articles exploring the continuation of initiative activities [76,78,82,84,93,96,99,102,106]. 'Continuation of activities' was demonstrated through multiple measures, including continuation of services (e.g. continued delivery of the HIV testing service [82]), tool use (e.g. continued use of a shared decision-making tool [84]) and continued delivery educational programmes for staff and patients (e.g. diabetes education programmes [93]). Three of these detailed partial continuation of activities [78,99,106]. For example, Redman et al.'s study using the LoIn scale found 'low to moderate levels' of routinisation were achieved, which suggested that the programmes had embedded into the host organisations [93]. Similarly, Ibrahim et al. discussed partial continuation of a grip strength measurement programme across five acute medical wards with weekly coverage of the programme ranging between 65% and 100% in one site and between 0% and 93% in another [78].
Maintenance of new organisational practices, procedures and policies that were started during programme implementation (7/14) Maintenance of new practices and policies was highlighted in six articles [50,66,79,82,93,99,102]. This variable was demonstrated through intervention integration within clinical guidelines [79,82], electronic records [66], through ongoing discussion at team meetings [66] and monitoring/evaluating activities [99]. For example, Toledo Romani et al.'s study using the LoIn scale found that elements of the intervention (community working groups and provincial/municipal coordination groups) had lost their separate identity and had become part of the programme's regular procedures [102]. Blanchet et al. followed two rehabilitation sectors using the Sustainability Analysis Process for 2 years. At follow-up, they noted that, although there were differences in the structure of social networks, the programme policies continued to be used [50].
Whether benefits or outcomes for consumers, clients or patients are continued (6/14) Six articles reported continued benefits to patients [66,76,79,84,102,106]. Detail on these benefits was variable across the studies. Some reported specific measures of continuation of benefits such as continued behaviour change or continuation of improved patient outcomes [102]. For example, Toledo Romani et al. used the LoIn scale and demonstrated maintenance of behavioural changes related to a dengue control intervention [102].
Johnson et al. used the NPT to study guideline compliance and found that the intervention resulted in maintenance of improvements in health outcomes for infants, specifically protein intake and weight gain [79]. Others reported continuation of benefits in terms of continued support of patients and families (e.g. a shared decisionmaking tool helping patients to choose treatment options [84] or family care maps supporting families and patients on the continuum of rehabilitation process [66]). Higuchi et al. used the NHS Sustainability Model and reported vague continuation of benefits. They described how sustained organisational perceptions of benefits existed 5 years after their initial initiative to introduce multiple nursing guidelines [76]. A number of studies also described the benefits to other stakeholders affected by the programme, such as their staff, communities and carers [65,66,79].
Maintenance of community-level partnerships or coalitions developed during the funded programme (5/14) Five articles reported the maintenance of partnerships and described how it was beneficial or challenging for the continuation of the programme [50,60,87,102]. This variable included the active continuation of stakeholder networks, champion engagement, leadership groups, steering committees and building further collaborative relationships. The maintenance of partnerships was highlighted in Diaz del Castillo et al.'s study where champions were a core part of the improvement delivery. The programme was grounded in the work of champions, who would continuously advocate, garner further resources, recommend strategies and implement activities [60]. Scudder et al. also discussed the importance of partnerships in sustaining a parent-child interaction therapy programme. They noted that the leadership groups and steering committees developed during the maintenance period were key sustainability outputs from the programme. Blanchet et al. found similar importance of developing strong partnerships and networks but noted that these relationships can also be a challenge to sustain. Within this study, one site documented the breakdown of their stakeholder network, which impacted their ability to continue to collect sustainability indicators [50].

Sustained attention to the issue or problem (3/14)
Three studies highlighted that attention to the issues or problem was sustained [60,102]. This was demonstrated through the use of national campaigns and goals [60], mass media and TV campaigns [102], and community meetings [102]. Diaz del Castillo reported that their programme was able to sustain attention through involvement in national campaigns where their 'Healthy Habits and Lifestyles Program' was included in the Ten-Year National goals [60]. Scudder et al.'s study also reported maintaining attention to their programme through a number of strategies, including 'word of mouth' and TV campaigns.

Programme diffusion and replication in other sites (2/14)
Two studies reported programme diffusion or replication [60].

Additional sustainability outcome variables
Throughout data extraction, it was noted that several articles discussed and reported sustainability outcome variables that did not fit into the six variables explored [17]. Three emerging outcomes reported across various studies are discussed below.

Capacity built (5/14)
This outcome was discussed within five studies [60,66,76,99,102] in relation to how the skills and capabilities developed throughout an initiative could continue to be utilised and developed after the end of initiative funding. This variable refers to the development of strengths, skills and abilities of staff, stakeholders and communities to maintain new organisational practices that were started, through their involvement with an improvement initiative [126]. This was demonstrated through integration of intervention into training programmes and orientations [66,99,102], and attending knowledge exchange events [76]. Diaz del Castillo et al.'s study specifically commented on the continuation of capacity built and skills gained in the initiative [60]. Authors reported that programmes had expanded and been maintained due to the quality of the instructors, which was seen to impact both people's continued attendance and perceived benefits. The programme maintained training for staff with a 'Teachers' Academy' created to select and train instructors for the programme [60].
Innovation and Adaptation (5/14) This outcome relates to the fidelity and flexibility of initiatives and the ability to respond and adapt to challenges and changing needs. While most studies that were reviewed did not describe adaptations or examine their impact on health-related outcomes, this variable was discussed in select studies [50,60,87,106]. Specifically, adaptations to system characteristics [50], funding changes [60], internal culture [87], existing needs and challenges and products (e.g. changing campaign images [106]) were noted. For example, Scudder et al.'s study found that, while initiatives were actively sustaining, no clear strategies were linked to their success but rather, "the common theme across initiatives was the importance of being responsive to existing needs and challenges". Diaz del Castillo also found that, for both programmes in their study, there was "a need and willingness to be flexible" [60]. Flexibility, in this case, meant adapting programmes to people's preferences, funding changes and diverse scenarios/conditions [60]. Blanchet et al. also noted how adaptation and resilience played a part in the sustainability of improvements with two rehabilitation sectors. They noted that "the structure of the rehabilitation system characteristics evolved over time and helped understand the changing nature of relationships between actors and their capacity to work as a system. This aided in the creation of consensus on a common vision of sustainability" [50]. To facilitate a greater understanding through future research, some clarity regarding adaptation and fidelity is necessary [3].
Ability to garner further funding (4/14) An initiatives' ability to garner further funds to continue the initiative and maintain gains was also reported as a sustainability outcome across several studies [50,60,99,102]. Toledo Romani et al. reported how funds for the activities of the community dengue intervention were included in the annual budget of the wider dengue control programme as an action towards sustainability [102]. This was also highlighted in the Diaz Del Castillo study where they described how 19 of 32 departments had co-funded programmes and six programmes operated using their own resources. They highlighted that "the creation of a specific PA (Physical Activity) unit in 2011 led to a dedicated budget, resources, goals and a capacity to translate policy into action" [60]. Blanchet et al. also identified gaining further resources as a mechanism indicating sustainability and noted that "financial support by an international organization facilitated advancement toward selfidentified sustainability goals" [50].

Discussion
While early implementation research was criticised for a lack of attention to the theoretical underpinnings, in the last two decades, numerous theories, models and frameworks have been produced [19]. Due to the vast number of approaches available, choosing the most appropriate one is increasingly difficult [30,31]. To understand the value of models and frameworks, we require supporting evidence to warrant their use [32]. This review provides healthcare professionals and researchers with the current evidence on sustainability approaches, including the benefits and challenges that arise through their use and the outcomes they have supported.
This review tracked sustainability approaches in the literature and found 68 articles detailing their use in peer-reviewed publications. It found that, while many sustainability approaches are available, few have demonstrated their practical application or benefits, with only 17 of the 62 available approaches showing use in peerreviewed publications [27]. Sustainability approaches were used to assess, evaluate, analyse and guide sustainability measurement and research. Benefits of use include improved understanding of the barriers and risks to sustainability, improved conceptualisation of sustainability and its constructs and improved ability to organise, analyse and interpret sustainably data. The literature has also documented challenges, including issues with approach constructs, difficulty in application and need for improvement in design [4,54,62,82,127]. These results support previous findings reporting that tools, frameworks and models may improve team knowledge and enhance action but they can also be complicated to use, hinder teamwork and take substantial amounts of time [21,46,48,54,62,82,[127][128][129][130][131][132]. The findings presented in this study, along with these previous reports, demonstrate the need for practitioners and researchers to have clear and realistic expectations when using frameworks, models or tools for sustainability. It is important to not only understand the potential benefits but also the challenges to realistically consider application and weigh potential issues with expected impact [3,35,36].

Implications for future measurement
This study has shown that, while there is increasing interest in the use of sustainability approaches, there is little reported evidence of their impact on sustainability outcome variables [21,26,36,116,133,134]. Although most sustainability approaches are intended to support or influence long-term success, only 21% (14/68) of studies reported any information on sustainability outcomes. This is similar to the results reported by Stirman et al. in their systematic review, which found only 22% of studies reporting information about sustained outcomes [3].
The low reporting of sustainability outcome variables is often attributed to poorly defined concepts of what sustainability means within individual studies [36]. This issue is compounded by the frequent reporting of sustainability as a dichotomous measure [135]. For example, in Leon et al.'s, study they concluded that their intervention had become sustained and was "embedded in practice (normalised) during a two-year period" but information on what was embedded, what specific activities were continued, and what other variables were sustained were not discussed [82]. These findings highlight that, although the use of multiple outcome variables to report sustainability outcomes is recommended, it is not common practice [17].
Future research should endeavour to clearly describe and outline expected sustainability outcomes and consider the use of multiple outcome variables to capture comprehensive sustainability narratives [17]. Reporting dichotomous or unclear outcomes contributes to underreporting of sustainability failures and successes, making it increasingly difficult to understand the true scale of the problem, compare outcomes across studies or provide generalised learning on sustainability measurement.
While this study examined six sustainability outcome variables, three additional variables were discussed in our results, namely Capacity built, Innovation and adaptation, and Garnering funding. Although not part of Scheirer and Dearing's original six variables, these outcomes have been suggested elsewhere as being potential variables that are sustained following healthcare initiatives [1,21,22,116,121,[136][137][138]. These additional outcomes may lend a new understanding of how sustainability outcomes can be viewed, assessed and reported in practice.
The role of continued innovation and adaptation in assessing sustainability is of particular interest to those adopting a 'process' view of sustainability. Much of the previous research considered deviations from intervention protocols to be implementation failure but, due to growing recognition that institutionalisation of activities may impede organisations from adopting more effective practices, there is now greater acceptance of the role of adaptation and capacitybuilding in sustainability [3,4,22,36,139]. Unfortunately, few studies explore what adaptations are made to interventions, why such changes are necessary, and how decisions to change intervention components are made [3,17,140]. Future research should prioritise studying this process to provide insight into how adaptations may impact outcomes and sustained services [3,141,142].

Implications for practitioners and researchers
It is recommended "that researchers specify the unit of data collection and analysis for both sustainability determinants and outcomes" [36]. This review seeks to support this further by incorporating emerging findings to guide measurement and study of sustainability in healthcare settings by providing an updated list of sustainability outcome variables (Table 6). It is hoped that having a clear list of potential outcomes to assess will support practitioners and researchers to clearly define the sustainability outcome variables of interest and consider how evidence for each outcome may be collected and analysed. It also aims to support the consideration of a wide breadth of sustainability outcomes to discourage unclear, over simplistic or dichotomous reporting of outcomes. While all variables may not apply to all settings, this list can serve as a baseline for sustainability outcome planning and measurement. To understand its potential value in achieving these aims, this list requires further testing and validation within future studies. Specifically, there is a need for future work to provide insight into how data on each of these variables can be collected and analysed.
Given this expanded list of outcome variables, planning for their collection and analysis may be a challenge. The most appropriate methods for collecting evidence related to each of these variables has not been determined but it is increasingly recognised that sustainability research should draw on mixed methods [3,13,17,143]. A mixed methods approach supports both quantitative methods key to measuring ongoing performance of initiatives (e.g. continuation of benefits or activities) as well as the qualitative methods to provide the necessary 'texture and detail' to describe intervention sustainability (e.g. continued partnerships or training) [143].

Future research
While this review presented the available evidence on the application and impact of sustainability approaches in healthcare, it also highlighted the remaining gaps in

I. Benefits
Benefits for patients, staff and stakeholders continue e.g. continued improvements in health outcomes for infants in protein intake and weight gain [79] II. Activities Initiative activities or components of the original intervention continue e.g. continuation of the HIV testing service delivery [82] III. Relationships, partnerships and networks Maintenance of relationships, partnerships or networks that were developed during the funded programme e.g. maintenance of relationships with champion groups to continuously advocate for and implement activities [60] IV. Procedures and policies Maintenance of new organisational procedures, and policies that were started during program implementation e.g. workflow integration achieved through ongoing communication and documentation in electronic records [66] V. Attention and awareness Attention and awareness of the problem or issue is continued or increased e.g. using educative messages through mass media and community meetings [102] VI. Spread Replication, roll-out or scale up of the initiative e.g. programme is scaled-up locally/nationally and exist in similar forms elsewhere [60] VII. Skills and capabilities Capacity built (skills and capabilities) within staff, stakeholders and communities developed throughout an initiative continue to be utilised e.g. intervention is incorporated into staff orientation [66] VIII. Innovation and adaptation Adaptation in response to new or changing populations, evidence, policies, or other contextual influences e.g. replacing intervention posters in public places or changing intervention images [106] IX. Garnering funding Gaining further funds to continue the initiative and maintain improvements e.g. funds for the intervention activities are included in the annual budget for illness prevention strategies [102] the literature. First, there is a need for future research to more rigorously report and explicitly describe how frameworks, tools and models inform development, organisation and implementation of research [7,19,27,34]. This evidence will help to support improved sustainability approach development and improve the reliability of sustainability studies [3,19,35,36]. Further work is also needed to build an understanding of underuse and misuse of frameworks, models and tools [29]. This study found that sustainability approaches were largely used for analysis and conceptualisation purposes to aid in interpretation and organisation of data. This does not reduce the value of these studies but highlights the challenge of understanding the impact of these approaches if they are not used as fully intended. For example, the aim of the NPT is to help understand how complex interventions can embed change and become 'normalised' in practice [113]. Although this approach has the potential to explain how complex interventions become routinely embedded, the evidence shows that it is often used to assess implementation processes only. Future studies may benefit from reporting on all aspects of the approach, including embedding and integration. There is also a need to empirically test the updated sustainability outcomes list to assess if it clarifies what continued, discontinued or changed for interventions and if it provides better insight into sustaining initiatives. This will be critical in understanding the process and potential difficulty of studying these diverse sustainability outcomes in practice.

Strengths and limitations
This is the first review to examine the application and impact of sustainability approaches in healthcare and the first review attempting to understand the impact of sustainability approaches on sustainability outcomes. There are a number of limitations associated with this systematic scoping review that are must be acknowledged. The first limitation is the potential for publication bias which may have manifested in a number of ways. First, the search strategy may have missed published articles if they had not referenced the use of an approach within their study. Second, this review only drew on the 62 approaches identified in a previous systematic review, which meant that approaches published after 2018 were not included in this work. This was largely a pragmatic decision as we believed the existing 62 approaches would provide sufficient evidence of the scope of application and impact of approaches. It was also believed that newer approaches would not have had sufficient time to build evidence of their impact within this short timeframe. Therefore, future work should continue to assess and collect information on emerging approaches and the evidence of their application. Finally, the review did not search the grey literature, which means that findings may be more representative of academic and research application and analysis.
While there is value in consolidating insight into the perceived experience of implementing such approaches, another limitation to this work is the reliance on reported benefits and challenges. Many papers did not discuss benefits or challenges and those that did may not have been comprehensive in their descriptions as this was not the primary aim of their work. This may have resulted in over or under reporting of benefits and challenges.
Another limitation is the disproportionate inclusion of studies using the NPT. As this review sought to give an overview of the available approaches and their use, attempts were made to balance results where possible by providing narrative examples and descriptions for all sustainability approaches throughout the manuscript.
Finally, it is also important to note that this work does not seek to critique the included approaches but instead to consolidate the reported evidence on their use. Many of the sustainability approaches have been developed to encourage more systematic work and guide empirical studies; therefore, it should be acknowledged that this area of research is still in its infancy. Many of the conceptual frameworks and models aim to instigate knowledge generation on the determinants that influence sustainability processes and outcomes. In order to do so, they need to be applied and tested across real-world settings [4,21,116,122]. Therefore, full insight into the impact of sustainability approaches will continue to be developed over time.

Conclusion
This review provides healthcare practitioners and researchers with a consolidated summary of the current evidence base on sustainability approaches, including the benefits and challenges that arise through their use and the outcomes they have contributed to. The use of a sustainability approach can improve the knowledge of sustainability constructs, aid in analysis and increase focus to promote sustainability. They can also create challenges associated with their content, relevance and applicability. These findings highlight the need for thoughtful consideration of the potential impact prior to use. They also emphasise the remaining gaps in the literature, with a substantial amount of work still needed to ascertain the value and contributions of many sustainability approaches. In order to understand if these approaches are providing the foundations for rigorous inquiry and advancing sustainability research, wellplanned studies recognising comprehensive sustainability outcome variables are needed.