Two frameworks were utilised for this review. Previously identified terms and the taxonomy and mapping of concepts [17] was utilised for the search strategy. The studies are described using this conceptual framework to show the relationship to SDM. The findings from this search are then synthesized using May’s theory of implementation [18] , to explain how the findings relate to clinical practice.

Embedding SDM as a new healthcare intervention within the usual clinical practice may be challenging. May [18] proposes implementation of a new intervention as a social process of collective action as it never refers to a single ‘thing’ that is to be implemented but rather a complex bundle of material and cognitive practices to be introduced into a social system. It follows that a process as complex as SDM needs substantial support and planning before and during widespread implementation into clinical areas to enable its success [19] . A theoretical framework of implementation characterises the elements of context and agency within a social system, that enable capacity to implement change [18] . May’s [18] framework provides a template to understand health care contexts and the elements involved which act as barriers or enablers to implementation of new interventions. The key components of May’s framework include potential, capacity, capability and contribution to implementation in practice.

A conceptual framework is an essential component for undertaking framework synthesis [20] . A framework guides decisions regarding analysis and assists in the interpretation of findings [21] .

The first author extensively reviewed the literature from August to December in 2015 and updated the search in November 2016. An anticipated complexity of the literature search was the multiple terms used independently and interchangeably for SDM. The heterogeneity of the concepts, definitions and terms used in relation to SDM has been described by other authors [22] . Makoul and Clayman [23] , found in a review of 418 papers devoted to SDM that the terms “patient preference” and “options” were terms used in more than half. Policy documents frequently refer to “patient choices” [24] and “consumer participation” [25] . Cooper [26] supports incorporating the broadest conceptual definitions possible.

Subsequently the search strategy was based on previously identified terms and concepts developed in a taxonomy and mapping of SDM [17] . Additional terms used to describe RAC internationally and traditionally were added (such as long term care and nursing homes). Because the focus was clinical practice, clinicians were defined by terms used internationally in the literature (care staff, nurses, allied health and medical staff). The search was adapted to each bibliographic database and Medical Subject Headings (MeSH) used where possible.

Search terms used for SDM and related concepts were;

“Shared decision making” OR “decision-making” OR “patient preference” OR “patient option” OR “patient choices” OR “consumer participation” OR “Consent” OR “guardianship” OR “advocacy” OR “proxy” OR “next of kin” OR “advance care plans” OR “patient centred care” OR “patient autonomy” OR “evidence based practice” OR “variations in care” OR “training” OR “education” OR “health literacy” OR “decision aids”.

To capture the relevant literature in residential aged care settings, SDM search terms were combined with AND for: “residential aged care” OR “nursing home” OR “long term care” OR “care staff” OR “medical staff” OR “allied health” OR “nurses”.

Bibliographic data bases searched included MEDLINE (Ovid), CINAHL, PsychINFO and Scopus.

Articles related to governance and occupational health and safety were excluded since the aim was to explore practice and processes relevant to SDM implementation in RAC. Studies conducted in settings other than RAC, such as community or transitional care, were excluded. Studies prior to 2005 were excluded to give a contemporary account of existing models and processes. Studies reported in languages other than English were also excluded. Qualitative and quantitative, or mixed methods research approaches were included, plus one case study due to its relevance to the topic under consideration. Systematic reviews were excluded but the reference lists were searched.

Quality appraisal of the methodology of the studies and risk of bias was not a central focus. Studies were critically appraised by the first author using the Critical Appraisal Skills Program (CASP) tool [27] , to determine validity of the aims and methodological quality.

The search yielded 374 articles, Medline-153, PsychINFO-105, Scopus-47, CINAHL-44, Pubmed-27. Experts in the field of SDM were also contacted to provide other sources of potential studies, which provided three additional articles. The total was reduced to 300 when 76 duplicates were removed. Of the 300 remaining articles, 211 were excluded after reading the abstracts because they were not studies related to the search for models or processes of SDM in aged care, but focused on tool development or discussion or alternative settings. Eighty nine full text articles were screened and a further 62 excluded for being irrelevant to the research question. Of the 27 remaining articles, nine screened negatively to the two initial screening questions of the CASP and so were eliminated. The final 18 articles are described using the conceptual framework and synthesized using May’s theory of implementation. A flow chart shows the selection process in Figure 1.

PCC is a dimension of quality care which is individualised to consider both the patient and their family as integral components of health care decision making and delivery of care [46] . As the essential component, or crux, of SDM [47] models of PCC are an important consideration in RAC. Six of the studies analysed concepts related to PCC [30] [34] [35] [41] [43] . These studies used qualitative [30] [34] , quantitative research methods [41] and mixed methods [43] to provide the perspective of residents and families [35] [41] and staff [30] [34] or both groups [43] to illustrate the opinions of all stakeholders. Time pressure and a task oriented focus of care was a key limiting factor in providing PCC. Families were excluded from the decision making process when a PCC workplace culture was absent, though the families viewed their participation as crucial. Staff propose that an existing workplace culture of PCC is an enabler to supported decision making and family involvement [30] [34] . Importantly, the perceptions of staff and families differed about the level of communication and information exchange that was taking place, with staff perceiving higher levels of PCC than families [43] .

Additional studies [29] [36] [39] included in the analysis illustrate that non PCC models of care result in a lack of SDM. A consistent finding was that paternalistic, task focused and evidence focused workplace cultures denied residents and families choice, from the perspectives of staff, residents and families [29] [36] [39] .

If SDM is to be successfully implemented in RAC it requires a workplace culture of PCC, but evidence suggests that social norms in RAC may not align with PCC, and that residents, families and staff have differing opinions about the workability of PCC.

Evidence is utilised in SDM to ensure that patients understand all the risks and benefits of associated care and treatment (informed consent) and to help them reach a decision based on their preferences [48] [49] . Like PCC, it is an essential component of the SDM process and utilisation of high quality research evidence demonstrates capacity for the implementation of SDM by workability.

Studies that utilised either qualitative [36] [42] or quantitative [40] designs as an intervention in RAC demonstrated contrasting results about potential SDM implementation. Perspectives of residents and families [40] , residents and staff [36] or staff alone [37] [42] are reported in each of these studies. A top down approach to utilising evidence [36] [42] , where staff held the knowledge of best practice, denied residents choice and autonomy, while a bottom up approach [40] , where families were provided with evidence of treatment risks and benefits increased family members knowledge and facilitated SDM with staff. Staff as reported in Sims-Gould et al. study [36] perceived that evidence based practice should take precedence over resident choice, while staff in Werner’s study [42] expressed moral and ethical issues about utilising evidence that is known to abrogate resident autonomy. In another study by Hanson et al. [40] , there is no knowledge of staff opinions of a bottom up approach to utilising evidence. Without staff perspectives of a bottom up approach, understanding of potential implementation is limited.

Four articles focused on advance care planning (ACP) or end-of-life care, two quantitatively measured [38] [44] staff perspectives, one case study [45] described perspectives of all stakeholders and three qualitative methodological approaches [31] [33] focused on residents and families. None of the articles describe utilising evidence to help residents or their families decide on treatment options. Two studies found a lack of provision of information to residents and their families, and reported that staff rarely discussed the risks and benefits of treatment options [28] [38] . Some families indicated that there was a lack of information provided to them to assist in decision making [33] . This suggests that poor knowledge may result in some families deferring decision making to staff, though other families expressed a desire to be part of the decision making process. Staff perceived it was unfair to expect families to make decisions without knowledge of the risks and benefits of the treatment option, and so believed that their expert knowledge justified them as decision makers [29] .

There is evidence that some patients prefer not to be involved and to leave decisions about treatment to clinicians [19] [50] . The reasons for this are multi-fac- torial, but an important consideration. Additionally, clinician and organisation readiness for the process of SDM is vital to implementation. Readiness to participate in SDM was highlighted in five of the studies regarding ACP and end- of-life care [31] [33] [38] [44] [45] , the study methodology and stakeholder perspectives having already been described but which includes a mixture of qualitative and quantitative methodologies plus a case study involving all stakeholder perspectives.

There are differences in the level of involvement residents and families wish to have in decision making in ACP and end-of-life care. Resident and family readiness to participate was shown to be a significant factor in the process of collaborative decision making. Perspectives of residents and families willingness to be involved in decision making varied within and across studies. Residents and families often give staff an imprimatur to make decisions on the resident’s behalf, rather than being involved [29] [31] . Importantly, some families were stressed and uncertain if expected to make a decision, but postulated that this may be a result of poor knowledge of the consequences of their decision [29] . Alternatively, some family members expressed dissatisfaction with being excluded from the decision making process [29] [31] . The study by Ampe et al. [38] demonstrated that readiness of families to be involved in SDM was assessed by inquiring about their preferred approach to receiving information to assist in SDM.

Staff and organisation readiness is paramount to effective SDM [29] [32] [34] [35] [36] [39] [41] . A lack of PCC was the limiting factor in staff and organisation readiness for SDM. Staff unpreparedness for SDM was highlighted in two of the studies [36] [39] , while organisational unpreparedness was perceived as a barrier in two [39] . Frequently, the unpreparedness or lack of readiness could be attributed to both staff and organisational factors [29] [32] [35] [41] . Organisational readiness was also identified as having policies to support SDM in practice by two studies [37] [38] .

There can be multiple factors contributing to staff and organisation readiness [32] which requires a huge organisation cultural shift [43] . Education was identified as paramount in staff, family and organisational readiness for implementing change.

Much of the broader literature on SDM, focuses on education for clinicians on how to execute the process of SDM [51] [52] , what SDM is [53] [54] and why it is important [6] . Education in SDM was not a focus of any of the articles retrieved, however, many of them utilised education to execute and embed models of care [30] [43] or enact organisational change to adopt new processes [44] [45] and are thus particularly relevant to SDM in RAC if it is to be implemented. Education in SDM was recommended in two studies, [28] [38] particularly about the importance of family involvement. The articles analysed which utilised an educational intervention were predominantly quantitative approaches [40] [43] [44] . Hanson et al. [40] differed by providing education to families not staff, while Bauer et al. [43] provided education to staff and families concurrently. All of the studies concerned with education excluded the perceptions of residents and focused on staff and families.

Staff and family education in collaborating together is illustrated as the corner stone for successful implementation of new interventions [43] . It is also important to seek staff opinion on factors which mitigate against collaborating with families [34] [43] . Discussion with family members on treatment options is crucial to SDM [28] [38] [40] . Education can be resource intensive, particularly if it focuses on embedding long term, sustained adoption of change [44] [45] .

The third component of Coulter and Collins SDM process [5] is documentation of the decision reached, communication between stakeholders and enactment of the treatment choice. Without this final component, the previous components have no validity in the process of SDM. Documenting and communicating residents preferences so that they can be enacted are the operational phase of SDM.

There were six articles concerned with the components of documentation, communication and execution of care in RAC settings, three articles used a quantitative research design [29] [33] [44] , a case study [45] , and three used qualitative methodology [29] . Arendts et al. [29] provides perspectives of staff, residents and families, while the other articles focus on staff [34] [44] or families [33] alone, or is purely descriptive [45] .

A support worker (facilitator or broker) was perceived as the key to implementation of Advance Care Planning (ACP) [44] [45] . The support workers role is to educate staff and families about process and act as a mediator. The support worker initiates discussion with all stakeholders and facilitates the implementation of goals and plans. Involvement of residents and relatives in understanding the process of an intervention is central to successful implementation. The importance of information exchange and communication between stakeholders is emphasized [34] . Barriers to execution of wishes include a lack of documented resident preferences [33] , or inadequate planning and communication [29] . Systematic operationalization of an intervention requires regular and ongoing education of all stakeholders [44] as well as auditing of existing processes, policies and strategic planning about organisational changes that may be required for success [28] [37] . High staff turn-over and changes to management of RAC is a threat to sustainability of interventions [44] . Blackford et al. [44] describe a system-wide approach to documentation, communication and dissemination of a residents wishes as paramount to success.

Capacity, in terms of RAC, is bound up in the social norms that govern behaviour within it. Paternalistic attitudes are a common social norm in RAC [29] [36] [39] [58] and the antithesis of SDM-there is an imbalance of power between clinicians and patients created by law and specialist knowledge [59] . Paternalism is frequently driven by clinician’s belief that they know what is best for patients. There is a tension between strong commitment by staff to keeping residents safe and violating their right to choice and autonomy [36] [39] . For SDM to be effective clinicians must relinquish their role as the single, paternalistic authority and train to become more effective coaches or partners [2] . The greatest challenge is turning the rhetoric into reality, changing workplace culture requires a huge shift in the common social norms of many RAC facilities [43] . Respect for a person’s right to make choices and participate in decision making is generally viewed as central to quality of life and well-being [30] , however existing social norms in residential aged care potentially prevents SDM implementation [37] .

A culture of person centred care is fundamental to SDM, and a recurring theme in the articles reviewed [30] [31] [34] [35] [41] [43] . PCC is a required social norm for implementation of SDM and SDM is viewed as a strategy that incorporates PCC principles into standards of care [60] . An audit tool has been developed to measure PCC in RAC settings, which aims to minimise staff bias of their perception of providing PCC [61] . Auditing PCC, as an essential component of SDM, is paramount to determine capacity for implementation. Implementation involves interaction with other agents, processes and contexts [18] . A workplace culture of PCC indicates capacity for SDM implementation, one factor of readiness. As well as organisational readiness, readiness of residents and their families is an essential component of capacity to implement SDM.

Material resources within a social system including their distribution and allocation are also a dimension of capacity. Similarly, shared commitment requires that participants are induced to participate, and Implementation Theory recognises that coercion, rather than motivation may be a strategy used to constrain others. SDM utilises evidence to inform the patient of the risks and benefits of various treatment options, to support their decisions [51] [62] , not to coerce patients into practices or treatments that conflict with their preferences and values. Patient preference is a warranted variation to best practice initiatives [63] .

Evidence can be utilised through the provision of patient decision aids. Although decision aids do not guarantee that decision making will be shared [64] [65] they offer a structured way to provide detailed information about treatment choices [1] . Decision aids are valuable for patients with low literacy levels [65] and so provide real hope for facilitating SDM with various population groups. Although decision aids have demonstrated effectiveness about increasing patient knowledge and risk perception, SDM is not dependent on them [6] . While increasingly available, a decision aid, or quality evidence may not necessarily exist for the condition under consideration, or the evidence may be rapidly changing. Clinicians must be able to access up-to-date evidence, evaluate quality and ensure that they are appropriate for the patient and their condition [66] , skills the clinician may not possess. Utilising decision aids is not straight forward [6] . The implementation theory framework suggests that agent’s capacity relies upon cognitive and material resources [18] . Knowledge of how to access, interpret and apply evidence will be crucial to successful implementation of SDM.

The provision of resources to both educate staff and families [34] [42] [43] [44] [45] , and promote sustainability increases capacity to implement SDM in RAC. The provision of education in SDM is widely supported [28] [37] [67] [68] . Paradoxically, the resources necessary to implement SDM, such as increased time, support workers and overall governance, may also be a barrier. RAC settings are typically busy, time poor environments [30] . Time constraints are a limiting factor for practicing PCC [34] , and also found to be true for SDM [37] [38] . Additional resources and dedicated funding are required to embed SDM practice within RAC facilities.

Lack of resident and family readiness for SDM is a barrier to potential implementation [29] [31] . Resident and family readiness relates to individual intention. Clinicians must first assess the patients preferred level of involvement in decision making [67] [69] . Previous authors of SDM studies [67] emphasize the importance of flexibility and warn that strict adherence to the principles of SDM risks another form of paternalism. Some patients prefer not to be involved and to leave decisions about treatment to clinicians [31] . It is recommended [68] that for SDM to become embedded requires a bottom up approach, where patients have an expectation that decision making will be negotiated together. The current regulatory pressures in many countries [54] to adopt SDM ignores that flexibility must be maintained, and that patient autonomy begins with the preferred level of involvement. Non participation in the decision making process, is in itself SDM, provided that risks and benefits of treatment options are clearly articulated and understood. Individual intention and collective commitment to SDM demonstrate potential to implement it. The object of an implementation process may be a modified way of thinking, where the agent intentionally makes things happen by action.

Like capacity, much of contribution and capability relate to PCC. SDM includes collaboration (collective action) with families as appropriate and emotional and physical support [70] . Contribution includes cognitive participation and reflexive monitoring. Joint action, such as sharing of information, unites two or more individuals towards a shared end. Documentation and dissemination of resident choices are essential to operationalize SDM [44] . Capability is concerned with the dimension of workability-operationalizing the intervention. Communication between all stakeholders is necessary for dissemination of information but there is evidence that clinicians and residents/families do not communicate well [59] . There is also evidence that enhanced communication can be achieved through a process of education of staff and families on how to collaborate more effectively [43] . To successfully implement and embed SDM will require RAC facilities to undertake reflexive monitoring or appraisal of the effects of SDM to ensure that property is shared, not a single entity. Reflexive monitoring frames how participants collect and utilise information about the effects of the intervention.