Shared decision making in social services – some remaining questions

ABSTRACT Recent years have seen increased emphasis on person-centring and user involvement in healthcare delivery and social services. In Sweden, the rights of users to participate in and influence care planning and services are stipulated in the Social Services Act, the Health Care Act and the Patient Law. Shared decision making (SDM) is presented as a supportive collaboration between users and staff, drawing on users’ preferences and values as well as the best available evidence to discuss options and reach a consensus on care decisions. However, studies on the implementation of SDM in somatic care illustrate challenges, such as staff trying to persuade users to agree to a particular option rather than offering opportunities to discuss various alternatives. The Swedish National Board of Health and Welfare recommend that SDM should be offered in social services support in relation to certain groups. However, the imperative that ‘SDM should be offered in the social services’ raises questions; in addition to the question of how SDM should be understood, we need to account for why SDM should be offered, who should offer SDM, when it should be offered, to whom and regarding which decisions. The questions concern justification as well as clarifications. The aim of this article is threefold: to discuss SDM and its components; to deliberate upon what SDM may imply to the social services and finally, to suggest some preconditions that should be present for an implementation of SDM.


Introduction
In recent years, there has been increased emphasis on individualization, person-centring and user involvement in healthcare delivery and social services. This emphasis has been expressed both at the policy level and by user organizations. The development of user involvement in health care and social services has been strengthened by clarifications in policy and in legislation. In Sweden, the rights of the user to participate in and influence care planning and services are stated in the Social Services Act and the Health Care Act and were strengthened in a new Patient Law in 2014 as well as in numerous policy documents. User participation is often described as including principles of respect for the knowledge of the client, coordinated and personalized services, and selfdetermination (McLaughlin 2009;Levin, Gewirtz, and Cribb 2017). Traditional decision making in health care and social services has been criticized as abusing power and excluding users' rights to be informed and consulted, and generally for underestimating the importance of users' experience, which may affect the suitability of offered interventions (Richards et al. 2013;Farelly et al. 2015;O'Connor, Stacey, and Rovner et al. 2009).
One process put forward to reach user participation is shared decision making (SDM) (Stovell et al. 2016). It originates from the field of somatic care and has been described as a process of supportive collaboration between users and staff, drawing on users' preferences and values and on the best available evidence to discuss multiple options and reach a consensus on care (Duncan, Best, and Hagen 2010;O'Connor, Stacey, and Rovner et al. 2009). As an example, Stovell et al. (2016) illustrates promising evidence that SDM supports users towards empowerment, including a subjective sense of involvement in treatment, self-efficacy and reduced coercion in community mental health services.
SDM diverges from compliance in that it assumes that two experts -the user and the practitioner -share their respective knowledge, experience and viewpoints and collaboratively decide on a treatment option (Stovell et al. 2016). However, studies investigating the implementation of SDM in somatic care also illustrate challenges, such as staff trying to persuade users to agree to a particular option rather than offering an opportunity to discuss various alternatives as prescribed in the SDM process (Land, Parry, and Seymour 2017;Elwyn et al. 2013;Woodhouse et al. 2017). Another obstacle in implementing SDM is that staff are not confident in clients' ability and willingness to share in decisions about their treatment (Morant, Kaminskiy, and Ramon 2015;Ramon et al. 2017;Slade 2017;Stovell et al. 2016).
Recently there has been a growing demand for the incorporation of SDM into social services provision in several countries, including the USA (Stovell et al. 2016), the UK (UKDH 2012) and Israel (Levin, Gewirtz, and Cribb 2017). SDM in social services is launched as a process in which social workers and clients work together to clarify treatment aims, establish goals and support self-management through sharing information about options and preferred outcomes with the aim of reaching mutual agreement on the best course of action (Coulter et al. 2011). Those promoting of SDM in social services and social policy regard SDM as a method that reflects the core values inherent to the social work profession, such as equality, fairness and just distribution of decision-making power (Braye 2000). Similar values are also voiced in the Global Social Work Statement of Ethical Principles from 2018, stating that 'Social workers work toward building the self-esteem and capabilities of people, promoting their full involvement and participation in all aspects of decisions and actions that affect their lives' (International Federation of Social Workers, 2019.. In Sweden, the National Board of Health and Welfare (NBHW) has stated through National Guidelines 2011 and 2018 that SDM should be offered in the social services' support to people with psychiatric disabilities. In this study, we are focussing on SDM in social services more generally, even if the national guidelines for psychiatric disabilities are sometimes used as a concrete example.
SDM is described both as an 'overall approach' in the work of social services and as a specific method of promoting participation and cooperation in decision making. The recommendation is given even though SDM has been investigated only in medical encounters and there is limited scientific knowledge of what SDM entails in social services practice (Schön et al. 2018). Although SDM is presented as a method that 'should be offered' as an overall approach or working method, the conceptualization of SDM within research and practice is diverse (Socialstyrelsen 2018). Furthermore, the values and actions associated with using SDM in a medical context have yet to be translated so that they can be applied in social work practice (Levin 2017;Munthe, Nykänen, and Sandman 2015;Schön et al. 2018). Knowledge is limited regarding what conditions contribute to the successful implementation of SDM in Swedish social work policy and practice, both generally and specifically when applied to the different areas of social work.
The call for SDM to be offered in the social services thus raises several questions: What does SDM mean, and what components are included in the method and the 'overall approach'? Why should SDM be offered? Who should offer SDM to whom, when and in regard to which decisions in the social services?
The questions concern different aspects of justification and clarification. If there is a call for SDM from a governing authority such as the NBHW, one 1 needs to understand what SDM is, and therefore clarification of both method and 'overall approach' is important. However, this understanding does not provide answers to the question of why SDM should be implemented. This second query has to do with justification in the sense of providing good reasons for the implementation of SDM. If reasons or arguments are provided, these can be evaluated according to strength and relevance.
The last questions, concerning who should offer SDM to whom, when and in regard to which decisions in the social services, are in need of clarification and justification as well. If answers to these questions are lacking or vague, at least two difficulties arise: a practical difficulty for the social workers when it comes to implementation and application, and a theoretical difficulty concerning the evaluation of the reasons or arguments for the call to offer SDM in the social services.

Aim, material and approach
The aim of this article is threefold: I) to discuss SDM and its components, II) to deliberate upon what SDM may imply for social services, and III) to suggest some preconditions that should be present for an implementation of SDM. The last aim will primarily be discussed in terms of the ideological and organizational fit of SDM in social services.
The three objectives are achieved in different ways. The first is achieved mainly via descriptions of SDM found in central literature on SDM (peer reviewed as well as 'grey literature'). 'Central literature' in this context refers to international literature on SDM by researchers as well as steering documents etc. on user participation, for the Swedish social services. The characterizations of SDM as well as the discussions concerning user participation and SDM in this literature, might be conceived of as those 'data' that this article present, interpret and discuss.
The second and third objectives are achieved by an explorative and argumentative approach. That the approach is 'explorative' means that the article discusses some implications that, to our understanding, so far are under-researched. That the approach is 'argumentative' means that the discussion and suggestions are characterized by reasoning and arguments. 2

What is shared decision making?
SDM is a mixture of ideas and practices that gradually emerged in health ethics discussions and that has taken a more concrete form since the early 2000s (Sandman and Munthe 2010). Still, in research literature, SDM is described in many different ways, and in some cases it is not discussed at all (Makoul and Clayman 2006). However, the Swedish NBHW outlines certain conditions for SDM (Socialstyrelsen 2014). First, at least two people are involved and they share all relevant information. Second, everyone must participate in the decision-making process and the decisions must be accepted by all parties. Finally, the process will end with an active decision being taken, even if it is a decision not to do anything. However, the opportunity to achieve a shared decision-making process also depends on the clinical context which includes the relationship and working methods in the service, the occupational responsibility of the provider and the client's preferences (Trevena & Baratt, 2003).
In this article, we will focus on the SDM procedure as it is presented by the Swedish NBHW, since the article discusses the structural and practical preconditions for the Swedish social services to implement SDM. The interpretation of SDM presented by NBHW does not deviate in any significant way from other definitions or understandings of SDM (Elwyn et al. 2013;Légaré 2014).
NBHW describes SDM as a method for user participation where 'the professional and the individual equally exchange information and jointly take a stand for different treatment decisions. The professional contributes knowledge of evidence-based treatment methods, explains their pros and cons and encourages the individual to take an active part in the decisions about interventions. The individual brings to the discussion his or her experiences of living with problems and describing needs, preferences and values' (Ibid., p. 8, our translation).
Here, it should be noted that in Sweden there is a demand that the social services work in accordance with an evidence-based practice (EBP) and the promotion of SDM should be seen in this light. What the professional should do, is to take the best available knowledge and/or research into account, as well as guidelines and available interventions. The third part of EBP consists of the service user's preferences, situation and experience and SDM can thus be understood as an interpretation or version of how to address and/or acknowledge this.
The professional and the individual should engage in an SDM procedure consisting of nine practical steps. Here, NBHW is referring to the procedure presented by Kriston (Kriston et al. 2010). The procedure is similar to other descriptions (see e.g. Elwyn et al. 2013;Légaré et al. 2014): (1) Disclosure that a decision needs to be made.
(2) Formulation of equality of partners. Among social services, there are of course some clients with e.g. disabilities, reduced autonomy, limited language skills, or young age that may reduce their ability to be actively involved. Due to this, various forms of decision support to clients are emphasized as crucial for SDM to be realized (Hamann et al. 2010;Schön et al. 2018).
It should be noted that tensions are inevitable within the EBP model, and also, in the SDM model. For instance, a conception of what the user ideally should be able to choose between, may be muddled by not that many locally available interventions. It is of utmost importance to distinguish the ideal conceptions from more practical suggestions on how to work in a more user-friendly way; while these initiatives perhaps do not live up to a desirable view on user participation, they may nevertheless have as a consequence that the user is more involved in decisions, than before. Participation is in this article considered as a matter of degrees, not a matter of all-or nothing.

Why should SDM be offered?
There are several arguments for the social services to promote user participation, but we will mention only a few here. Well-known arguments concern the promotion of autonomy and the empowerment of users. There are also arguments related to the experiential knowledge or expertise of service users. User participation may also be argued for as a means to foster democratic citizens and as a means for the welfare services to achieve legitimacy in the eyes of the citizens (Dahlberg and Vedung 2001). There are also economic reasons, such as cost-reducing arguments, where user participation is viewed as contributing to making our welfare services more cost-efficient (Eriksson 2015). In the Swedish Social Services Act, we also find several values that, taken together, express a particular perception of what it is to lead a good life; this perception includes components such as meaningfulness and active participation in community life. Participation is thus emphasized in general as well as concerning areas in the welfare system, such as the social services.
Anyone who wants to strengthen the argumentation for SDM (or any other method) in particular however, needs to find reasons that support SDM (or any other method) as the desirable method for individual user participation (Nykänen 2019). Historically, the conceptualization of SDM emphasized the value of including clients' knowledge and preferences in planning and carrying out social services policy as a means to realize civil rights, participatory democracy, antioppressive concepts and social empowerment (Beresford 2013;Ramon et al. 2017). Recent conceptualizations of SDM contain a combination of a sharing of knowledge and expertise, a sharing of power and a contextual awareness that corresponds with both forms of sharing within the professional relationship between users and providers (Levin, 2017).
This article deals primarily with what SDM may imply for social services and what preconditions should be present for an implementation of SDM. Regarding the reasons for a method to be considered as a desirable method for user participation, we consider (at least) three reasons to be of importance when determining what model should be implemented in the social services: it should have been evaluated scientifically, the results of the evaluation should be favourable, and the model's components should harmonize with the norms and values in the legislation (such as a country's social act), a precondition being that the legislation is morally sound (Nykänen 2019).
With regard to the first reason or criterion; that the method should have been 'evaluated scientifically', this can be perceived as a quite strong requirement. However, we do not presuppose a particular evaluative method here, thus tentatively opening up the space for what should be considered as a good enough 'scientific evaluation'.
SDM has been evaluated in randomized controlled trials (Stovell et al. 2016) showing promising evidence that SDM supports users towards empowerment, even if these evaluations so far have been carried out in psychiatric encounters. There is also a certain 'fit' between the SDM components and stated quality criteria in the Swedish Social Services Act, as well as statutory care planning (Nykänen 2019;Knutsson and Schön 2020).

Is SDM appropriate for the social services, and if so, how?
Now, is SDM appropriate in relation to the work performed by social workers and their clients when it comes to decisions about support or interventions? 'Appropriateness' relates here to the space between what the social services, according to central documents, such as legislation and national and local guidelines, should do (and is recommended to do) and what the social services actually do. 'Appropriateness' thus has to do with 'the fitting', in a 'formal sense'; what we are discussing is how well or un-well such things as methods and processes fare when it comes to their relation to legislation, guidelines and practice. Another way of expressing this could be in terms of a 'desirable coherence', that is, coherence between legislation, guidelines, new methods to be implemented and practice is a good thing. 3 Here we consider two broad kinds of 'fit': 'ideological' and 'organizational'. As mentioned earlier in this paper, the core principles of SDM are considered to resemble the core values of social work, namely, striving for equality, fairness and just distribution of power, in this case in decision-making situations (Levin, Gewirtz, and Cribb 2017). It is about creating conditions for cooperation between the client and the staff and that the client should be able to participate in decision-making (Knutsson and Schön 2020;SFS 2001, 453). It may thus be claimed that there is a good ideological fit between SDM and the core values stated for social work or social services. Now, we will discuss at greater length the aspect of organizational fit. A formal decision model, such as SDM may be practically unrealistic if it does not have an organizational fit or if the organization is unable to support it. Such models rarely slip naturally into practice because they carry with them assumptions about the organizational context which may not match the actual setting in which they are used (e.g. Berg 1997). However, we should not ascribe too much weight to the organizational fit since organizations are not static, but can be modified through systematic work. As studies of formal decision models have shown, getting a decision model to work requires negotiation with many elements within the organizational context (ibid), meaning that organizational fit is not given but achieved. Moving SDM to the social services means that we must consider how and where it can fit this new context.

Fitting SDM among distributed decisions
One way to estimate the appropriateness of SDM is by looking at the organizational landscape concerning decisions in the social services; in other words, by looking at the 'decisional topography' (Munthe, Nykänen, and Sandman 2015). Decisions in the social services tend to be distributed across time, people, and different levels of the social service organization (ibid; Björk 2019), which may create challenges for determining when and by whom SDM should be applied. In Swedish social services, several different units make decisions concerning services for a particular user. Often, an investigating unit makes the formal decision, whereas a providing unit may make additional decisions regarding the more concrete content of the services. Additionally, many smaller, step-wise decisions are often taken in regard to a particular service user (ibid). This triggers the question of which decisions SDM should be applied to. It may appear cumbersome to apply SDM -with all its nine steps -across the entire decisional topography.
Another aspect of distributed decisions is the practitioners' ability to make decisions within the social service organization. First, social workers tend to perceive restraints regarding their decisional capacity. A study aimed at supporting implementation of SDM in social services and in psychiatric care illustrates how social services staff in particular feel uncertain when decisions are made and consider themselves powerless to make decisions in their professional role (Schön et al. 2018). This is despite the fact that these staff members were employed mainly as case managers, a role that involves planning, coordinating and implementing actions with the client (ibid). Second, apart from perceived decisional constraints, there are actual legal and material constraints, which may affect the realization of SDM in the social services. Compared with physicians, social services staff are more dependent on their organization in the sense that they are not entitled to make all kinds of decisions that may be relevant in an SDM situation (Evetts 2009). That is, social workers do not make decisions single-handedly, but are dependent on decisions made at other levels of the organization. For example, they can rarely make independent decisions regarding social welfare payments and out-of-home placements because of regulations in the Swedish Social Services Act. 4 Moreover, social workers are also circumscribed by the availability of internal and procured external services within the organization. This illustrates that SDM as an intervention directed at individual social workers is highly dependent on organizational decisions and arrangements. The third step in SDM includes the presentation of treatment options, but in social service organizations with a poor supply of treatments, this means a limitation in the ability to 'share' treatment decisions with the client. It may create situations in which a client requests services that are not available or that the social worker is not permitted to make decisions about. In other words, even if the social worker has the best intentions in line with SDM, the organization sets limits as to how much decisions can be 'shared' with the client. Thus, organizational support in terms of a good treatment supply can be said to be an important precondition for implementing an SDM worthy of its name.
Here, we can actually find support for participation models aimed at the organizational level and how they can act as a complement to SDM: by participating, for instance, in a service user council, users may hopefully be able to influence what treatment options a particular social service unit should offer. The participation on the organizational level thus ideally could influence the degree of participation in SDM on the individual level.

Fitting SDM into the policy level
One way to try to answer the question concerning which decisions SDM should be applied to is to identify what formal decisions are being made in the social services according to policymakers. As noted above, the Swedish Social Services Act's introductory paragraph states that social services, for the sake of democracy and solidarity, should promote people's economic and social security, equality of living and active participation in community life. The services shall focus on releasing and developing individuals' and groups' own resources and be based on respect for human rights to self-determination and integrity. In the preamble to the law, it is stated that voluntarity and selfdetermination, normalization, continuity, flexibility and proximity should be guiding principles in the services (SFS 2001:453 :453).
In order to facilitate the work of social services and the application of the Swedish Social Services Act, the NBHW has developed a flowchart on how to handle the investigation and decision making (Socialstyrelsen, 2015, p. 34, our translation): However, the formal decision ('Make decision') shown in the flowchart under the heading 'Decision making' is not visualized as a decision that the user can participate in. This is probably because such final decisions are formally made by a social welfare board, and the SDM process, rather, is thought to take place during investigation and assessment and in the executing and implementing phase. However, given the weight ascribed to user participation in other documents, this absence of the user or client in the flowchart may be perceived as striking. Even if the flowchart is intended to provide an ideal picture of how decisions should be made, we may add that implementation of SDM also may have such an 'ideal aspect': to include the user not only in special guidance documents on user participation may be perceived as, at least, symbolically important.
NBHW states that SDM may be appropriate in cases where there are several different effective interventions to offer (Socialstyrelsen 2014, 8). On the other hand, in an example, such as the national guidelines for psychiatric disabilities, the confirmation of when SDM is to be offered seems to be more extensive, including the whole decision-making process illustrated above. NBHW there states that the general purpose of SDM is to promote recovery by helping the individual play an active role in making decisions concerning his or her health and life situation (Socialstyrelsen 2018).
This illustrates a divergence between SDM and NBHW's representation of decisions in the social services that needs to be reconciled in order to facilitate an implementation of SDM. If policymakers are unclear about how SDM may fit into their own policies and representations, it is likely that this ambiguity will be reproduced in the social services as well. In sum, what does this tell us about the appropriateness of SDM in the social services? The distributed nature of decisions certainly creates challenges regarding when and for which decisions SDM shall be used, something which is further accentuated by NBHW's flowchart. This is not to say that SDM is inappropriate, but rather that much work remains to be done to make SDM appropriate, for example, by finding its place in the 'decisional topography'. Empirical studies have shown that, concerning SDM, there are implementation gaps of different kinds both internationally (Levin, Gewirtz, and Cribb 2017;Slade et al. 2017) and in Sweden (Schön et al. 2018). Making SDM fit the social services will likely require structural change, in terms of service delivery, as well as cultural and attitudinal change among staff and users (Stovell et al. 2016).

SDM as a method and SDM as an 'overall approach'
In the former sections, we have presented the components of SDM and some of the arguments for user participation in general and for SDM in particular. We have discussed when and in regard to which decisions it could be used. We have also discussed the preconditions for SDM to be introduced into the social services.
This section will address SDM both as an 'overall approach' and as a method. The issue needs to be addressed since SDM sometimes refer to the method including the nine steps, and sometimes to something vaguer, that is, an 'overall approach'. The SDM method is included to convey information about different support options, clarify the person's own preferences and then make joint decisions on treatment or support based on individuals' and staff's knowledge as well as scientific knowledge. A prerequisite for SDM is that the user's preference is obtained and that the user is given information, in accordance to his/her specific needs, on the options that are available. Another core principle is that the decisions taken should not be changed without the involvement of the individual (Slade 2017;Levin, Gewirtz, and Cribb 2017).
But SDM is also described as something that acts as a basic foundation for other interventions offered by the social services, promoting user participation and recovery (Levin, Gewirtz, and Cribb 2017). Examples of such methods are case management, supported employment according to Individual placement and support (IPS), and Housing First, interventions that are all based on users' preferences and active participation. Coordinated care planning is also an effort that is based on users' needs and requires shared decision making. In a comparison with the steps that coordinated care planning should follow, the coherence with SDM is good. From this perspective, SDM can be regarded both as a specific method, but also as a more basic way of working for the social services.
SDM as an 'overall approach' or a basic foundation seems to include the supposition that SDM should be favourable to the work in some way, but in what way? And, if SDM does not necessarily need to take the form of an SDM procedure, like the one presented above (the nine steps), what does it mean in practical terms to say that 'we apply SDM as an overall approach'?
In our view, one of the challenges for the implementation of SDM in social services is to learn how to soundly navigate between two rocky islets. One islet is that the conception of SDM as an 'overall approach' may entail a risk of it collapsing into the 'we already do that syndrome' (Schön et al. 2018). Professionals may regard the call for SDM as an overall approach as somewhat unnecessary since their view is that they already let SDM permeate their work. The other islet consists in following the SDM procedure (the nine steps described above) slavishly in situations where such fidelity is not necessary; this latter course of action or way of working may entail a risk of it collapsing into the 'manual-based tyranny' that many professionals and researchers in social work have reacted against, for instance, when it comes to certain aspects of evidence-based practice.
However, the fact that SDM as a method may not always be appropriate or fitting does not mean that the SDM components can be abolished in the conception of SDM as an overall approach. For the approach to qualify as an overall SDM approach, some criteria need to be fulfilled. 5 Our suggestion is that, for an overall participation approach to qualify as an overall SDM approach, the SDM components should be present. The components are found in the SDM method, including equality of partners; presentation of treatment options and their risks and benefits; identification of parties' preferences; shared decision making, etc. Even if these components are not part of a formal procedure, as long as they are present, then it is legitimate to say 'we already do that'. If the components are not present, the approach does not qualify as an SDM approach, but rather, as perhaps a 'participation approach' in general.
However, establishing that an approach qualifies as an SDM approach does not necessarily mean that this particular SDM approach is the most desirable of the different SDM versions. As human beings, and professionals, we may interpret 'information', 'sharing' and all the other components differently. We may argue that for an SDM approach to qualify as a desirable (or better) version than another version, these components need to be present in a certain way and/or to a certain degree. For instance, do the users or clients have the option of getting the important information they need before meetings with staff? And in what way do the professionals obtain the information they need about the client? And to recap: is there a good treatment supply, so that there actually are different treatment options for the professional to present to the client?
We also need to account for how facilitation of SDM as an 'overall approach' relates to different levels, such as the organizational level (for instance, a particular social services unit) and the level of the individual social worker. If we have the individual social workers in mind, it is not far-fetched to say that the facilitation of SDM as an 'overall approach' has something to do with the attitudes of the staff.
A classic way of describing an attitude is that it has to do with different evaluative responses, cognitive (what I know or hold to be true, thoughts and ideas about a certain object), affective (my emotions, feelings etc.) and behavioural or intentional (I tend to act in a certain way when it comes to the thing the attitude concerns) (Eagly and Chaiken 1993).
Translated into our theme, the attitude has to do with knowledge about SDM, emotions concerning SDM and the tendency to act in an SDM way. The work of implementing SDM as an 'overall approach' when it comes to attitudes can thus be directed at these attitudinal components.
A second point is also worthy of consideration: sometimes slavish fidelity to a certain procedure may be a good thing. As social workers, if we want to make sure that we really do include SDM in our work, perhaps we need to educate ourselves to take all the SDM components into account, and in a desirable way, in relation to fidelity. When Aristotle argued that we learn the things we have to learn by doing them, he was considering the virtues (Aristotle, trans. 2009). However, this approach to learning may be expanded into other areas as well; to be predisposed to take the SDM components into account in daily practice may very well be the result of following the SDM procedure on several occasions. The presence of an overall desirable SDM approach in the workplace may be due not only to the fact that the social workers place a high value on user knowledge and encourage user empowerment, but also to their tendency to actually act according to SDM principles.

Conclusions
The aim of this article was to discuss SDM and its components, to deliberate upon what SDM may imply for the social services, to suggest some preconditions that should be present for an implementation of SDM and in relation to the last aim, to primarily discuss it in terms of the ideological and organizational fit of SDM in social services.
With the help of previous research, policy texts and national guidelines from NBHW, we have presented the SDM components and the SDM procedure. Thereafter, we introduced some of the arguments for user participation in general and for SDM in particular. We concluded that the fact that a method for user participation has been scientifically evaluated, that the results of the evaluation are favourable, and that the components of the method harmonize with the norms and values in the legislation, constitute good reasons for a method to be considered a desirable method for user participation. Although SDM has only been evaluated concerning certain encounters, we consider it at least to be worthy of attention by virtue of its having been evaluated in RCTs.
We then posed the question of whether SDM is appropriate in relation to the work performed in the social services. We presented the notion 'appropriateness' and suggested that this, in this context, is to be interpreted as a desirable coherence between legislation, guidelines, new methods to be implemented and practice. Thereafter, we discussed the issue of ideological fit and organizational fit. Whereas our analysis proposed a good ideological fit, several questions remain regarding the organizational fit. We pointed to the distributed nature of treatment decisions within the social services and a divergence between SDM and policymakers' representation of decisions, something which sparks questions regarding which decisions SDM should be applied to.
In order to support SDM implementation, policymakers need to clarify which decisions SDM should be applied to. Any confusion at a policy level is likely to be reproduced in practice. For practice, there is also a need to think over which decisions are made and where SDM might fit in. As we pointed out, social workers cannot always make decisions single-handedly, and in addition, they seem unsure about their decision capacities. Facilitating implementation may require raising awareness about social workers' decision capacities and their limits. Apart from this, it may also require handling situations in which the social worker cannot make a decision on her own. As SDM involves the social worker, it is likely that situations arise in which the social worker cannot perform all steps of SDM alone, but has to await or is dependent on other decisions being made within the social service organization. This issue still remains to be handled.
We also argued that in order to take SDM and its implementation seriously, the prevailing views on when the user or client should be involved need to be challenged.
Finally, we discussed the difference between SDM as a certain method or procedure and SDM as an 'overall approach', the latter found in the Swedish National Guidelines. We suggested that for a participation approach to rightly be characterized as an 'SDM overall approach', the SDM components need to be present. But for this approach to also be a desirable SDM approach, the components need to be present in a certain way and/or to a certain degree.
The notion of participation (in general as well as in SDM in particular) as a matter of degree, and not an all-or nothing, remains to be more researched. 'A matter of degree' can refer to the scale between ideal SDM and SDM in practice, as well as to the scale 'within' the different SDM components (for instance, the seventh step in the SDM procedure, 'negotiation', can be more or less well performed). The insight that collective models for participation (such as service user councils) on the organizational level, can have a bearing on the degree of participation in SDM for the individual, is another point we identified and that also should be a part of a more thoroughgoing analysis of participation in SDM.

Notes
1. The question of who it is that needs this could be given several answers, for instance: the social worker; the client; the organizational management of the social services; the researcher etc. 2. The fact that the authors come from different academic disciplines (practical philosophy and social work) has implications for the article. Analyses of pros and cons, and reason-giving argumentation (for a particular view, theory, method etc.) is an essential part of philosophical (and normative-ethical) writings, but not as common in many of the social sciences. The article might therefore be considered a 'hybrid discussion'. Such an approach has pros as well as cons; while multi-disciplinary work often contribute with new knowledge for the respective parties, the methods traditionally used in each of the disciplines, may not appear as familiar as usual. 3. Whether the coherence is desirable or not, has to do with the content of the legislation, regulations and/or methods etc. If the content was morally reprehensible, it would rather be desirable (from a moral point of view) that social workers did not follow the rules or implement the new methods. 4. It states that most decisions within the social services are, by default, made by a social welfare board. Some decisions can, however, be delegated to social services staff.
5. This, however, is not the same as a fidelity evaluation concerning the method SDM. What we are searching for here are criteria for distinguishing between what is and is not something that can be considered an SDM approach.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The work