‘How shall we handle this situation?’ Social workers’ discussions about risks during the COVID-19 pandemic in Swedish elder care

Abstract Within a context where New Public Management [NPM] has become increasingly influential in shaping everyday working practices, social workers often handle risks in their everyday work using formalised bureaucratic procedures, among other strategies. As the COVID-19 pandemic progressed, rapid changes occurred in Swedish elder care that social workers were required to address in their everyday work. Intra-professional case conferences amongst social workers provide one opportunity to discuss individual viewpoints and obtain suggestions from colleagues on how to proceed with a case. These discussions have so far received little scholarly attention. In this study we used a data set consisting of 39 audio-recorded case conferences to analyse social workers’ intra-professional discussions about risks during the COVID-19 pandemic. In the case conferences, social workers discussed the risks that were accentuated by the pandemic, such as the risk of spreading COVID-19 to clients, the risk of unmet care needs amongst clients, risks related to accountability, and the risks pertaining to blurred boundaries between different organisations. The collegial discussions in case conferences included opportunities for social workers to use their collective professional experience and competency to establish creative solutions ‘on the go’ and to discuss various ways of handling and balancing different risks while continuing to carry out their work in the changing and unknown situation. Our findings highlight the importance of collegial support in social work in dealing with accentuated risks during the pandemic.


Introduction
Intra-professional case conferences are an important arena for social workers to deal with work-related difficulties and problems they encounter when assessing clients' care needs and eligibility for various care services. Little is known about the content and function of these conferences (Taghizadeh Larsson, Olaison & Österholm, in-press article) but what is known from previous research is that during these conferences social workers consult their colleagues regarding difficult cases, they evaluate their work processes (Bingle & Middleton, 2019;Dall, 2020) and they share reflections, information, and emotions regarding different clients or cases (Forsberg & Vagli, 2006;Riemann, 2005). Discussions and reflections amongst colleagues concerning risks have been described as essential for identifying "risky situations" or potential errors within the organisation (Sicora, 2017). Furthermore, discussions amongst colleagues have been described as valued by practitioners for the opportunity they provide to inform each other about a client's situation and to make a collective assessment of a client's risk status (Hardy, 2017). However, little is known about the specific ways in which social workers, in intra-professional case conferences, discuss and frame the risks they encounter in their everyday work and how they seek collegial support on how to manage or handle these risks.
During the COVID-19 pandemic, established social work practices regarding elder care changed with the intention of protecting older people from becoming infected with the COVID-19 virus. Existing research has shown how elder care services were restricted or even suspended during the pandemic (Banks et al., 2020;Nilsson & Olaison, 2020) and physical meetings between social workers and clients were transformed into telephone or video meetings (Cox, 2020). This resulted in various consequences, such as the difficulties faced by clients to engage in conversations (Levin-Dagan & Strenfeld-Hever, 2020) and ethical challenges regarding privacy and integrity (Banks et al., 2020). Furthermore, studies conducted during the pandemic found examples where older people had rejected or terminated services due to the risk of becoming infected with COVID-19 (Miller & Lee, 2020).
These challenges for social work practice related to the pandemic have arisen within a context of an ongoing transformation from 'welfarism' to 'neo-liberalism' (Parton, 1996), where New Public Management [NPM] has received increased emphasis within policymaking and has shaped and re-ordered social work practices. Throughout this transformation the notion of risk has expanded so that professionals need to consider not only the more obvious risk of danger for clients but also the risks posed to them as professionals, such as being held accountable for their actions or failure to act (Parton, 2017). To handle risks associated with accountability, administrative organisations use formalised bureaucratic procedures where equal treatment of clients (Taylor, 2017) in the assessment process is pursued in line with legislation, organisational regulations, and municipal guidelines (Brodin, 2017;Wittberg & Taghizadeh Larsson, 2021). Nevertheless, collegial support and sharing of professional experience have been described as vital features of decision-making in social work practice as professionals consider how to proceed with a specific client or a case (Bingle & Middleton, 2019;Dall, 2020).
However, much is still unknown regarding the function and importance of intraprofessional case conferences. In particular, there is a need for studies on how social workers discuss and obtain support from colleagues on how to manage and deal with risks in intra-professional case conferences amid contexts influenced by NPM logics. The aim of this study was therefore to analyse social workers' intra-professional discussions about risks they encountered in their everyday work, focusing on the specific context of Swedish elder care amid the COVID-19 pandemic.

Changing understandings of risk in social work practice
Risk has at times been presented as a neutral probabilistic calculation of an events´ outcome of either positive or negative character. Nowadays, risk is primarily associated with negative outcomes of a situation (Parton, 1996), such as harms, losses, dangers, and hazards (Douglas, 1992). Parton (1996) argues that this shift in the understanding of risk is a result of a change from welfarism to neo-liberalism in the relationship between the state and its citizens. Welfarism evolved in the post-war era, with resources in society distributed by the state to guarantee a minimum standard of living for all, based on the status of being a citizen. During this era, '[p]rofessional experts were invested with considerable discretion and trust' (Parton, 1996, p. 99) to assess and cater for various needs and to protect individuals from harmful or dangerous situations. As the change occurred from welfarism to neo-liberalism, the relationship between the state and its citizens also shifted, from a collective membership in the form of a status of social citizenship where the individual had rights and obligations to an individualised and personalised active citizenship. With this shift came the individual responsibility of citizens regarding their consumption of welfare services and the necessity to cater for their own needs. A change in responsibility for professionals also followed, as they were now required to identify vulnerable individuals in need of support and to distribute scarce resources to those unable to provide for their own needs (Parton, 1996).
Similar changes have taken place within the Swedish welfare system. A system of care management was introduced in the mid-1990s with the intention of facilitating the coordination of support for older people (Andersson & Kvist, 2015) and increasing legal certainty (Blomberg, 2004). This process was closely connected to a trend of increased economy and market orientation driven by NPM policies (Christensen & Laegreid, 2001). One of the core elements of the reforms were to expose the welfare services to the idea of competitive governing with partial privatisation, which assumed that the more the welfare state was forced into privatisation, the more efficient the organisation would become (Shanks, 2016). In line with this development, programmes were introduced with the intention of increasing individual choice and control of clients -transferring responsibility for their consumption of welfare services (Szebehely & Trydegård, 2012).
The growing culture of holding someone accountable has been discussed extensively in relation to risks (see, for example, Douglas, 1992;Kemshall, 2010). Because of the risk of being held accountable -not only for one's actions but also for the lack thereof - Parton (2017) argues that the welfare state, through its institutions and representatives, is not trying to make the 'right decision', but rather, a 'defensible decision' (Parton, 2017). To handle the risks related to the accountability of professionals and different organisations in social work practice, formalised bureaucratic procedures and policies are used to guide professional decision-making and to make decisions transparent to clients and members of the public (Taylor, 2017), but also tacit moral judgements on deserving and non-deserving clients could be used as logic for decision making (Warner & Gabe, 2004).

Needs assessment practice in Swedish elder care
Social work practice in Swedish elder care is governed by the Social Services Act [SSA] (2001:453), which aims to ensure a reasonable standard of living. Decisions made in relation to the SSA should respect the principle of autonomy and integrity, and thereby all decisions regarding future elder care services -despite the risk of potential negative outcomes -are to be made by the client. Therefore, formally, social services cannot be imposed on a client against his or her will (Sandberg et al., 2019).
The SSA does not contain any detailed regulations concerning the needs assessment process. Instead, the responsibility for organising and distributing elder care is largely decided at a municipality level, governed by local politicians and implemented by social workers (Brodin, 2017). In Sweden, the responsibility of providing health and care services for older people is divided between different authorities (municipalities and regions) (Dir. 2020:142). Swedish elder care services are tax-subsidised and provided to eligible citizens through a needs assessment process (Moberg, 2017). An assessment of the client's needs is undertaken and information is gathered from the client, significant others, and representatives from other health and care agencies. In most municipalities, social workers are guided by local guidelines where municipalities have specified services catalogues that list services that can be provided (Wittberg & Taghizadeh Larsson, 2021).
The assessed need of the client is often negotiated between the client and the social worker to fit available services (Olaison, 2017). During the assessment process, social workers often meet with their colleagues in intra-professional case conferences; that is, conferences where social workers who work at the same local social work agency meet to discuss, consult and obtain support from their colleagues on how to proceed with a case or how to handle challenges in their work practice. Case conferences are internal meetings amongst colleagues, and hence no clients or representatives for clients participate. These conferences differ from informal consultations or hallway discussions, as they are scheduled on a regular basis at the local social work agency.

Data collection
This study is based on 'naturalistic data' (Potter & Shaw, 2018) consisting of 39 audiorecorded intra-professional case conferences. Naturalistic data include 'authentic' forms of speech that would have occurred regardless of whether they were subject to research or not (Potter & Shaw, 2018). Naturalistic material preserves the studied phenomenon itself, which makes it possible to study how a phenomenon occurs in situ (Potter & Wetherell, 1987) with fewer problems pertaining to artefacts being produced in the data due to the presence of a researcher.
Data were collected in four Swedish municipalities located in different regions as the COVID-19 pandemic developed (from April 2020 to April 2021). An initial contact was established with the operating managers at these local social work agencies to inform them about the aim and method of this study. All operating managers gave consent so that we could inform the social workers at their offices about the present study and ask if they were willing to participate. Information was then given to the social workers verbally in a meeting, and in written form. They were informed about the aim and method, what was expected of them, their right to decline participation, confidentiality and so forth. Written informed consent was given by all social workers individually before the data collection began. When the social workers were informed about the project, we urged them not to name or give personal information about clients or other professionals that could be tracked back to a specific individual. Occasionally, personal information was provided; this information was erased from the data and pseudonymised in the examples presented in the Findings section.
The initial plan was that a member from the research group should be present at all case conferences to conduct observations and handle the audio-recorder. Because of the pandemic, these local social work agencies did not allow visitors. One agency had converted their physical case conferences into videoconferences. Technical barriers to ensure confidentiality prevented anyone other than employees at the agency from participating in these videoconferences. Hence, the procedure of collecting data had to be changed. The group leaders responsible for leading and organising the case conferences were provided with an audio-recorder. Thus, the group leader decided which conferences to include, and when to begin and end the audio-recordings. All audiorecordings were transcribed verbatim by a professional transcription firm.
All names of social workers, clients and places in the present paper are fictitious to ensure confidentiality. An ethical review of this study was carried out by the Swedish Ethical Review Authority (Dnr. 2019-05965).

Participants
In total, 49 social workers participated. Their experience in adult social work ranged from one month to 31 years (average 5.5 years).
The audio-recordings of the 39 case conferences ranged from three to 106 minutes (average 33 minutes). In total, 137 cases could be distinguished in the transcriptions. We present a more detailed presentation of the collected data in Table 1.

Data analysis
The transcriptions from all audio-recordings were scrutinised to identify and extract data sequences where COVID-19 was mentioned or referred to. Thematic discourse analysis (Braun & Clarke, 2006;Singer & Hunter, 1999) provided the methodological framework for analysing the material (Potter & Wetherell, 1987). When conducting this type of discourse analysis (Singer & Hunter, 1999), the analysis focuses on identifying features and underlying ideas in conversations to create the meaning of the studied phenomenon (Braun & Clarke, 2006). The conversational context (such as utterances before and after the analytical claim that were needed to understand the sequence as a whole) was incorporated with all data sequences to make sure that utterances were analysed in  (Braun & Clarke, 2006). In total, 92 data sequences were extracted from the 39 case conferences. All 92 data sequences were included in the analytical process. The analytical process was data driven. All extracted data sequences were read repeatedly and discussed by all authors in relation to how social workers discussed the risks they encountered in assessing the needs of clients during the COVID-19 pandemic in Swedish elder care. The first author condensed and coded all the data sequences. Condensation of risks discussed, for whom, the consequences thereof, and how to handle these risks was conducted to summarise content relevant to risks brought up by social workers and how they had discussed this in relation to their practice. All sequences were condensed and coded with one or several different codes. These codes were used to label differences emerging in the data relevant to the social workers' discussions about risks. As the analytical process was data driven, codes were given based on the sequence's contents. Examples of codes used are 'fear of COVID-19', 'responsibility', and 'unmet care needs'. From the condensations and codes, risks that social workers discussed were grouped into four themes, which are presented in the Findings section. To corroborate the saliency of the themes and increase reliability and consistency all authors discussed the themes and revised these until consensus was reached of what themes were the most pertinent to capture risk discussions in the collected data. The analysis of the original transcripts was conducted in Swedish. The data fragments presented in the Findings section have been translated by a professional translator.

Findings
We structure our findings section around four themes concerning risks that social workers discussed during the case conferences. Two of these themes revolve around risks relating to when the social workers met clients, and the different difficulties they encountered in assessing their care needs. The other two themes pertain to risks of a more organisational nature, relating to accountability and boundaries between different organisations in Swedish care for older people.

The risk of spreading COVID-19 when meeting clients
The social workers discussed the risk of spreading COVID-19 when meeting older vulnerable clients in their everyday work. To reduce the risk of spreading the virus when meeting clients, the social workers we observed in case conferences discussed solutions for adjusting the assessment process in accordance with the rapidly changing situation and local and national guidelines. These social workers suggested some changes that were designed to reduce the risk of spreading COVID-19 when meeting clients; for instance, rather than visiting a client in his or her home, telephone or video meetings were considered as alternative ways of conducting assessments. If the client was deemed to need a face-toface meeting, to facilitate participation in the assessment process or for the social worker to complete the assessment of the client's care needs, the social workers suggested alternatives such as meeting the client in specific rooms at the residential care facilities, or speaking through windows, over fences or on benches outdoors during their discussions. If the meeting was to be conducted indoors, they discussed how to use personal protective equipment. The potential risks or consequences of these changes, such as clients being unable to go outside due to physical restrictions or to find their way back home after a meeting due to cognitive impairments, were also raised in relation to these changes.
In the following example, the social workers were discussing together how to reduce the risk of spreading the COVID-19 virus to clients and at the same time how to organise assessment meetings in accordance with the client's abilities. Central to the example is the collegial scrutiny regarding compliance with local and national guidelines. Furthermore, this scrutiny is raised in relation to both local and national guidelines implemented to reduce the risk of spreading COVID-19 within elder care. In the example above, the social workers were negotiating alternatives together as to circumvent both local guidelines (to conduct assessments through video meetings) and the national curfew at residential care facilities implemented to protect older people from COVID-19 by reducing the number of visitors. SW1, responsible for the case, argued for a meeting in person with the client to gain a sense of how to proceed. This was questioned by SW2, who suggested that the assessment should be carried out through a video meeting. SW1 stood her ground, however, sharing her stance about the client and what would be in his best interest, as she stated that it would be difficult for him to participate in a video meeting. As an alternative, SW3 suggested a meeting through a window. This was also questioned in relation to the national curfew. To circumvent the national curfew, in order to meet the client in person, SW5 suggested that he could be taken outside by staff at the care facility and that the social worker could make sure to keep a physical distance between them (between the client and the social worker) to reduce the risk of spreading the COVID-19 virus to the client. This approach was finally approved by SW2.
In the remainder of the excerpt, we can see a discussion among the social workers of the practical logics regarding how to undertake that specific meeting. The sequence indicates that the discussions in case conferences could serve an important role by allowing collegial scrutiny regarding how to conduct one's everyday work in accordance with both national and local guidelines when meeting clients to reduce the risk of spreading the virus but, concurrently, how to work around these same regulations in a creative way. An underlying concern apparent in the interaction above was the risk of unmet needs, to which we now turn.

The risk of unmet care needs
Apparent in the case conferences in our sample were discussions among social workers regarding the risk of clients' living with unmet care needs. In these case conference contexts, social workers discussed an initial decline in the number of clients applying for support, because of the risk of becoming infected with COVID-19 when using services. Social workers also described that they encountered clients who, for fear of becoming infected, and the consequences they might suffer, declined or terminated contact with services so that they could practice social distancing and thereby reduce their risk of infection. Consequently, the implicit risk raised here was that of unmet care needs. In the excerpt below, the social workers discuss a client who had declined support, despite the presence of care needs that should have been catered for through home care: SW1: What was it with her? She was pretty ill there. But she wanted food distribution because she did not want a lot of people visiting her and infecting her because she had been admitted. I do not remember why she was in the hospital. But it is hard for her to accept support too. So this was such an approach. GL: Which seems to work pretty well, then? SW1: It has worked great, apparently. SW2: So it works really well. SW1: Should be able to continue for some time so that she uses the services. Because she has care needs, according to the son; she needs help, but she has a little difficulty to accept that she needs help, and then it has to do with the fact that she does not want to let people into her home, which is good.
In the example above, the social worker discussed a client who had accepted some services but declined support in the form of home care, to reduce the risk of becoming infected with COVID-19. This provides insight into how social workers balanced recommendations for social distancing, to safeguard older clients from COVID-19, with clients' assessed care needs in case conferences during the pandemic. In the example, the client was described as exercising her right to self-determination by refusing to grant home care professionals access to her home, with the potential consequence of not having her care needs met. Furthermore, SW1 suggested that the client was practising social distancing and thereby complied with current regulations, which was judged as a responsible act, but that she probably had care needs that were not being attended to because of this.
Another issue raised in the discussions in the case conferences was that some clients who continued to request services were described as reluctant to meet social workers face to face during the pandemic to avoid being infected with COVID-19. Social workers expressed that when they were unable to visit clients, it was difficult for them to assess the person's care needs as they had to rely on spoken information in the needs assessment and could not observe the client or take into consideration the client's home environment in the needs assessment. Thus, in the case conferences there were discussions about whether there was a risk -especially if the client was applying for the first time -that the person's actual care needs were not being catered for as the assessment procedure was not conducted in the same way as before the pandemic. looks like they live in an apartment, it is a bit hard to access, and he will not give any more information.
In the example above, the social workers discussed a situation where they had assessed that a care need existed but, due to the client's unwillingness to meet the social worker during the pandemic, they could not assess to what extent, or how to cater for that care need. To complicate things further the client was not previously known to the social worker.
In contrast to these examples above, in the discussions social workers also referred to occasions when clients and/or significant others who, despite being vulnerable to the COVID-19 virus and needing to follow recommendations regarding social distancing, stressed their need for services that had been suspended during the pandemic. These services had been suspended either to minimise the risk of spreading the virus forward or to prioritise available resources elsewhere. Still, a few significant others were described in the case conferences as insisting that the needs assessment process should be carried out as usual despite the risk of negative health consequences associated with COVID-19. Hence, in discussing this matter, social workers balanced risks associated with COVID-19 with the risks of clients living with unmet care needs, with consequences such as worsened health, physical functioning, or well-being. In the example below, the social workers discussed a client who argued for access services despite potential negative outcomes regarding COVID-19.
SW: She is a nice lady that cares about her abilities and explained to me last time that she had to walk, she had to exercise her legs, or else she would be in a wheelchair and she did not want that. And she noticed that when she had been locked down quite a bit because of Corona, that her legs could not carry her the same way and she was afraid [. . .] Well she has to go out, she has to exercise. [. . .] She can walk up the staircase but she cannot go down the stairs.
In this example the social worker suggested that the client requested services despite the risk of becoming infected with COVID-19, because not being able to exercise her legs would have negative consequences for her continued mobility in everyday life. It also illustrates how social workers, during case conferences, requested support from their colleagues on how to balance the regulations implemented within elder care to ensure social distancing with the client's arguments for using services to address her care needs. Hence, the social workers in our study could often be heard discussing social distancing in relation to the client's right to self-determination and the right to make choices about her own life, despite the risk of negative outcomes. This weighing up of different risks and rights took place amid a wider context of being held accountable amid NPM frameworks.

The risk of being held accountable
Another central theme that the social workers discussed in the case conferences was the risk of being held accountable. Decisions to suspend, restrict, or reorganise services to protect clients against the risk of spreading COVID-19 to each other were made at a strategic level in the municipality organisations. In the case conferences, social workers discussed the risk of being held accountable for these decisions, seeking support by discussing with their colleagues how to proceed, and expressing feelings of vulnerability to criticism from clients, significant others, other healthcare staff or the public, as they could not grant services in accordance with assessed needs or conduct needs assessments as usual.
In the example below, a social worker elaborated on the risk of encountering clients requesting services that had been restricted, due to a strategic decision to prepare the municipal organisation for an initial surge of COVID-19 patients during the pandemic. The importance of action by operating managers to relieve pressure on the social workers was emphasised: SW: Should the physiotherapist make her assessment now, she is not fully rehabilitated yet at the short-term care facility. She really needs to be there for the rest of the month. Uh, at the same time so. yeah and I said like I said 'we are going to vacate the facility' and as I said 'I will extend her stay' so I talked to both the daughter and the wife on Thursday, but I said 'we must prepare for the service (at home) if they must admit other patients  well then this must be fixed'. They do not want to receive help at home. They are not there yet [. .

.] We have given all information possible from our profession. Here operation managers are needed to really give information that this is what it looks like. Because there has been no general information given to the municipality's residents that it is like this [. . .] So we said that 'it is best if it comes on a higher level' because we are completely powerless. And they will get the information from all professions, this is what it looks like, this is what the municipality has decided and these are the consequences.
As can be seen in the example above, the social worker tells her colleagues about a client who needs further rehabilitation to manage in her ordinary housing. In this municipality, a decision had been made to vacate short-term care facilities to prepare care places for a potential surge of COVID-19 patients. Due to these changes, the social worker, in frustration, described a situation where she was caught in between the regulations decided higher up in the hierarchy and the client's wish to be rehabilitated at a shortterm care facility. This was depicted as a frustrating situation, as both the client and her significant others were portrayed as unwilling to accept changes made in relation to the COVID-19 pandemic, and who were described to some extent as being unreasonable, as they were demanding services as usual. The social worker used the case conference as a forum where she urged her operation managers to relieve pressure on them by publicly giving information and justifying the reorganisations made in relation to the COVID-19 pandemic. This communication from higher up in the organisation would mean that the social workers would not have to 'carry the can' for decisions taken elsewhere in the care organisation.

The risk of blurred boundaries between different organisations
In negotiating the multiple pressures and demands noted above, the social workers discussed the risk of blurred boundaries between different organisations and their respective responsibilities in the case conferences. Social workers discussed boundaries encountered between different organisations and talked about the consequences of not working in close collaboration with other organisations. Information-sharing about a client's COVID-19 status was depicted in the case conferences as challenging in relation to integrity and confidentiality, as, in Sweden, there exists a division of responsibility between municipalities and regions for care (municipalities are responsible for care and some health care administrated in the person's home) and healthcare services (such as primary care and acute health care). In some of the case conferences, it was articulated that information about health conditions and diseases fell outside the responsibility of social workers, as their objective is to assess care and service needs in relation to municipal elder care, and not to assess the healthcare needs that could be related to primary care or hospitalisation. However, information about a client's COVID-19 status was discussed as important, to avoid the risk of placing a person with COVID-19 at the 'wrong' care facility, thereby spreading the virus amongst vulnerable individuals. Furthermore, social workers at the case conferences described a lack of relevant medical knowledge to assess the seriousness of a client's disease or whether a client was infected or not with COVID-19. This uncertainty therefore brought up the necessity of a forming collaborative relationships with healthcare personnel. That is more of a medical concern, we cannot always tell that. But there has to be more like a nurse in the community medical service that decides that, somehow.
In the example above, the social workers were discussing and negotiating their responsibility and mandate to determine whether a client was infected with COVID-19 or not. In the sequence they discussed how to obtain information about a client's COVID-19 status, information described as important to reduce the risk of admitting clients with COVID-19 to care facilities. Furthermore, the social workers negotiated which questions they, as social workers, could ask about their client's health condition. In this collegial discussion, they moved to conclude that they, as social workers, could only ask questions of a more general character, but in such a way that they could distinguish whether the client had healthcare needs that required being addressed by health professionals. The social worker further argued that if there was an underlying medical condition, they lacked medical knowledge and thus would have to involve the community medical service.
During the case conferences, social workers also turned to their colleagues concerning issues related to boundaries between different organisations, and potential risks, as collaborations were described as difficult or more challenging amid the COVID-19 pandemic. Apparent in our case conference data were examples where social workers discussed their mandate and responsibility in relation to the organisation's responsibility and purpose, both from their own professional perspective and from the perspective of staff working at other organisations. Thus, to provide the client with the right support and tackle risks in relation to COVID-19, there were discussions about how to increase collaboration with healthcare professionals and providers of services in order to benefit from different professionals' competencies. In the example below, the social workers discussed a certain occasion when they lacked support from community medical care, as the client had not yet been assigned to that organisation. In the example above, SW1 described a previous situation where she needed support from the community medical services to make a medical assessment of a client's COVID-19 status. Here, SW1 sounded frustrated while telling her colleagues about a previous situation, where the nurse declined to help and work with her as the client was not assigned to the nurse's organisation. The nurse in this context was described as 'square' as there was no room for manoeuvre or pragmatism to collaborate with SW1. The social worker added that different organisations/agencies within the elder care system need to help each other out, work together and use each other's professional competencies to assess the clients' health status so as to safeguard clients during the COVID-19 pandemic. The boundaries between the different organisations were discussed, in this instance above but in several other instances in our data, as creating situations of risk that the social workers had no mandate to influence.

Discussion
In this study we aimed to analyse social workers' intraprofessional discussions about the risks that they encountered in their everyday work during the COVID-19 pandemic in contexts of Swedish elder care. All the identified themes, to some extent, touched upon the risk of social workers being held accountable for their actions or of not acting, which in turn limited the social workers' discretion towards clients when making individual professional judgements regarding how to continue their practice during the pandemic; such as, for example, how they balanced recommendations for social distancing with clients' requirements for support to address specific care needs, or how collegial scrutiny was used to discuss how to act in various situations.
The types of risk discussions presented in the findings above pre-date the pandemic as they were to some extent enduring and already known. However, the specific context of the pandemic makes some of these risks, uncertainties, organisational boundaries and defensive practices more explicit and we have sought to illuminate these through their discussion in case conferences. These case conference discussions remain underresearched in the literature and show very clearly how social workers cope with risk and uncertainty by accessing collegial advice and support, not least in how to handle a new and drastic situation occurring when data was collected (April 2020-April 2021) when much was still unknown regarding the COVID-19 virus.
In previous research about social work practice, protective measures and controlling the functions of individuals have been described as reducing risks and protecting clients in elder care (Bornat & Bytheway, 2010;Taylor, 2006). In our findings, various aspects of risks were navigated by the social workers as they sought to find solutions for how to reduce the risk of spreading the virus while continuing to carry out their work of assessing older clients' care needs. For example, social workers discussed that assessment meetings might be conducted outside, through digital means or in specific rooms. Similar solutions for reducing the spread of the virus have been described in previous research about COVID-19 and social work practice (Cox, 2020;Nilsson & Olaison, 2020). Our findings suggest that measures of a protective nature, as discussed by the social workers in the case conferences, were primarily a result of changes in guidelines created at a strategic level in the municipalities' care organisations to address the COVID-19 pandemic (for example, placing infected older people on certain wards, and suspending or restricting services), and this was something that the social workers had to take into consideration when carrying out their work.
These findings reflect Kemshall's (2010) arguments regarding the linkage of situated rationality with cultural norms and power structures in institutional risk assessments, whereby choices are often constrained. It was evident from our findings that the social workers, during the case conferences, together and as best as they could, tried to navigate the various risks they encountered in their everyday work during the COVID-19 pandemic, combined with exposure to new guidelines produced at a higher level of the organisation. This included discussions on situations that became difficult for them to handle, such as how to assess clients' care needs and how to cater for them, which could have potential negative outcomes. Thus, the obvious risk of clients living with unmet care needs was brought up recurrently in the case conferences.
In Sweden, multiple professionals and organisations are involved in the provision of elder care services and hence, to avoid fragmented care, representatives from different organisations are requested to collaborate with each other and coordinate their services to clients (SOU 2020:80;Sundström et al., 2018). The importance of interprofessional collaboration and coordination of elder care services has been discussed previously but continues to face several challenges (Backhouse et al., 2017). Our findings suggest that social workers encountered and struggled in dealing with blurred boundaries between organisations as well as within the same agency during the pandemic, and this hindered collaboration, leading to different potential risks, such as spreading the COVID-19 virus amongst vulnerable clients or misjudging clients' medical status. Furthermore, discussions about increased tensions between social workers and representatives from other organisations could be seen in the findings, which might affect collaboration and coordination negatively. To handle potential risks related to the virus and infection, a closer collaboration between various professions with different competencies is stressed as part of the findings of this study.
The neoliberal turn in social work practice has increased the risk of professionals being held accountable for their actions or the lack of them by the organisations they represent (Lymbery & Postle, 2015), where bureaucratic formalised procedures are applied as part of making defensible decisions (Parton, 2001). Hence, researchers have argued that the discretion of individual professionals is limited in line with these procedures. The findings from our study concerning social workers' intra-professional discussions suggest that changes occurred as a result of the pandemic, both in relation to the municipality organisations (altering services, changing the assessment process) and the clients' actions (declining support or assessment of their care needs). To handle these changes and the related uncertainties, social workers discussed with their colleagues how they would act in accordance with new and changing guidelines, but also used their shared discretion to find new creative solutions that would allow them to provide services in accordance with the clients' care needs.
Previous research has described the function of consulting or evaluating the work process with colleagues at a case conference (Bingle & Middleton, 2019) or in team meetings (Dall, 2020). Our findings add to this the function of collegial scrutiny at intraprofessional case conferences where social workers discussed their potential actions regarding how to handle an individual client's case. Here we stress the importance of collegial discussions during a pandemic as something more than identifying 'risky situations' or potential errors within the organisation (Sicora, 2017), or having oneʼs colleagues' judgement on an assessment or decision (Hardy, 2017). We argue that collegial discussions in case conferences also include the opportunity for social workers to use their collective professional experience and competency to establish and legitimate creative solutions 'on the go' (such as how to assess care needs, which services to grant, how to circumvent local and national guidelines) and to discuss various ways of handling different risks while continuing to carry out their work in this changing and unknown situation. The social workers in our study did this while simultaneously scrutinising their potential actions to handle the risk of being held accountable for their use of discretion.

Limitations
There are limitations with the present study. The spread of COVID-19 has varied between different municipalities in Sweden. Data collection occurred in municipalities one and two at the same time as the 'first wave' of COVID-19 emerged. Both of these municipalities prepared for a surge in COVID-19 cases, but only the first was severely affected during the data collection phase. In contrast, data collection in the third and fourth municipalities occurred when the spread of COVID-19 was relatively low in Sweden and, consequently, fewer relevant data fragments were extracted from these municipalities. Data collection in the fourth municipality occurred when the spread of the virus had once again increased but, in contrast to the first two municipalities, it seemed that the initial lack of understanding and fear of COVID-19 amongst social workers and older people in need of social services had settled.
The data corpus from municipalities three and four mainly concerned talk about clients. In the first and second municipalities the data corpus also included a general discussion about services and the organisation of services during the pandemic. We attribute these differences to various reasons, such as group leaders deciding when to begin and end the recordings, closer proximity to the beginning of the pandemic, or different purposes of these case conferences. What occurs in intra-professional case conferences and how they are organised in (older) adult social work is still an underresearched practice that is in need further scholarly attention.

Conclusion
Intra-professional case conferences were an important arena for collegial consultation for social workers to discuss how to proceed with a client or a case as changes occurred continuously in Swedish elder care during the pandemic. In their discussions, social workers brought up risks that were accentuated by the pandemic, such as the risk of clients living with unmet care needs, or how to reduce the spread of COVID-19. These findings add to our understandings of the importance of social workers addressing the risk of being held accountable for their actions, or for not acting, by exploring the dynamics of how this handling of risk was undertaken together with colleagues in intraprofessional case conferences, as these social workers negotiated a rapidly changing practice regulated by local and national guidelines. While the literature on professional work amid neo-liberal governance and new public management has tended to emphasis professionals coping on their own, our findings also indicate that increased demands related to accountability, as part of a transformation from welfarism to neo-liberalism, may impact on social workers' (in)actions (Parton, 2017) in unknown, critical situations by making them less inclined to act individually and rapidly.