Designated officials’ approaches to judging serious incidents: an analysis of incident reports in care for older people, disability services and family services in Sweden

ABSTRACT There are various systems for reporting incidents that have harmed service users depending on the country and legislation. In Sweden, the duty of staff to report mistreatment covers three areas: care for older people, disability services, and individual and family care. This study examines Designated Officials’ judgements concerning reported incidents of serious mistreatment in social services in Sweden. A qualitative analysis of incident reports covering all three areas shows that investigators used five approaches to judge the seriousness of an incident. The ‘individual approach’ is seen as a blaming strategy of individual staff, and the ‘quality development approach’ highlighted systematic failures, with little focus on protecting individual service users. The ‘medical approach’, predominantly used in care for older people, measured physical injuries but neglected service users’ own experiences. The ‘empathic approach’ incorporated service users’ perspectives as interpreted by staff, often resulting in service users being stereotyped as victims. The ‘legal and human rights approach’ was least common in the analysis, despite echoing the core values of social work that are highly important for social workers. We conclude that the assessment process for reports needs to be reconsidered and the meaning of harm in the context of social work re-evaluated.


Introduction
In social care and social services there are multiple systems for reporting serious incidents that have caused harm for at-risk service users, and many European countries align their reporting procedures with legislation. It is mostly various forms of abuse and neglect that are reported and acted upon (Preston-Shoot, 2016). The need to define a seriousness threshold for practitioners has been pointed out in international research (Manthorpe & Martineau, 2011), and the question of 'how poor does poor or neglectful practice have to be for adult protection systems and processes to apply?' (Collins, 2010, p. 5) has been raised.
The core values and ethical codes of social work, such as social justice, human dignity and selfdetermination, should permeate all social workers' judgements (Bisman, 2004;Reamer, 2018). However, research points to a range of factors that influence social workers operating as Designated Officials in threshold decisions, for example the type of abuse and whether the service user consented to the proceedings (Trainor, 2015). Platt and Turney (2014) highlight factors such as policy, organisational circumstances and pressurised working conditions as significant in investigating officials' decisions, but above all they stress the consequential features and context of the social work services. Besides exploring factors that influence decision making, researchers have analysed how decisions are made in social work. To arrive at an adequate decision after sorting through large amounts of information, investigators have to use knowledge from various sources, making use of their discretionary power (Samsonsen & Turney, 2017).
Designated officials' definitions of risk, available resources and other external factors lead to thresholds described as 'fuzzy junctures', which have been addressed in child protection (Keddell & Hyslop, 2020). The difficulties of making threshold decisions also apply in Sweden where there is a specific duty for staff working in social services, including disability services and care for older people, to report mistreatment. The legislation has been criticised because of the ambiguity over what they are supposed to report (Kjellberg, 2020) and investigating officials have asked for more advice on how to assess degrees of seriousness to arrive at a judgement. The Designated Officials investigating incident reports in Sweden are mostly trained social workers employed by local municipalities. As such, they are accountable to the local municipal board, which authorises investigations and reports to the regulatory body. Throughout this paper we refer to the above as investigators.
This article aims to provide further understanding of how Designated Officials determine the seriousness of an incident, drawing from an explanatory analysis of investigators' judgments of serious incidents in social services in Sweden.

Context
In Sweden, the specific duty for staff to report mistreatment first appeared within care for older people and was incorporated in the Social Services Act (SSA) in 1999. In 2005 the duty to report was expanded to include disability services. Since 2011 the provision has applied to all social services, covering three areas: care for older people, service and support for people with disabilities, and individual and family care. Care for older people encompasses home care, municipal daily activities for older people and residential care homes for people over 65. Disability services consist of daily activities for people with varied degrees of disabilities, personal assistance, group homes, and service homes. Individual and family care includes voluntary support for a range of different services, for example financial assistance, social care for children and young people, shelters for homeless people, substance abuse and dependency care and support for victims of crime. Compulsory care for young offenders and people with substance abuse issues is carried out in residential homes run by the Swedish National Board of Institutional Care. This government agency uses the same incident reporting system but is not included in this study.
Incident reporting is mandatory for staff working within the social service agency and others who have appointments within the agency or complete tasks on behalf of the service user, such as case managers. Reports are therefore made by all kinds of staff, not only social workers. Reports can be initiated by complaints from service users or their next of kin, but only staff can file reports. However, service users can file complaints directly to the management or to the Health and Social Care Inspectorate (HSCI). The social service agency is also obliged to investigate all reports from staff. If an investigator concludes that a reported event has (or presents a risk of) serious consequences for an individual's life or physical or psychological well-being the investigation is sent to the HSCI, which can take disciplinary action against social services. The aim of the regulation is to protect at-risk service users, amend and improve services and enable staff to report all mistreatment without risk of reprisal, thus increasing quality of care. However, the regulation has undergone several changes over the years and has been discussed in terms of a change from a rights-based discourse, in which staff have a duty to safeguard the rights of at-risk service users, to a discourse of obligations in which staff are part of systematic quality assurance (Kjellberg, 2020). Investigations have been criticised for not addressing structural shortcomings and an inability to align amendments with causes (Björne et al., 2021). Across Sweden there are approximately 1200 serious incidents reported to HSCI each year (HSCI statistics, 2013(HSCI statistics, -2021, most of which concern care for older people, which is also the largest area.

Materials and methods
This analysis is part of a larger study on incident reporting by staff working in care for older people, disability services, and individual and family care in Sweden. The larger data set consists of all reports (n = 1105) by staff within these three areas in the region of Västra Götaland over one year (2019). While the analysis was informed by the reading of all 1105 reports in the larger data set, the analysis presented in the following paper is specifically based on 97 (of 156) incident reports that were deemed serious by investigators: 41 relating to care for older people, 28 to disability services and 28 to individual and family care. The sample was varied, covering small and large municipalities and different types of incidents.
A serious incident report typically includes a description of the incident itself followed by a path trajectory of events and problems leading up to the incident with some or all of the following features: interviews with involved staff and service users; descriptions of relevant policy documents, laws and routines; a description of the work experience and education of the staff involved; and the consequences of the incident for the service user(s) involved. The assessment aims to identify contributory factors, explaining what caused the incident and determining whether it is likely to happen again. The investigator suggests actions needed to avoid further incidents, for example new routines or information for staff. Lastly, the investigator compiles the collected evidence to summarise the causes and consequences of the incident and make a judgement of its seriousness. The focus of the analysis in this article is the executive summary and judgement of the incident.
The incident reports vary greatly in quality and length, from four to approximately 30 pages each. Some of the investigations involved several interviews, including with the victim, various documents and a clear and detailed judgement of the incident's level of seriousness. Others were less detailed. In most of the reports it was possible to identify factors in the incident that the investigator considered important for the final judgement of its seriousness.
The study was approved by the Swedish Ethical Review Authority, number . In Sweden, incident reports are subject to public access and municipalities must hand over reports on request. Municipalities must remove all personal identifiers and have the right to withhold reports if they may compromise the confidentiality of anyone involved. There was only one occurrence of a report being withheld due to confidentiality. We asked for gender breakdowns but only a small number of municipalities responded to this request. The reports are sometimes written using gender-neutral language so as not to disclose the gender of the victim. In addition, most municipalities also withheld the names of areas and care services. In the quotes presented in this article all names of people, places and specific organisations (such as care homes) have been removed.

Analytic framework and analysis
We carried out a thematic analysis from a social constructionist perspective (Braun & Clarke, 2006). The themes presented constitute descriptive categories as well as discursive framings of incidents and harm. We are following Biesel and Cottier (2020, p. 18), who underline that all deviance within social work, such as 'errors and mistakes', are discursive constructs shaped by, e.g. 'social expectations, legal regulations, professional values, norms and attitudes as well as quality standards'. In addition, discursively constructed categories such as harm and abuse are ambiguous and lack clear-cut boundaries (Gee, 2005). When investigators assess what should be deemed a serious incident or harm, they must fill these categories with situated meaning, which is understood in its specific context. We pay attention to different categories of service users as contextual differences. For example, in modern Western societies there is an expectation that special protection and care be provided for children.
Analytical tools were borrowed from discourse analysis, an analytical approach that explores how discourses are used flexibly as a form of action/negotiation in conversations (Potter & Wetherell, 1987). Margaret Wetherell (1998, p. 400) defines a discourse as a form of 'interpretive repertoire': 'a culturally familiar and habitual line of argument comprised of recognisable themes, common places and tropes (doxa)'. In other words, interpretive repertoires are situated discourses used as framings, or methods for making sense, and to organise accountability, that is to present an argument others recognise and accept as true. We focus on how care staff and social workers are appropriating the duty to report, drawing on different discursive resources linked to social work. Thus, investigators are not simply executing policy according to existing guidelines, they are constructing the meaning and threshold for a 'serious incident', including what can be considered serious consequences for service users and who can be blamed for an incident. We summarise these framings as the investigators' different discursive approaches.

Thematic analysis
The thematic analysis was conducted in six phases (Braun & Clarke, 2006). The first, familiarisation with the data, and second, generating initial codes, were conducted simultaneously. Both authors were familiar with the data having previously read and coded all the reports. A total of 60 reports were copied into the analytic computer programme NVIVO and thereafter read, reread and analysed inductively. Multiple notes and codes of minor details were created in the computer programme.
In the third phase, searching for themes, broader themes and subthemes were created, all of which related to factors that investigators highlighted as important for the final judgement of seriousness: actions leading to an incident, faults within the organisation, physical and mental consequences, legal rights of service user(s), dignity and reputation of the organisation.
In the fourth phase, the themes were revised and reduced. and 37 further reports were analysed to provide greater resolution to the themes and confirm their relevance. The fifth and the sixth phases were conducted iteratively. Themes were further revised and analysed simultaneously during the process of writing up. The first author conducted all the phases in the analysis and the second author took part in phases four, five and six. As shown in Table 1, this process resulted in two sections with five themes in total (a-e).

Results
The findings are presented following the order of the themes in the table above with excerpts from the investigations. In each executive summary, certain factors were presented as particularly problematic, which together composed an argument, or discursive approach, to explain the incident's seriousness.

Framing the incident
The first section presents two separate approaches for explaining why an action or lack of action should be considered serious.

The individual approach: 'staff member's competence and work ethic on trial'
Under the individual approach, a serious incident is considered to be an episode, or sometimes a series of episodes, of unacceptable behaviour by an individual member of staff. Investigators assessed a specific action, or lack of action, and deemed this a violation of the rules (breach of confidentiality), routines (faulty processing, neglect of duties) or ethical grounds of the organisation (sexual relationship with service user, use of force, abusive language). Alleged violations were described as 'dangerous', 'reckless', 'sloppy' and 'misconduct', which at the same time passed judgement on individual staff members as being incompetent, lazy and lacking a work ethic. In the following extract, an act committed on an individual's initiative is evaluated and presented as dangerous, implicating the individual staff member's lack of judgement: Giving a service user his/her own prescribed medicine is not only inappropriate but can also be dangerous. The fact that it happens between an employee and a service user makes it even worse. When asking the employee, he/she cannot see that he/she has made a mistake […]. It is the investigator's assessment that this is an individual employee who has acted on his/her own initiative outside working hours. (Elderly Care,2) When investigators applied the individual approach, it was to put an individual staff member's competence and/or work ethic under scrutiny. According to Potter and Wetherell (1987), the evaluation of an act is linked to how the motive behind it is interpreted. If, for example, a violation is thought to be intentional and done out of laziness, it will be considered more morally reprehensible. It was important to investigators whether an incident appeared to be a conscious act or an unfortunate mistake. The staff member's personal responsibility for an incident was sometimes evidenced with facts showing why an action should be considered a violation without mitigating circumstances. Examples of such evidence include the organisation being 'fully staffed' on the day of the incident or 'well implemented routines' with which 'everyone was familiar' being in place. 'It appears that the responsible planner is aware of established routines, but that he/she has failed to act in accordance with these' (Elderly Care, 186). Interestingly, in some reports, failures to report incidents were treated with the same gravity as violations of the rules. In one report, in which a suspected case of sexual abuse was under investigation, the investigator focused more on the social worker who had failed to report the incident than the incident itself or the possible perpetrator.
The staff member who observed the incident has not fulfilled his/her responsibility to report the incident urgently to the manager, nor reported deviances according to routine, nor reported it according to the SSA, which is judged as a serious misconduct. (Disability,10) To sum up, this framing was formulated as a trial of the staff member's competence and work ethic, and acts were deemed serious through the evaluation of certain conditions: the action breached spoken and/or explicit rules or laws that the staff member in question could reasonably be assumed to have knowledge of (or a lack of knowledge revealed a serious lack of competence), that it was a conscious and intentional act and there were no mitigating circumstances, such as situational pressures (i.e. shortage of staff) pushing staff to break rules. Thus, what seemed to determine the level of seriousness of an incident under the individual approach was whether the act could be understood as reprehensive and whether the staff member involved could be considered responsible. Under this approach, individual staff were seen as free agents, incompetent and/or immoral, and their actions independent of the organisation, which in turn remained free from scrutiny.
The quality development approach: 'the system failure' Under the quality development approach, investigators focused on organisational failures. The adjudged seriousness of an incident appeared to depend on the level of system failure to which it could be connected. The duty to report was framed as a systematic method of ensuring quality rather than a way to identify problematic individual actors, which is characteristic of the individual approach. One investigator explicitly criticised the duty to report as a way to 'cover up' for system failures rather than expose them, indicating that the quality development approach was perceived as the 'right use' of the law, further supporting the analysis of the two approaches as being separate: 'The duty to report is part of the systematic work on quality development in social work. Its purpose is to expose, identify and take measures for system failures, not finding scapegoats to potential faults' (Disability,27).
Under the quality development approach evidence was given that deficits were 'numerous' and 'considerable' and that these occurrences would not end without corrective measures. Incidents were framed as part of 'a pattern' (in some reports an 'escalating' pattern) of repeated failures 'over time'. In the quote below, a number of similar incidents over a given period are provided as evidence for a larger problematic pattern that originates in a systemic failure: Although the degree of seriousness of the consequences for each individual service user could not be determined in the investigation, an extensive number of events in similar areas have occurred and been reported in the care home during a relatively short, coherent period. All events are judged to have a similar causal background in serious deficiencies in, among other things, structure, follow-up and systematics. (Elderly Care,14) Moreover, within this framing the responsibility of involved staff was blurred. The investigators provided explanations for the actions of individual staff and framed their actions as reasonable given the situation. One investigator highlights: 'Strained work situation and insufficient resources, constant work with emergency solutions, 'putting out fires' and few possibilities to plan ahead' (Individual/ family, 6). Responsibility was partly attributed to factors outside the organisation, such as difficulties of recruitment, and partly to management, including references to 'faulty leadership' or more indirectly to 'the responsibility of the organisation'.
Taken together, serious incidents were recurrent; they had happened before, and would most likely happen again, and were therefore a symptom of an underlying system failure. In contrast to the individual approach, responsibility was not attributed to individual staff, who were presented as exposed, strained and struggling and in significant need of managerial support and resources.
Having presented the two different approaches used by investigating social workers to determine the seriousness of incidents, the next section focuses on their interpretations of the consequences of incidents for service users.

Framing consequences for service users
According to government policy (NBHW, 2014), an incident's seriousness should be assessed in relation to its consequences for the service user. To be deemed serious, an incident needs to present an actual and obvious risk of harm to a service user's physical and mental health, life and security. Accordingly, consequences for service users were central in most investigators' explanations of why an incident should be treated as serious. Investigators used three different approaches to explain harm and its importance in their final verdict.
The medical approach: 'harm to life and health' One recurrent approach used by investigators to judge harm to service users was to describe level of injury or 'risk of harm' from a health/medical perspective. This approach was particularly prevalent in care for older people. Investigators assessed the consequences for service users as a seemingly objective variable that could be measured on a scale or continuum from 'life-threatening', 'considerable' or 'profound' ('serious injury', 'malnutrition', 'dehydration') caused by neglect of essential care or medical attendance to discomfort caused by, e.g. omission of hygiene routines. Some investigators used a standardised numeric scale to evaluate the seriousness of an incident, ranging from level one ('discomfort') to level four ('death'). Under this approach, harm to service users was typically described in generic terms, such as: 'the faults [within the organisation] have brought serious consequences to the individual's physical and mental health.' In other words, from a medical perspective harm was treated as being self-evident with no need for further explanatory details.
Moreover, under this approach prevalence was given special importance. Examples included cases in which consequences were deemed more serious if more than one service user had been affected: 'this is a serious incident as approximately 450 individuals who receive this care […] have not been given the help and care they need' (Disability, 24). There were also examples in which an investigator judged 'less serious' examples of poor practice and neglect as being more serious when multiplied. In the excerpt below, an investigator uses the example of not brushing a service user's teeth, which if continued could cause 'serious harm to the individual': The fact that service users stay in bed longer than necessary and that they do not receive food and fluids on time can lead to equally serious consequences. A missed event such as tooth brushing on one occasion may not in itself be a serious shortcoming, but overall, even minor deviations that are described can lead to poorer health for the service users in leading to serious consequences. (Elderly Care,28) In all the examples above, the investigator attempted to set a threshold for harm using measurements of physical and mental injury. Thus, the logic underpinning these descriptions of serious harm is that of an objectively measurable variable linked to medical health. If one accepts this rationale, it is not necessary to acquire the perspective of the person affected. Indeed, service users' own experiences of harm were given little attention in these assessments, and they were presented as a physical body somewhere between the points of being injured and unharmed. However, there was one exception to the practice of measuring harm on a gradient scale, which was in cases involving children. In a number of investigations, the presence of a child or a potential threat to a child's health and development was used as a single signifier for an incident being described as serious. In one report a disabled child was hit by another child in the same care home, leading to distress but no physical injury; in this case the most important factor in the investigator's judgment was the fact that the service user was a child, 'because this is a matter where children are involved and from a child's perspective there are high requirements to secure the child's home environment. What happened is to be seen as a serious mistreatment' (Disability, 6).
The empathic approach: 'service users' experiences and needs' In contrast to the above, the focus of this approach was how service users may have experienced an incident. Investigators explored service users' experiences of an incident by having conversations with them or using observation and interpretation. However, there were a few reports in which a service user gave his/her own view of an incident, and more often than not it was an investigator's interpretation of staff reports of service users' feelings, experiences and needs that determined if an incident was 'serious': feeling 'sad ', 'anxious', 'upset', 'scared' or 'humiliated'. 'Feeling unsafe' was often highlighted and linked to service users' needs. In some cases a service user's emotional response was treated as the most important signifier of a serious incident: The motive for the assessment of seriousness is above all that the service user clearly expresses strong concern about being in the daily activities despite the fact that the service user has previously been positive and enjoyed these activities. The incident has threatened and violated the individual's security and integrity. The staff perceive the service user as affected by anxiety and poor night sleep as a result of the inappropriate and unprofessional treatment. The service user is dependent on support and help in different, daily situations and activities due to functional variation, but the support has not been carried out in a professional manner. (Disability,17) In addition, experiences linked to a lack of social affirmation were highlighted, such as feeling 'excluded', 'abandoned' or 'unwelcome' or 'not feeling seen or heard'. The psychological harm inflicted on a service user was emphasised. In the quote below, a manager's meeting with a potential victim is used as evidence: What is written in the report is confirmed when the manager talks to the affected service user, and the service user bursts into tears when talking to the manager and says that he/she feels sad, offended and insecure. (Disability,14) The different examples given above suggest that the interpretation of an incident's seriousness is closely related to the afflicted person's experiences, and this can be viewed in two different ways. On the one hand it is a sympathetic approach that highlights the user perspective in the construction of harm, on the other, the experiences presented are often the interpretations of investigators who do not necessarily represent the feelings and needs of service users. From either perspective, the picture presented of service users is one of victimhood, accentuating vulnerability, neediness and suffering, and evoking sympathy and pity. Disability care stands out as the single most represented area under this approach. It is also linked to the following and final approach to explaining harm presented in this article.
The legal and human rights approach: 'dignity, legal-social rights and the breach of trust' The last, and least represented, approach used to explain harm to service users relied on a legalhuman rights discourse. Ethics and moral considerations have been an implicit component of all the above but here they represent a separate approach. The reports included under this approach differ depending on the area and type of incident, but they all revolve around a breach of legal rights and ethical values, i.e. 'dignity', 'social justice', 'respect for individuals' integrity' and 'self-determination'. The underlying argument in these cases was that social work is closely related to upholding the human rights and social dignity of service users. Sweden incorporated the Convention on the Rights of the Child in 2020, which may be why cases involving children were considered serious to such a high degree. When care staff and social workers failed to follow protocols, it was associated with a potential serious legal and/or ethical offence. The first recurrent ethical concern was determining acceptable quality of life. In these cases, investigators' judgements relied on a discourse of human dignity. This was especially true in relation to older people's living standards in cases that were no longer a question of mental and physical health: 'Physically, no visible injury has been discovered, but the unworthiness of lying in one's own stool for a long time is not acceptable' (Elderly Care, 20).
Social justice was typically raised in relation to disability services. For people with disabilities, the right to live a life as similar to others as possible was a recurring theme. This is exemplified in the excerpt below relating to a service user's access to a social life through support from social services: People with extensive disabilities should be able to live a life as similar to other people's lives as possible, and participate in society. This means that the measures that people receive must compensate for the difficulties and obstacles that the disability entails. In this case, the organisation has not lived up to the above, which has had serious consequences for the service users' trust in the service. In order to show that the organisation takes this very seriously, the incident is reported to HSCI. (Disability,23) Breaches of service users' legal rights and right to integrity were deemed serious in relation to the exercise of public authority. This could include, for example, 'delayed legal processing' for incarceration and 'delayed child investigation', but most commonly it related to a 'breach of confidentiality' over sensitive information. On the latter, none of the reports elaborated on the consequences for service users, rather harm was assessed in vague terms: 'it [breach of confidentiality] could supposedly have resulted in considerable consequences for the individual if the information was widespread' or 'this concerns sensitive information about personal circumstances' (Individual/family,9). This could be understood with regard to the stigma, shame and damaged social reputation associated with being subjected to the exercising of authority.
In this approach, service users are presented as people deserving of respectful treatment, people with legal rights, integrity and self-determination, but whose self-fulfilment and dignity relies on the reliability of social services. In turn, negative consequences of rights-based violations were presented in terms of 'damaged trust' in an organisation and a reason for individuals to avoid seeking help. People's trust in authority is essential for successful social work. In the reports, incidents that undermined moral rights and human decency were described as not only harmful for the individual but to the reputation and credibility of the organisation. Thus indirectly, the importance of preventing scandals was highlighted with regard to social services' mutual dependency on public trust.

Discussion
The aim of this article was to further current understanding of how Designated Officials (trained social workers) determine whether incidents of mistreatment are serious, drawing on a thematic analysis of investigators' judgements of serious incidents in social services in Sweden.
Similar to the findings of an analysis of serious case reviews in England (Manthorpe & Martineau, 2011) and in child-protection cases (Collins, 2010), there were no clear and consistent thresholds in the reports for why an incident should be deemed serious. Investigators argued for an incident's seriousness on the basis of several recurrent principles related to the construction of the incident itself and its causes and to the framing of the consequences for service users. The latter corresponds with the definition of seriousness given in legislation. The results showed five different approaches to the framing of incidents and harm to service users: the individual, quality development, medical, empathic and legal/human rights approaches.
Investigators used two contrasting approaches to explain what caused an incident. The individual approach aligns with previous criticism of the practice of the duty to report as a form of blaming and shaming (Kjellberg & Höjer, 2020), with little focus on structural shortcomings (Björne et al., 2021). Under this approach, an individual staff member was usually blamed for inappropriate behaviour. In some cases, failure to report was treated as a major offence, confirming this approach as a discourse of obligations for staff rather than a rights-based discourse in which the role of staff is to safeguard the rights of at-risk service users (Kjellberg, 2020). The focus of the quality development approach was to detect system failures and learn from mistakes: a form of trouble-shooting aimed at improving the organisation. This rationale seemed to concur with the emerging systemfailure approach as well as being an expression of the discourse of quality assurance found in other research (Kjellberg, 2020;Manthorpe & Martineau, 2011). More closely linked to the aim of legislation than the individual approach, quality development emphasised the underlying causes of incidents and avoided simplistic explanations that blamed individual staff. The advantage of this is that problems in the organisation can be revised and lessons learnt, ideally before real harm is done to service users (Manthorpe & Martineau, 2011). However, from our analysis, 'the rule of thumb' (Taylor, 2017) for this approach appeared to be that serious incidents should stand out as obvious system failures, characterised by multiple incidents and substantial faults within the organisation. In effect, this means that service users could be subjected to recurrent harmful incidents before there is any real response. Moreover, it shows a tendency to not consider poor practice to be serious (Collins, 2010) unless it has reached a higher level of system collapse. As pointed out by Collins (2010), a 'no blame culture' needs to be balanced with an adult protection perspective with a much greater focus on victims.
In addition, inability to respond to structural harm or frequent abuse without adequate action has been stressed both in international research and within the Swedish context (Björne et al., 2021;Trainor, 2015). The results suggest that the quality development approach does not provide longterm improvements in organisations.
The three approaches presented concerning the protection of service users were shaped by the diversity of discourses around social work but also stereotypes of service users. This shows that the definition of harm and abuse is ambiguous (Dixon et al., 2010) and that investigators have interpreted the same legislation in different ways across the various areas of social services in Sweden.
The most common approach when investigators considered consequences for service users was the medical discourse. Investigators concentrated on how much physical and emotional pain a service user suffered. As noted in child protection cases, the seriousness of an act is usually graded on a continuum, which may result in less harmful incidents of poor practice not being dealt with (Collins, 2010). The medical approach was used extensively in care for older people and was a reactive assessment mostly based on medical opinions. One way of understanding this is as a manifestation of ageismthe absence of a service user perspective and social justice discourse (Jönson, 2016), as well as the tendency to treat older people as a homogenous group with only physical needs (Carey, 2018). In contrast, any harm to children was normally considered serious due to their health and development, but this was not always measured on a scale. However, there are many disadvantages with a medical discourse such as the exclusion of non-physical phenomena, ageing and older people's needs and personal experiences (Carey, 2018). Therefore, measuring harm as physical injury only captures one dimension of harm that service users could be subjected to in social work.
In the empathic approach, harm was acknowledged without the precondition of obvious pain from physical injuries and was based on service users' own perceptions. It was frequently used in disability services, with service users often presented as insecure, helpless and in need of affirmation, although not necessarily corresponding with what they felt or needed. Although attempts were made by investigators to position service users at the centre of incidents, the descriptions given often relied on stereotypes. A common feature of ableism is 'paternalistic prejudice', judging disabled people as warm but incompetent, an attitude that manifests in pity, compassion and sympathy but at the same time results in a lack of respect and infantilisation (Coleman et al., 2015;Nario-Redmond et al., 2019).
The moral base is the core signifier of social work, and emancipatory goals and ideals of social justice distinguish social work from other health-related professions such as medical care (Bisman, 2004). However, the least common discursive approach used to determine harm was the legal and human rights approach, and violations of ethical values were presented differently depending on the incident and area. Generally, investigators positioned service users as having the right to selfdetermination, dignity and social justice and argued that these rights had been breached. Harm was defined as the experience of exclusion and discrimination; these reports mainly concerned the area of disability. As noted by Jönson (2016), researchers and activists within the area of disability have developed a critical perspective on disability, in contrast to care for older people. Disability has been framed as a form of exclusion and discrimination rather than an individual and medical phenomenon (Barnes, 2020). When judging harm to older people using the legal approach, investigators' major ethical concern was a lack of 'dignity' related to the physical environment. This confirms previous research on attitudes about older peoples' needs (Carey, 2018).
To summarise, by using different approaches investigators constructed their arguments to underline an incident's seriousness based on certain dimensions while ignoring others. In this way, investigators acted as gatekeepers for the types of harm and incidents that could be considered serious. It can be questioned whether their processing of reports is sufficient to protect service users and highlight an organisation's shortcomings. While the quality development approach provides the opportunity to learn from mistakes at the organisational level, it could also result in a high threshold for harm to individual service users. Moreover, framing harm through different discourses also places service users in different subject positionings. In our analysis the assessment of harm included harmful but common stereotypes of the disabled (see Coleman et al., 2015) and older people (Jönson, 2016). Most concerning was the latter, which relied heavily on the medical discourse to reduce service users to a physical body. This is not to say that physical health is unimportant, but rather that a medical understanding of harm cannot encompass all core values that are important for social work and risks ignoring older people's perspectives and needs. In addition, assessing harm by leaning on the medical discourse and stereotypes risks perpetuating discriminating and damaging organisational structures rather than questioning them. This said, the issues highlighted here should not be understood as problems that begin with investigators, instead they must be addressed on an organisational level. Factors such as organisational circumstances, legislation and policies (Trainor, 2015) are likely influencers of investigators' judgements of incidents and harm.
In conclusion, there is a need to reconsider the assessment process for reports in social services in Sweden and further explore service users' experiences of harm in cases of mistreatment. The results are important for practitioners and social service organisations in both Sweden and internationally suggesting that mandatory requirements to report mistreatment are not sufficient to protect at-risk service users. It is still necessary to listen and pay close attention to service users' own experiences and to improve safeguarding measures. Different assessment tools may also have to be used within different areas of social services. A critical re-evaluation of the meaning of harm in the context of social work that goes beyond medical issues for all service users seems to be a priority for the future.

Limitations
The reports analysed did not contain complete information, leading to a lack of gender analysis. Depictions of, e.g. people with disabilities are likely to have been shaped by gender, but this was not possible to analyse. Other factors that could have influenced decision making, such as economics and organisational culture, etc., could also not be analysed due to a lack of information.

Declaration of interest
The authors report there are no competing interests to declare.

Funding
This work was supported by FORTE: [Grant Number 2020-00195].