Introduction

By 2021, an estimated million United Kingdom (UK) people will have dementia, despite a recent fall in incidence [1]. One third of UK people with dementia live in care homes and at least two thirds of care home residents have dementia [2]. A number of inquiries regarding abuse and neglect of residents have influenced public perceptions of care homes [3]. The UK Department of Health defines abuse as “a violation of an individual's human and civil rights by another person(s)" (2). Abuse is defined by the impact, rather than intention of actions or inactions on an individual. A carer may unintentionally be neglectful if they are unaware of a care recipient’s needs.

Conceptual models developed to try to explain elder abuse consider victim vulnerability, abuser stress, psychopathology or impairment, intra-individual dynamics and societal attitudes [4]. Most care home residents have dementia, rely on others for personal care and many exhibit challenging behaviours, factors associated with higher risk of abuse [5, 6]. Carer stress is associated with: low job satisfaction, long hours, low pay, physical demands, staff shortages and minimal education and training [4, 7] and lower empathy which may be linked to lower care quality [8]. In family carers, experiencing violence and aggression from people with dementia was a significant predictor of acting abusively, possibly because staff react defensively or find managing aggression stressful. This may also be true for paid carers and could explain why people with dementia who have more neuropsychiatric symptoms are at increased risk of abuse [9].

In previous surveys, a quarter of relatives of care home residents reported incident(s) of physical abuse [10]; 16% of 114 long term care staff reported committing significant psychological abuse [11]; and over 80% of 577 nursing home staff observed abuse although far fewer admitted to acting abusively [12]. As care workers reporting abuse face potential adverse legal, employment and social consequences, recent surveys have elicited care worker experiences anonymously. In an Israeli study of 510 care staff completing an anonymous questionnaire, just over half admitted abuse and 70% had witnessed maltreatment in the past year; more abuse was reported by staff who experienced more burnout and worked in larger facilities with higher staff turnover [13]. In a small UK survey in 5 nursing homes, most respondents (n = 138, 88.5%) reported witnessing or suspecting abuse in homes where they had previously worked [14]. In both these surveys, staff were asked to identify incidence of “abuse” or “maltreatment”, so behaviours not identified correctly as abusive were undetected. There have been no previous anonymous surveys of elder abuse to elicit behaviours such as ignoring residents’ needs or making them wait for care, without requiring staff to identify behaviours as abusive. As many professionals do not correctly identify abusive behaviours, this is an important omission [15].

The MARQUE (Managing Agitation and Raising Quality of Life) cohort study, primarily devised to consider the relationship between agitation, carer coping and quality of life, enabled us to examine care home abuse on a large scale. We included our recently developed measure for anonymous reporting of abuse and neglect in care homes in baseline data collection and used this to describe the prevalence of abuse and its relationship with care home and staff characteristics. We based our hypothesis on the conceptual model of abuse outlined above, and in particular the likelihood that abuse and neglect occurred most often where there was greater resident vulnerability, abuser stress strained intra-individual dynamics. We hypothesised that staff were more likely to report abusive behaviour in care homes:

1. where staff were more stressed, as indicated by higher mean staff burnout depersonalisation scores;
2. with more residents and
3. fewer permanent staff, because staff-resident dynamics were more challenged;
4. where residents with dementia were potentially more at risk of abuse because they had higher mean neuropsychiatric inventory scores; and
5. in care homes with lower environment quality scores.

Materials and methods

Measures

Trained research assistants conducted assessments at the care homes. Data was collected in the baseline MARQUE study interview [16].

Care home measures.

We recorded characteristics of the included care homes (see Table 1 for details recorded); where care homes comprised units each staffed by discrete staffing groups, we recorded these as separate care home units. The physical environment of care homes was assessed by the TESS (Therapeutic Environment Screening Survey for Nursing Homes and Residential Care), an instrument with good inter-rater reliability and concurrent validity [18]. The TESS score is the sum of 15 items, each scored 0–2 (facility maintenance, cleanliness, handrails, call buttons, light intensity, light glare, light evenness, hallway length (shorter is better), homelikeness, room autonomy, the presence of telephones, tactile stimulation, visual stimulation, privacy, and outdoor areas). Higher scores indicate better environmental quality [19].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of care home studied.

https://doi.org/10.1371/journal.pone.0193399.t001

Carer measures.

Staff completing measures for the MARQUE baseline survey were also asked to complete the revised Modified Conflict Tactics Scale for professional carers (primary outcome), a measure of helpful and potentially abusive behaviour perpetrated or witnessed by staff which we have developed and piloted with care home staff and found to be acceptable and have content validity [20]. It is based on a previous measure used extensively among people with dementia and family carers [6, 21, 22]. It comprises 10 potentially abusive items and six positive care items (Tables 2 and 3. Care staff were asked whether each item had, in the last three months happened “never”, “almost never”, “sometimes” “most of the time” or “all of the time”. The carers self-completed this questionnaire anonymously, in private and sealed it in a blank envelope. We recorded which care home the person worked in. We could not identify individual participants, but notified the care home manager where staff reported that residents had been hit or shaken. If residents or staff told us or we witnessed potentially abusive behaviour this was handled according to our protocol in line with UK safeguarding procedures.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Proportion of care staff reporting abusive behaviour at least sometimes by care home type.

https://doi.org/10.1371/journal.pone.0193399.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Number (%) of care home staff reporting that they had seen or carried out each of the potentially abusive behaviours studied at least sometimes in the last 3 months.

https://doi.org/10.1371/journal.pone.0193399.t003

The following measures were completed as part of the MARQUE baseline assessment; since they could not be directly linked to anonymous abuse reports, they were analysed as mean values at the care home unit level:

The Maslach Burnout Inventory (MBI): a commonly used measure of burnout in care home staff, with adequate psychometric properties [23]. Burnout describes physical, mental and emotional exhaustion that may be accompanied by a change in attitude, from positive and caring to negative and unconcerned. The MBI comprises three subscales—asking how often certain feelings or behaviours associated with emotional exhaustion, depersonalisation and personal accomplishment are currently happening, on a 7 item scale, from never to every day. It does not specify a time period, as burnout is conceptualised as an enduring state [24]. We hypothesised prior to analysis that the depersonalisation (unfeeling and impersonal response towards care recipients) subscale would be associated with abusive behaviour.

The paid carer working most closely with each resident with dementia completed the following measures:

• The Dementia Quality of Life (DEMQOL) proxy, a responsive, valid and reliable measure of quality of life in people with dementia in the last week. It is a 31-item interviewer-administered questionnaire [25].
• Staff gave information so the researcher could rate dementia severity using the Clinical Dementia Rating (CDR), a reliable and valid instrument for rating performance in Memory, Orientation, Judgment and Problem solving, Community Affairs, Home and Hobbies, and Personal Care. This information was used to classify residents’ dementia severity into: very mild, mild, moderate or severe [26].
• The Neuropsychiatric Inventory (NPI), a validated instrument was used to evaluate neuropsychiatric symptoms in the past 4 weeks. 12 domains, each rating a discrete symptom are scored between 0 and 12 (calculated as the product of symptom’s frequency (0–4) and severity (0–3)) with higher scores meaning increasing severity. A summed scored can be calculated for total neuropsychiatric symptoms [27].

Results

Positive behaviours

The majority of care staff reported that most of the time staff spoke nicely to residents during personal care (98%, n = 1490), enjoyed spending time keeping them company (57%, n = 886) and spent time to get to know them (63%, n = 966). By contrast, 520 (34%) of care home staff were never or almost never aware of resident being taken out of the home for their enjoyment; 196 (13%) had never or almost never experienced relatives being involved in care planning, and 234 (15%) had never or almost never been aware of an activity planned around a resident’s interests (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Number (%) of care home staff reporting that they had never or almost never seen or carried out each of the positive behaviours studied in the last 3 months.

https://doi.org/10.1371/journal.pone.0193399.t004

Associations with potentially abusive behaviour

A greater odds of staff reporting that any potentially abusive behaviour happened at least sometimes was associated with working in a home with higher staff burnout depersonalisation scores (1.19, 95% CI 1.05–1.35). Contrary to our hypothesis, numbers or ratios of staff to residents, environmental quality and neuropsychiatric symptom severity in residents with dementia were not associated with abuse. There was a small association with staff reporting higher resident quality of life (1.03, 95% CI 1.01–1.06) (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Adjusted and unadjusted associations of proportion of staff reporting abusive behaviours^* occur at least sometimes and other factors studied (p<0.05 in bold).

https://doi.org/10.1371/journal.pone.0193399.t005

Discussion

This is the first large survey of abuse and neglect in English care homes, and the largest survey to date of abuse and neglect in care homes in any country. Just over half of respondents reported that potentially abusive or neglectful behaviour towards residents occurred at least sometimes. Some abuse or neglect was reported in all but one of the 92 care home units. Neglectful behaviours were most common, and very few care home workers reported actual or threatened physical abuse. While only 5% of carers interviewed reported verbal abuse, at least one carer endorsed this item in over half of the homes. As there may be under-reporting, due to fear of reprisal or lack of awareness of abuse or neglect occurring, these are likely minimum estimates of the prevalence of abuse and neglect. Our worrying findings accord with previous studies [13]. Under section 42 of the Care Act (2014), safeguarding adults is a statutory duty in England. In 2015–16, 39,485 cases of vulnerable adult care home residents (of all ages) experiencing, or being at risk of abuse and neglect were investigated and concluded. Most reported cases related to neglect (34%), with physical abuse next most common (27%) [29]. As nearly 15,000 care homes in England are registered with the Care Quality Commission (CQC) [30], our findings suggest that the neglect investigations currently initiated in care homes may be the tip of the iceberg [31]. Physical abuse was rarely reported in our survey, but constitutes a significant proportion of reported concerns, suggesting cases are more likely to be reported to statutory services.

While the concept of person-centred care [32] is widely accepted by researchers and care home managers, there appears to be a gap in its implementation in care homes. Staff are often caring for residents who resist care, or are verbally or physically aggressive, with little training or support. Where emotional distancing and objectifying behaviours are used as ways to cope with this, there is clear dissonance with the concepts of person-centred care. Attributing aggression to dementia rather than the person may, for example, help staff to be less distressed by it, but perhaps normal inter-personal interactions become more difficult in circumstances where some behaviours are viewed at two different levels; those of the person and those of the dementia [33].

We explored putative associates of abusive behaviour at the care home level, based on our selected conceptual model of abuse occurrence as related to staff stress and distress, resident vulnerability and intra-individual dynamics [4]. Abuse or neglect, as hypothesised, was reported more in homes where staff experienced more burnout and depersonalisation feelings towards residents, mirroring findings with dementia family carers [34]. More abuse was reported in units where staff rated the quality of life of residents with dementia higher. This might be a chance association, or perhaps indicative that staff experiencing more burnout were more likely to act abusively or neglectfully and empathise or notice less when residents were distressed. Staff may conflate life quality of residents with care quality. Perhaps those experiencing more burnout worked to a less person-centred, more task focussed model, rating residents’ quality of life more highly when care tasks were completed to schedule.

Our hypotheses that abuse happened more in care home units where residents with dementia had more neuropsychiatric symptoms that increased their vulnerability, or with fewer permanent staff per resident and more residents because staff had less time to deliver person-centred care, were unsupported. Perhaps this was because most homes operated at minimum legal staffing levels, with additional staff employed only when highly agitated residents required 1:1 care. Staff numbers tend to be higher in care homes where more residents have clinically significant agitation [16]. Having more staff per resident in a unit probably indicates that residents need more care, not that staff work under less pressure, with more time to provide person-centred care. Our hypothesis that environmental quality was associated with less abuse was also unsupported. While important for many residents and for staff, a more pleasant environment was insufficient to reduce likelihood of abuse.

Positive care behaviours greatly outnumbered abusive or neglectful behaviours. Carers reported that most of the time, staff spoke nicely to residents, and spent quality time, trying to know them better. However, activities that took more planning and time—involving families in care planning, and planning activities around a resident’s interests—were less frequent. A third of care workers reported that residents were rarely taken out of the home for their enjoyment. Perhaps staff lacked the time to plan activities, or this was not perceived as within their role or power. Taking residents out may have been avoided due to the possibility of falling, or precluded by staffing levels at minimum levels. Absence of pleasant events and activities may in itself be neglect—albeit unintentional neglect. Reducing abuse and neglect in care homes, requires, in our view, a systemic approach, where they are not perceived as the commission or omission of a direct carer, but as occurring in a system. In its widest view, this involves a society that has delegated care of the most vulnerable to carers undertaking challenging, stressful work for low pay with minimum training and support.

Our study demonstrates that anonymous reporting of abuse by care home workers is acceptable and feasible, and it could be a new, useful indicator of care quality for home managers, regulatory bodies and researchers. The only existing evidence-based interventions to reduce elder abuse target use of physical restraint, and were developed in countries where restraint is acceptable in some circumstances [35]. There are no interventions known to effectively reduce less tangible abusive behaviours, partially due to difficulties measuring outcomes that paid carers are unwilling and residents with dementia unable or unwilling to report.

Our study suggests that that future abuse interventions should focus on reducing staff burnout and depersonalisation; introducing true person-centred care by encouraging staff to explore residents’ personal histories, current and past interests and build pleasant interactions into care, reducing objectification of residents would from our findings, be rational strategies. We are currently evaluating such an approach. We found no association between care home size, staffing numbers and abuse likelihood, perhaps because the care homes included were similarly staffed, or because level of staff training and support, which we did not measure, matters, rather than numbers of staff alone.

Our measure of abuse and neglect was developed with care home workers; it is valid, and the use of anonymous reporting reduced potential for desirability bias. We did not require staff to identify behaviours as maltreatment, so we could capture potentially abusive behaviour that observers or perpetrators may not have identified as abusive. We cannot estimate the proportion of staff acting abusively from our findings, because we did not ask staff to differentiate witnessed from committed abuse [36]. Our work with care staff developing the measure suggested this would have reduced its acceptability. The measure is intended as an indicator of abuse prevalence within a home rather than a way of directly identifying abusing staff. Previous surveys that have distinguished between observed and perpetrated abuse found that observed abuse was reported far more frequently, either because staff were unwilling to admit acting abusively or because abuse was perpetrated by few but seen by many [37]. In a more recent study, many staff were willing to report perpetrating abuse, so we will consider introducing this distinction in future studies [13].

Our questionnaire did not capture behaviour intent or context. We did not recruit a random probability sample of care homes, and those homes participating may have been more highly functioning, with more open cultures and proactive managers and thus with lower abuse rates than those declining participation. This may have biased our findings and potentially underestimated overall levels of abuse. We included regular bank workers but excluded agency staff who did work in the homes on a regular basis. We told staff that we would advise care home managers if concerning abuse were reported, although individual reporters were anonymous. This may have deterred some reporting due to fear of investigations, dismissal or blame. While abuse is always of concern, we told the home when reported abuse crossed a ‘red line’. We informed managers there was “concerning abuse” if any physical abuse was reported at all, or if any of the abusive items were reported as happening “most” or “all of the time”. Our results indicate that the proportion of carers reporting this level of abuse was low. We analysed associations between mean scores and abusive behaviour reports at care home unit level, as abuse was reported anonymously; power for these analyses was therefore limited by the care home unit sample size.

Conclusions

Clinicians treating care home residents should be aware that neglect is common in care homes and that person-centred activities such as trips out or activities tailored to residents’ interests are often happening very infrequently; as these are possible targets for psychological interventions to improve wellbeing. Measuring harm to vulnerable people anonymously brings ethical dilemmas, as abuse detected cannot be directly addressed. Not measuring it, so it remains invisible, is a greater harm.

Acknowledgments

We wish to thank all the care homes, the residents and their families and the staff. We also would like to thank all the other UCL and research network researchers involved in the study, all members of the steering committee (chaired by Prof Sube Banerjee) and all members of the Community of Interest (chaired by Matt Murray from the Alzheimer’s Society).