Staff-Reported Strategies for Prevention and Management of Resident-to-Resident Elder Mistreatment in Long-Term Care Facilities

Abstract

Background

Resident-to-resident elder mistreatment (R-REM) in nursing homes is frequent and leads to adverse outcomes. Nursing home staff responses may significantly mitigate R-REM's impact, but little is known about current practices.

Objective

To identify common staff responses to R-REM

Methods

We interviewed 282 certified nursing assistants (CNAs) in 5 urban nursing homes on their responses during the previous 2 weeks to R-REM behaviors of residents under their care.

Results

97 CNAs (34.4%) reported actions responding to R-REM incidents involving 182 residents (10.8%), describing 22 different responses. Most common were: physically intervening/separating residents (51), talking calmly to settle residents down (50), no intervention (39), and verbally intervening to defuse the situation (38). Less common were notifying a nurse (13) or documenting in behavior log (4).

Conclusions

Nursing home staff report many varied responses to R-REM, a common and dangerous occurrence. CNAs seldom documented behaviors or reported them to nurses.

The clinical importance of resident-to-resident elder mistreatment (R-REM) in long-term care facilities has just begun to be recognized in the literature (Lachs, Bachman, Williams, & O'Leary, 2007; Pillemer et al., 2012; Ramsey-Klawsnik, 2004; Rosen, et al., 2008; Rosen, Lachs, & Pillemer, 2008, 2010; Rosen, Pillemer, & Lachs, 2008; Shinoda-Tagawa et al., 2004; Sifford-Snellgrove, Beck, Green, & McSweeney, 2012; Sifford, 2010; Sifford & Bharucha, 2010). Previous research has focused on resident abuse by staff (Pillemer & Moore, 1989; Schiamberg et al., 2012), an important but relatively infrequent phenomenon, and resident aggression towards nursing home staff (Gates, Fitzwater, & Meyer, 1999; Zeller et al., 2009).

Despite the relative lack of academic focus, R-REM potentially has a high prevalence and serious consequences. Nearly 50% of nursing home residents have dementing illness (Harris-Kojetin, Sengupta, Park-Lee, & Valverde, 2013), often leading to behavioral disturbances and aggression (Patel, 1993). Residents with behavioral disturbances living in close proximity and interacting frequently commonly leads to negative verbal, physical, and sexual interactions that have high potential to cause physical or psychological distress in the recipient (Rosen, et al., 2008; Rosen, Lachs, & Pillemer, 2008).

In addition to other job responsibilities, nursing home staff must attempt to minimize R-REM and manage its consequences when it occurs (Rosen, Lachs, & Pillemer, 2008). Very few training materials or guidelines exist for staff prevention, intervention, and management of R-REM, so staff have been forced to develop informal strategies to respond to this phenomenon (Rosen, et al., 2008; Rosen, Lachs, & Pillemer, 2008).

Limited literature exists examining staff responses to aggressive resident behavior in long-term care facilities (Enmarker, Olsen, & Hellzen, 2011), but we know of no research evaluating the relative frequency of staff-reported actions and strategies to manage R-REM. The goal of our research was to more completely describe and characterize current nursing home staff experience with and responses to R-REM.

Methods

Overview

This study was performed in 5 large, not-for-profit urban nursing homes in New York. We interviewed a convenience sample of 282 certified nurse's assistants (CNAs) who worked the day shift (28.1% of the total CNA workforce at these nursing homes). We inquired about the R-REM behaviors of 1,688 nursing home residents, representing 96.4% of the resident population. We requested that each CNA report on how they responded to R-REM from residents for whose care they are primarily responsible. Research staff conducted individual private interviews with CNAs regarding the behavior of each resident under study. These interviews lasted an average of 16 minutes. Written consent was obtained from all participants. This study was approved by the Institutional Review Board (IRB) of the Weill Cornell Medical College. When individual facilities had their own governing IRBs, approval was additionally obtained from those IRBs.

Interview Instrument

Interviews were conducted using a structured questionnaire first inquiring whether the CNA had witnessed any form of physical, verbal, or sexual R-REM by any of the residents under their care within the past two weeks. For each event, we asked CNAs to report on any response or intervention they attempted. Staff members were able to report up to three responses for each witnessed behavior. These responses were recorded by the interviewer. This structured questionnaire is part of a larger interview we designed to measure aspects of R-REM. We have described this instrument in detail elsewhere (J. A. Teresi et al., 2013). The questionnaire/data collection instrument for the current study is available from the authors upon request.

Data Analysis

Staff reported twenty two distinct responses to R-REM behaviors during the interviews. Each response was assigned to a category to allow more in-depth analysis of staff behaviors. These categories were generated based on consensus among the investigators after several meetings. Categories created included: Single Incident Management/Reaction; Anticipation of Future Incidents/Long-Term Prevention; Reporting/Documentation/Escalation for Potentially Definitive Management; and No Intervention/Allowing Behavior to Continue. Frequencies of staff responses were analyzed both individually and within categories.

Results

The majority of residents were female (72.3%) with a mean age of 84; 61.9% of residents were white, a lower proportion than found among national nursing home samples (Feng, Fennell, Tyler, Clark, & Mor, 2011), which reflects the ethnic diversity of New York City. The residents had a mean length of nursing home stay of 3.33 years. Additional demographic and clinical characteristics are shown in Table 1.

Table 1. Demographic and Clinical Characteristics of Nursing Home Residents (n=1,688).

Characteristic	N	%/Mean	Standard Deviation
Female	1219	72.3%
Male	469	27.7%
Race
Black	321	19.1%
Hispanic	284	16.9%
White	1043	61.9%
Age (years)		84.3	(9.9)
Education (years)		12.1	(4.1)
Length of stay (years)		3.3	(3.2)

A total of 282 CNAs participated in the study, reporting on responses to R-REM in 1,688 residents under their care. 97 CNAs reported 22 responses to R-REM incidents involving 182 residents during the previous two weeks. These responses are displayed in Table 2.

Table 2. Nursing home staff responses to resident-to-resident elder mistreatment behaviors during previous two weeks (97 staff members reporting responses to the behaviors of 182 residents).

Response Category	Response	Number (%) of Nursing Home Staff Using this Response	Number (%) of Residents to whom Response was Directed
Single incident management/reaction	Talked calmly to aggressive residents--settled them down	37 (38.1%)	50 (27.5%)
	Physically intervened/separated residents	35 (36.1%)	51 (28.0%)
	Verbally intervened to defuse situation	34 (35.1%)	38 (20.9%)
	Asked resident to quiet down	32 (33.0%)	36 (19.8%)
	Redirected or distracted residents	24 (24.7%)	33 (18.1%)
	Moved the resident to another seat	20 (20.6%)	21 (11.5%)
	Removed resident from dining room or public area/changed seating arrangements	17 (17.5%)	19 (10.4%)
	Removed one resident from room to the nursing station	12 (12.4%)	16 (8.8%)
	Offered to find resident an equally good seat	7 (7.2%)	7 (3.8%)
	Explained to residents the nature of communal living	6 (6.2%)	8 (4.4%)
	Explained to resident that the other resident is confused/demented	5 (5.2%)	5 (2.7%)
	Tried to convince residents to compromise	4 (4.1%	4 (2.2%)
	Any single incident management/reaction response	92 (94.8%)	163 (89.6%)
Anticipation of future incidents/long-term prevention	Anticipated residents needs	9 (9.3%)	11 (6.0%)
	Established a routine	8 (8.2%)	9 (4.9%)
	Watched residents vigilantly	5 (5.2%)	5 (2.7%)
	Nylon barricade with alarm to prevent room entry by wanderers, at night	1 (1.0%)	1 (0.5%)
	Any anticipation of future incidents/long-term prevention response	19 (19.6%)	22 (12.1%)
Reporting/documentation/escalation for potentially definitive management	Notified the nurse/CNA	10 (10.3%)	13 (7.1%)
	Notified nursing supervisor	5 (5.2%)	5 (2.7%)
	Made entry in behavior log	4 (4.1%)	4 (2.2%)
	Notified social services/advocated to change resident room	2 (2.1%)	2 (1.1%)
	Requested medication alteration	1 (1.0%)	1 (0.5%)
	Any reporting/documentation/escalation for potentially definitive management	18 (18.6%)	21 (11.5%)
No intervention/allowing behavior to continue	Any report of no intervention/allowing behavior to continue	34 (35.1%)	45 (24.7%)

Staff responses were most commonly Single Incident Management/Reactions, with staff reporting 12 different responses within this category. Staff reported much less frequently employing responses in the categories of Anticipation of Future Incidents/Long-Term Prevention (4 different responses) and Reporting/Documentation/Escalation for Potentially Definitive Management (5 different responses). Many CNAs also reported that they that they did not intervene and allowed the behavior to continue.

The most common individual responses were: physically intervening/separating residents, talking calmly to settle residents down, no intervention, and verbally intervening to defuse the situation. Moving the residents and redirecting or distracting them were also commonly used approaches.

Discussion

CNAs have developed many informal responses to prevent and manage R-REM, a potentially common and dangerous occurrence. The responses described in this study are widely varied, from allowing an argument to continue to notifying social services for a room change to installing nylon barricades with alarms. Clearly, the CNAs who provide around-the-clock bedside resident care and must frequently manage R-REM have developed multiple strategies to protect residents and themselves. In one illuminating example, a CNA attempted 8 different strategies across all 4 categories to control the R-REM behaviors of a single resident. This aptly illustrates how certain residents may exhibit R-REM behaviors frequently and that managing them is complex and challenging. (Table 3)

Table 3. Staff responses to the resident-to-resident elder mistreatment (RREM) behavior of a single resident during the previous two weeks.

Response
Anticipated residents needs
Asked resident to quiet down
Notified nursing supervisor
Notified the nurse/CNA
Offered to find resident an equally good seat
Redirected or distracted residents
Verbally intervened to diffuse situation
No intervention

As no research exists evaluating the efficacy of strategies to manage R-REM behaviors, we have relied for this analysis on existing literature on the management of aggressive or agitated behaviors in cognitively impaired nursing home residents, hypothesizing that R-REM may be a variant or an extreme form of these behaviors. While this approach ignores other potential factors that may contribute to R-REM, including medical illness, mental health issues, personality disorders, predatory behavior, and power and control issues between residents, it may serve as a starting point to understand this behavior. Nearly 50% of nursing home patients have dementing illness (Harris-Kojetin, Sengupta, Park-Lee, & Valverde, 2013). Physically aggressive behaviors typically occur in the later stages of dementia, when a resident is no longer able to communicate verbally (Dewing, 2010). As cognitive abilities and language skills decline, behavior may become the resident's primary source of communication (Dewing, 2010), and disruptive behaviors, such as R-REM may be an expression of unmet needs (Algase et al., 1996; Sifford-Snellgrove et al., 2012; Sifford, 2010). For example, an incident in which a resident becomes violent over food may be related to hunger as an unmet need. The “person-centered” approach to dementia care (Kitwood & Bredin, 1992) stresses identifying these unmet needs when a cognitively impaired nursing home resident behaves aggressively and attempting to address them. As an example, in our study, one resident became involved in R-REM because she was constantly shouting, “ow, ow, ow,” prompting other residents to interact aggressively with her. It was determined later that she had an open wound on her back.

Many of the staff responses reported in this study -- including talking calmly to residents to settle them down, offering to find a resident an equally good seat, and anticipating resident needs -- may be understood in this context. Many other non-pharmacologic approaches not reported by staff in this study, while not directly studied for R-REM, have been suggested in the literature, and may be considered as possible interventions for this behavior. These include interventions such as rubbing in moisturizing hand cream (Deudon et al., 2009), aromatherapy (C. G. Ballard, O'Brien, Reichelt, & Perry, 2002), playing the resident's preferred music (Sung, Chang, & Abbey, 2006), psychomotor/game therapy (Verkaik, van Weert, & Francke, 2005), pet therapy (Elliott & Milne, 1991), and art therapy (Johnson, Puracchio, & Shore, 1992). Notably, however, many of these represent primary prevention of aggressive behavior and may not be effective in the midst of a highly-charged episode of R-REM.

Encouragingly, CNAs only reported requesting medication alteration to manage the behavior of one resident during the previous two weeks. Recent literature has shown that, while anti-psychotic medications (including risperidone, olanzapine, and quetiapine) have modest efficacy in reducing aggressive behavior (Ballard & Waite, 2006), they have significant potential for adverse events, including stroke, orthostatic hypotension, extrapyramidal symptoms, muscle rigidity, and increased risk of fall and hip fracture (Jalbert, Eaton, Miller, & Lapane, 2010). In addition, if aggressive and agitated behavior represents a resident's only remaining method to express discomfort, the use of these psychotropic medications may suppress the behavior while failing to meet the underlying need (Kovach et al., 2006). Therefore, current recommendations suggest that non-pharmacologic methods be used as a first-line treatment for dementia-related behavior problems, with medication reserved as a last resort for residents who pose a risk of severe distress or physical harm to staff or other residents (Ballard & Waite, 2006; American Association for Geriatric Psychiatry, 2003). These recommendations have were strengthened by Centers for Medicare and Medicaid Services (CMS) initiatives to reduce the use of anti-psychotics by 15% by the end of 2012 (Mitka, 2012). Despite this, in a recent survey, 69% of nursing home physicians reported believing that nursing staff request medication too quickly in the care of aggressive residents (Cohen-Mansfield & Jensen, 2008b).

Also, staff in this study did not report using physical restraints to manage R-REM behavior. Mechanical restraints have been associated with increased mortality as well as injury, incontinence, pressure ulcers, and negative psychological effects, including depression (Evans, Wood, & Lambert, 2003). Unfortunately, physical restraints continue to be used in some long-term care settings (Feng et al., 2009; Hamers, Gulpers, & Strik, 2004).

Our research found that staff members took no action in response to the R-REM behavior of nearly one quarter of residents. This finding is consistent with previous research in agitated behaviors among nursing home residents, which found that no nursing assessment or intervention was undertaken 25% of the time (Souder & O'Sullivan, 2003). This suggests that many nursing home staff members may potentially regard R-REM behaviors as a normal, unavoidable part of resident-to-resident interaction and, therefore, do little or nothing when incidents arise.

On the other hand, more egregious forms of abuse were likely to receive some attention. For example, the resident who hit both a resident and his visiting wife was moved. However, in that instance, it was known that the resident did not like anyone to sit next to him, and this invasion of personal space could have been avoided by careful monitoring. Often, however, CNAs may simply not know how to respond. This is potentially dangerous, as staff inaction may allow R-REM to escalate.

Staff only documented the R-REM behaviors of 2.2% of residents in a behavior log. Additionally, few instances of R-REM were documented in the charts or incident/accident reports. This lack is not surprising, as many nursing home staff often have little time to complete basic care tasks. It is also perhaps reflective of the attitude that R-REM is normal resident behavior, making reporting unnecessary. Nevertheless, not documenting these behaviors is perilous, as staff members who care for the resident in the future may be unaware of their potential for R-REM. Also, team members from other disciplines who don't interact directly with CNAs will be unable to participate in development of prevention and management strategies. This pattern is particularly true for nursing home physicians who frequently spend little time at the bedside in the nursing home due to their other professional responsibilities and rely on chart notes to evaluate a resident's longitudinal care and participate in care planning. In addition, if CNAs don't report R-REM when it occurs, the prevalence and importance of the phenomenon will continue to be underappreciated and underestimated. Therefore, universal documentation of these behaviors by care staff should be strongly encouraged.

In this study, CNAs notified a nursing supervisor in response to the behavior of only 2.7% of residents engaging in R-REM and notified social services in response to only 1.1%, suggesting a reluctance to escalate these issues or to engage in a multi-disciplinary approach to R-REM management. As we interviewed only CNAs, our study was limited in its ability to assess whether physicians and mental health professionals ultimately participated in the management of residents exhibiting R-REM behaviors, as typically involvement of these consultants is requested by nursing supervisors and outside the scope of practice of most CNAs. That CNAs very infrequently reported this behavior to their supervisors, however, strongly suggests little involvement.

Despite these findings, geriatric physicians may play an important, multi-faceted role in R-REM control when consulted, including determining whether a treatable medical problem is causing the behavior, giving support to staff, synthesizing information from different multi-disciplinary team members, keeping the family informed, and setting realistic expectations for treatment (Cohen-Mansfield & Jensen, 2008c). In fact, in a recent survey of nursing home physician responses to agitated resident behaviors, physicians reported providing some sort of intervention in 91% of cases (Cohen-Mansfield & Jensen, 2008a). In addition to the prescription of psychotropic drugs, physicians reported using at least one non-pharmacologic treatment in 75% of cases (Cohen-Mansfield & Jensen, 2008a). Physician assessment may also uncover other symptoms that may be contributing to R-REM behaviors, such as inadequately treated pain, delirium, constipation, or infection, all of which are common and difficult to assess among cognitively impaired nursing home residents. Notably, physicians report prescribing pain medication, antibiotics, and/or laxatives in 40% of cases (Cohen-Mansfield & Jensen, 2008a), and referring the resident to at least one other health care professional in 58% of cases, most frequently to a psychiatrist (Cohen-Mansfield & Jensen, 2008a). Physicians believe that, ideally, their role in care for behavioral problems associated with dementia should be increased (Cohen-Mansfield & Jensen, 2008c). Our findings also suggest that, to involve themselves in the prevention and management of R-REM, physicians will need to be proactive in inquiring about challenging resident behaviors.

Despite the breadth of innovative strategies employed by staff in this study, the vast majority were reactive and involved defusing the existing situation rather than preventing future events. Reactive strategies are important to avert escalation and to protect residents, but they are unlikely to reduce future R-REM behaviors. This is particularly concerning because much R-REM is likely unwitnessed by staff (Rosen, Pillemer, et al., 2008), and no reactive intervention is possible. CNAs did, however, report four responses that anticipated future incidents and may be part of a long-term prevention strategy. These included: anticipating resident needs, establishing a routine, nylon barricade with alarm to prevent room entry by wanderers, and watching residents vigilantly. Though staff only employed these strategies in response to 12.1% of residents engaging in R-REM behaviors, using them shows that CNAs acknowledge that R-REM is a chronic problem for which prevention may be superior to only managing issues as they arise. These and related strategies may serve as an initial foundation for development of future evidence-based prevention and management recommendations.

Limitations

This study has several limitations. Results from staff of 5 nursing homes in a single urban area may not be generalizable to all nursing homes. Further, as in any study of this kind, self-reporting of responses to R-REM incidents may be unreliable. We attempted to limit recall issues by limiting our inquiry to the two weeks prior to the interview.

Nursing home care staff members also may be biased respondents, particularly if they are targets of abuse themselves (Rosen, Pillemer & Lachs, 2008). Nursing home staff may underreport due to fear of reprisals from superiors or resident family (Rosen, Pillemer & Lachs, 2008). We have attempted to minimize this potential bias by conducting private anonymous interviews with each CNA. In addition, it is likely that nursing home staff only witness a small sample of resident interactions, and absence of staff may actually precipitate incidents and allow them to escalate (Rosen, Pillemer & Lachs, 2008). Nevertheless, CNAs are the front-line care providers with the greatest potential to provide insights into R-REM and how to prevent and manage it.

Our study only includes staff from the day shift, and responses to R-REM may be different in the evening and overnight. Additionally, previous research has suggested that all staff with resident contact, including employees with nonclinical responsibilities may assist in R-REM prevention and resident protection (Rosen et al., 2008), but our research was not designed to measure whether they are engaging formally or informally in any of these responses.

Our study does not evaluate the severity or type (e.g. verbal, physical, or sexual) of R-REM to which the staff responses were directed. Given that the appropriate responses to an R-REM episode are dependent on the nature of the incident, this limits our ability to contextualize our findings.

Conclusions

Our results, from the first study to quantitatively evaluate staff responses to R-REM, suggest that, despite the lack of any proven strategies to manage R-REM, staff are not relying only on chemical restraints but have also adopted many varied informal strategies to manage it. Most of these strategies, however, are reactive, and care staff seldom document the behavior or report it to supervisors or colleagues in other disciplines. More education is needed to encourage conceptualizing R-REM behavior as preventable. R-REM (as a subset of difficult behaviors in general) may be prevented by anticipating resident needs and heightening staff awareness of environmental influences on behavior. More education is needed to encourage a multi-disciplinary team approach to care planning to address the complex and multi-causal nature of R-REM behavior. In particular, participation of physicians/physician extenders and mental health professionals, while challenging in many nursing homes as they don't participate in regular care meetings but serve as consultants only, is critical to the development of holistic care approaches for these individuals. Ultimately, the effectiveness of existing responses and the potential for increased multi-disciplinary collaboration needs to be examined in greater detail, with a central goal of future research to develop comprehensive evidence-based interventions to assist staff and protect residents.