Nursing Home Violence: Occurrence, Risks, and Interventions

Author Affiliations:

Dr. Ryan Hall is an Affiliate Instructor at the University of South Florida, and is a 2006 Rappaport Fellow; Dr. Richard Hall is Courtesy Clinical Professor of Psychiatry, University of Florida, and Affiliate Professor, Department of Psychiatry and Behavioral Medicine, University of South Florida, Tampa; and Ms. Chapman is Research Assistant to Dr. Richard Hall.

____________________________________________________________________________________________________________________________________ 

Nursing Home Violence Prevalence

Nursing assistants working in long-term care facilities have the highest incidence of workplace violence of any American worker, with 27% of all workplace violence occurring in the nursing home (NH).^1,2 Aggressive and violent behavior, which is often seen in the NH, includes repetitive demands, verbal outbursts, sexual advances, and physically aggressive acts^2,3-6 (Table I). Over time, such behavior creates a stressful environment for other residents and staff. Nursing home studies show that repetitive patterns of aggressive disruptive behavior occur regularly in 43-85% of NHs surveyed.^3,7,8 This prevalence is likely an underestimate due to many episodes of aggression not being reported (ie, an estimated 55-80% of violent episodes).²

Staff surveillance studies show that 70% of NH staff are assaulted at least one time per month.⁵ Certified nursing assistants (CNAs) are physically assaulted on average nine times a month.^5,8,9 Approximately half of all NH staff have been injured by these attacks at least once during their careers, with 38% of those who are injured requiring medical treatment for the injury.^2,5,8

Research has indicated that 75% of assaults against NH staff occur during periods of close staff–resident contact, such as during resident transfers/turning (26-33%), or when assisting with activities of daily living (ADLs), such as dressing changes (43%), toileting (9%), feeding (12%), and bathing (19%).^3,5,10 Assaults reported during these times include grabbing/pinching/hair pulling (38-40%), scratching/biting (4-28%), hitting/punching (12-51%), pushing/shoving (8-8.2%), hitting with object/throwing objects at staff (3-9%), kicking (2-27%), and spitting (1-11%).^2,3,5 Typical verbal aggression includes verbal insults (18.1%), verbal threats (10.7%), and sexual advances (0.7%).³ In a study by Gates et al,10 5% of aggressive behavior (including verbal and physical assaults) resulted in injury to the staff.

Patient Risk Factors for Violence

The residents most likely to assault NH staff are those diagnosed with a dementia or other organic brain syndrome.^5,10 Residents with mild-to-moderate cognitive impairment are 2.59 times more likely to physically assault staff as compared to cognitively-intact residents, while residents with a “severe cognitive impairment” are 8.26 times more likely to become aggressive.³ What is interesting is that the odds ratio for verbal assaults is much lower in individuals with severe dementia, respectively 1.85 for mild-to-moderate impairment and 1.48 for severe cognitive impairment.³ This indicates that residents with dementia may be much more likely to attack without the typical verbal escalation that is seen in cognitively-intact residents, and, therefore, may be more dangerous.

The two most common medical diagnoses found in individuals who assault staff are Alzheimer’s dementia (approximately 90%) and cerebrovascular incidents (approximately 5%).^2,10 This is not surprising since some studies have indicated that 70% of individuals with dementia experience agitation. Reports suggest that as many as 93% of NH residents with dementia experience some form of agitation at least once a week.^10-12

Other patient-driven risk factors in NHs include: male gender; being a younger to moderately-aged individual (age 65 to 84 yr); having a psychiatric disorder such as depression, mania, or schizophrenia; and being socially withdrawn from other residents.³

Once a resident has shown a tendency for violent or aggressive behavior, it is necessary for NH staff to recognize and assess specific predictive factors for that resident (Table II). Identifying specific factors for an individual allows for the development of a multidisciplinary treatment plan (eg, environmental changes, behavioral changes, medication changes) to protect the individual, other residents, and staff. Aggressive, violent residents who have a pattern of physically assaulting other residents or staff should be removed to a more structured, controlled environment, such as a geriatric psychiatric unit, until the behavior is carefully evaluated and treated.

Violence Risk Factors for Staff Who Provide Direct Care

Risk factors for staff who participate in resident care and are assaulted include a high resident-to-staff ratio, increased staff workload (vocational and physical strain), elevated staff anger scores on standardized psychological tests, a perception of being untrained or not competent to deal with residents with dementia and/or aggression, and being a younger-aged staff member (not related to time on the job).^2,5,8-10,13 Many of these factors are thought to relate to a sense of limited time and patience to complete a task. When elderly residents with dementia are hurried, they are more likely to become agitated and potentially violent.¹⁰

Factors that were not significant in predicting attacks included the length of time someone worked in a given NH, working on a specialty dementia unit as compared to a general ward, and having previous training on how to handle residents who have aggression.^8-10 If the training received was perceived as useful by the staff member, he or she was less likely to be attacked. The risk for assault on a specialty unit was comparable to the risk of being assaulted while employed on a general NH unit because specialty units usually have a higher staff-to-resident ratio, hire employees who choose to work on the higher-risk units, and have specifically designed physical spaces that better accommodate residents with dementia. Specialty unit staff also receive additional education, have a higher level of overall satisfaction with their work and colleagues, and are better able to apply their specialty training than staff who work on general units.⁹

Staff Training and Education

A specific staffing group that has been targeted for education in NHs are CNAs. It is estimated that 70-90% of all resident interactions are with CNAs.^2,4,6 Studies show that the patients of CNAs who have undergone behavioral management skills training require less physical and chemical restraint due to agitation than do the patients of CNAs who have not had such training.^4,5 These training programs focus on defining violence and assaults, elucidating the reasons for violent behavior, enhancing communication techniques, and teaching de-escalation skills and methods.^4-6,8 Skills taught in these programs include improving communication, such as appropriate eye contact; providing positive statements; using simple one-step commands; allowing the elderly residents to respond to statements before offering physical assistance; and allowing elderly individuals to independently attempt to engage in tasks where appropriate. Negative communication skills, such as giving multiple commands and announcing multiple activities, can confuse residents with dementia and are discouraged.^4,6 Staff are encouraged to use behavioral modification techniques, including distraction and staff “timeouts” to reduce staff frustration and anger, and to diminish negative staff behavior such as arguing with residents.^2,4,6 General education is also provided about dementia and why it can lead to resident agitation and aggression as a symptom of the disease.^8,10

A study by Burgio et al⁴ found that behavioral management skills training was most effective at decreasing negative behaviors in CNAs working in NHs. A study by Fitzwater and Gates⁵ found that such training of CNAs reduced assaults by approximately 46% in the period immediately following the training.

Nursing homes experience a higher rate of personnel turnover than other healthcare facilities, which requires them to provide constant re-education for their personnel to ensure maintenance of staff competence and ability.^2,3 The annual NH staff turnover rate varies from 25% to 150%.¹⁰ The level of staff training varies from institution to institution, but many studies have found that only 45-65% of CNAs have received training on how to handle residents with agitation.² Such training is essential, since some studies have shown that the receipt of training in how to handle these residents reduces both the staff turnover rate and resident violence, and improves the living situation for all residents.² Longitudinal studies show a reduction of violent incidents, but this effect wanes with time unless training continues.²

Environmental Interventions and Family Interactions

Environmental changes such as re-orientation by staff, frequent visits from family, and reduced levels of environmental stimulation are often helpful in reducing agitation.^1-6,13,14 Family visits are helpful for many reasons, such as helping to calm fears of abandonment and isolation in residents, maintaining communication with NH staff, and helping staff feel more connected to NH residents.¹⁴ Residents who have family visits are less likely to be dehumanized by staff.¹⁴ While family visits, in general, are helpful, there can be cases where family visits can destabilize a resident with dementia. Here again it is important for facilities to try to determine which environmental cues may be causing episodes of agitation and find environmental solutions to address the needs of the resident and family. Anecdotally, having the NH frequently call the family to update them on a resident’s status, having a staff member brief a family about their loved one before a visit, or having a staff member whom the resident is comfortable with accompany the family during all or part of a visit can help reduce the level of apprehension/agitation for both the family and the resident with dementia.

Restraints

Traditionally, NHs have relied on physical restraints (eg, geri-chair, Posey vest) or pharmacologic (eg, neuroleptic) interventions to diminish aggressive behaviors.^3,4,9 The problem with restraints and medications is that they can result in physical injuries and unwanted side effects for the patient.⁴ A cross-sectional study by Voyer et al³ found that the adjusted odds ratio (OR) for physical violence in residents who were treated with neuroleptic medications in NHs was 1.74 (95% confidence interval [CI] of 1.38-2.19). Residents who required restraints were 1.79 (95% CI of 1.37-2.33) times more likely to physically attack staff than those who did not. It is understandable that being restrained may make a resident with delirium or confusion experience more agitation, but residents are often placed in restraints or prescribed neuroleptic medication to help control an already present physical aggressivity.

When it comes to using restraints, the least restrictive level should be used.^14,15 Generally, the hierarchy of restraints goes from a Posey vest or reclining chair with or without a lockable tray (eg, geri-chair) to a two-point restraint (eg, both wrists). Restraining residents in a seated position may be dangerous and can cause syncope or stroke if the resident has low or unstable blood pressure.¹⁴ Although these techniques still have a place in NHs, since the 1980s there has been a great effort both legislatively (eg, Omnibus Budget Reconciliation Act) and by NH organizations to reduce reliance on these techniques as the first line of intervention.^4,16

Pharmacologic Treatments for Agitation, Aggression, and Violence

Traditional Pharmacologic Treatments
The two classic medications that have been used to treat aggression in the elderly are benzodiazepines for acute aggression and neuroleptics for long-term agitation and aggression.^11,15,17 Various studies of NHs have found that approximately 50-60% of residents were taking some form of psychotropic medication (studies using data from before 2005 black box warning for neuroleptics was instituted), with approximately 15-22% of residents taking anxiolytics/hypnotics and 23-25% taking neuroleptics.^18,19 The percentages of residents taking neuroleptics had remained constant over the last 10-15 years, but the frequency with which atypical neuroleptics (eg, risperidone, olanzapine) have been prescribed has been increasing as compared to typical neuroleptics (eg, haloperidol, fluphenazine).^18,19 Benzodiazepines and neuroleptics have side-effect profiles that can make their use in elderly individuals difficult or dangerous. It is often best to try environmental and behavioral approaches first to control aggression and agitation, and to rule out medical causes of delirium, such as a urinary tract infection in acute cases, before resorting to the use of new psychotropic medications.¹⁴

Benzodiazepines are generally metabolized through the liver, with the notable exceptions of oxazepam and lorazepam (renal metabolism with minimal liver metabolism). Benzodiazepines are metabolized more slowly by elderly persons than by younger individuals, resulting in higher blood levels for a given dose/body weight.^20-22 The half-life and duration of effect are enhanced, and these drugs may thus have a greater impact on residents’ cognitive abilities.

Benzodiazepines can also lead to worsening of memory problems, produce hypotension, increase the risk of falls, and produce a state of dependence, which could result in seizures if they are suddenly stopped or withdrawn too quickly.¹⁵ In addition, they may disinhibit a well-functioning patient and produce a paradoxical worsening of aggression or precipitate delirium.^15,23 In certain situations (eg, alcohol withdrawal, hysterical agitation), benzodiazepines can reduce agitation. They are less effective when used for the long-term reduction of aggression, such as is needed by many NH residents.^17,24

Neuroleptics, both typical and atypical, can result in the side effects of: dystonia; postural hypotension; exacerbation of closed-angle glaucoma; akathisia; neuroleptic malignant syndrome; sedation; cardiac conduction delays (prolonged QTc); increased risk of cerebrovascular events (CVEs), particularly during the first 3 months of administration; and paradoxical worsening of cognitive function.^{15,20-22,24,25}

The decision to start pharmacologic treatment for agitation related to dementia has become more complicated with the addition of a black box warning for risperidone and other neuroleptic agents used for persons with dementia.²⁶ A meta-analysis by Schneider et al,²⁴  published in the Journal of the American Medical Association (JAMA),found an increased OR for death of 1.54 CI, 1.06-2.23; P = 0.02) for individuals treated with atypical neuroleptics versus controls. This finding may not apply only to atypical neuroleptics. Although the 2002 Cochrane review on haloperidol found that haloperidol effectively treated aggression with no findings of increased mortality in individuals with dementia, the JAMA meta-analysis reported a mortality OR of 1.68 (CI, 0.72-3.92; P = 0.23) for haloperidol based on two studies.^11,24 A retrospective cohort study (N = 22,890) by Wang et al,²⁷ published in the New England Journal of Medicine (NEJM), also found the typical neuroleptics to have the same, if not a higher, mortality risk than the atypical neuroleptics. Since the original JAMA and NEJM studies, additional meta-analysis and retrospective studies have continued to show as great a risk, if not a greater risk, for typical neuroleptics.^28,29

A 2006 Cochrane review on the use of atypical neuroleptics for agitation and psychotic features in individuals with Alzheimer’s disease found that risperidone significantly improved both conditions, while olanzapine improved only agitation.²⁵ The Cochrane review also found a significant incidence of CVE and extrapyramidal symptoms (EPS) for both medications.²⁵ In 2006, Schneider et al²⁶ reviewed the efficacy and side effects of the atypical neuroleptics for all types of dementias. They found that risperidone and aripiprazole showed improvement on the efficacy rating scales but olanzapine did not. This meta-analysis also found that the greatest benefit from neuroleptic medications, which they described as “modestly effective,” occurred when they were used in persons suffering from severe dementia as compared to those with milder forms of dementia.²⁶ There was no indication of an increased risk of falls, self-injury, or syncope in the neuroleptic-treated group versus the control population.

The recent phase 1 analysis of the Clinical Antipsychotic Trials of Intervention Effectiveness-Alzheimer's Disease (CATIE-AD) showed improvement for olanzapine and risperidone on the Brief Psychiatric Rating Scale (BPRS) hostile suspiciousness factor and the Neuropsychiatric Inventory (NPI) total score.³⁰ There was no significant difference between neuroleptics and placebo on cognition, functioning, care needs, or quality of life, except for worsened functioning with olanzapine as compared to placebo. The overall conclusion of the study was that atypical neuroleptics were potentially effective in treating targeted symptoms such as anger, aggression, and paranoid ideation, even though there was no significant improvement in quality of life as compared to placebo. It needs to be remembered that the study population were community-dwelling patients, and that there may be a more substantial improvement for NH residents, who typically have more severe dementia.²⁹

In pooled study data of NH residents (N > 1100), risperidone was more effective in reducing violent behavior and psychosis in geriatric residents with dementia (mixed population group of Alzheimer’s disease, vascular dementia, and other dementias) than placebo while maintaining a similar safety profile (total adverse events reported for risperidone 84.3% vs placebo 83.4%), with exceptions for somnolence (21.6% vs 13%), peripheral edema (11.6% vs 3.4%), fever (7.9% vs 3.6%), urinary problems (14.5% vs 11.5%), EPS symptoms (16.3% vs 11.6%), and CVE (3.9% vs 1.6%)].7 In the safety profile study, 4.4% of the residents taking risperidone died during the 12-week study period as compared to 3.4% of those taking placebo.¹⁷ This study also points out another potential complication in determining the effectiveness of these medications in the studies conducted to date: There appears to be a very large placebo response/effect in NH residents studied, which is approximately 2.5 times higher than would be expected, assuming the classic 30% placebo response rate.

A consensus conference of geriatric mental health experts reviewed the evidence regarding safety and efficacy of neuroleptic medication for treating dementia-related symptoms of agitation and aggression.³¹ They found that there was enough evidence to suggest increased concern for safety, but that the lack of a stronger evidence for drug efficacy was in part caused by problems in the clinical trial design.

A question that often arises is what to do with residents who enter a NH already taking neuroleptics or who have recently been started on one. Should the medication be left in place or stopped as soon as possible? In a recent population study by Kleijer et al³² in The Netherlands, it was again found that current and recent exposure to neuroleptic medications was associated with an increased risk of the individual experiencing a cerebrovascular adverse event as compared to individuals not taking a neuroleptic (OR 1.7; 95% CI, 1.4-2.2). What was surprising with this study was that there was an inverse temporal connection between the length of time individuals were taking the neuroleptic medication and their risk for experiencing a CVE. For individuals who took neuroleptics for less than 1 week, their OR was 9.9 (95% CI, 5.7-17.2). By the third month of neuroleptic treatment, the risk was comparable between the two groups (OR 1.0; 95% CI, 0.7-1.3). This finding suggests that individuals who are already stable on a neuroleptic medication should not have the medication discontinued simply due to concerns about a possible cerebrovascular event. However, the literature is still confused on this point, as is evident from a study by Gill et al,²⁸ who reported an elevated risk for a CVE 180 days after being started on a neuroleptic medication. However, in their findings they cautioned that there were “unequal rates of censoring” (sample bias), which may have affected the results.

Neuroleptic Risks and Benefits: What Should be Communicated to Patients and Families?
A Food and Drug Administration black box warning (a warning indicating a significant risk of serious or life-threatening adverse effects) revealed a risk of death in drug-treated residents of between 1.6 to 1.7 times that seen in placebo-treated residents, based on pooled data from 17 studies that averaged 10 weeks in duration and used varying atypical neuroleptics. Although there does appear to be an increased mortality rate, caretakers need to remember that there are risks for not providing pharmacologic treatments to residents who are still experiencing aggression after environmental and behavioral attempts to calm and control them have failed.^25-27,33,34 Violent residents can hurt themselves and others, as the NH data demonstrate. Violent NH residents with agitation are at increased risk for falling, becoming malnourished, not having their ADLs attended to as frequently as needed, becoming isolated, and suffering health complications from the refusal of medication.

Before starting neuroleptic medication, physicians need to have a risk/benefit discussion with their patient and, where indicated, the patient’s family or guardian. Current data suggest that persons taking atypical neuroleptics sustain a three-to-fourfold increase in their risk for cerebrovascular incidents.^7,25 Another study suggested that for every 9-25 individuals with dementia who are helped by these medications, one may die sooner due to the medication.²⁴

Other Pharmacologic Medications
Other medications, such as the selective serotonin reuptake inhibitors (SSRIs), trazodone, anticonvulsant medications (eg, valproic acid, carbamazepine), antagonists to the glutamate NMDA receptor (eg, memantine), and the acetylcholinesterase inhibitors, have been tried as treatments for aggression in the elderly.^17,35,36 Although there have been some promising case reports, open-label trials, and even a few placebo-controlled trials, none of these medications consistently have been found to be as effective (“modestly”) as neuroleptics for the treatment of acute or prolonged agitation and aggression in persons with severe dementia.¹⁷

SSRIs, particularly sertraline and citalopram, have been helpful in the subpopulation of persons with dementia who are also suffering from depression (approximately 25-30% of all Alzheimer’s dementias), but have not been shown to be particularly beneficial for persons with dementia only.^12,17,37-39 The particular improvements noted in persons with dementia and depression include improvement in patient mood, decreased behavioral disturbances, and decreased caregiver distress.^37-39 Improvement in cognitive function has been reported in some studies and seems more likely to occur in women treated with SSRIs; however, additional research is needed to better identify the characteristics of those persons who demonstrate significant improvements in cognitive function following SSRI treatment.^37-39

Valproic acid appeared effective in 66% of case reports and open-label trials, but the placebo-controlled trials reported mixed results.^17,35,40-42 The Cochrane review⁴⁰ and the review by Sink et al¹⁷ of valproic acid in persons with dementia for control of agitation did not find enough evidence to recommend its use.

Memantine and acetylcholinesterase inhibitors have been shown to have beneficial effects on cognitive abilities, ADL functioning, and maintenance of behavioral functioning in 6-month studies.^43,44 Neither class of medication has demonstrated efficacy for decreasing aggression or violence once agitation is present in individuals with severe dementia.^43,44 However, there are beginning to be some small open-label studies and one double-blinded study that have reported improvement of behavioral symptoms in individuals with Lewy body dementia.²⁹

Trazodone has been anecdotally used by many physicians for the treatment of aggression and agitation in the elderly. Kitamura et al45 recently published a small (N = 13) retrospective study in which they found that 6 out of the 13 patients had an improvement with “aggression and negativism” in a caregiving situation. However, the Cochrane review³⁶ and the review by Herrmann and Lanctot⁴⁶ did not find enough data currently available to recommend trazodone for the treatment of agitation and aggression.

Summary

Other residents and staff members in NHs are frequently the victims of geriatric violence usually perpetrated by residents with delirium, dementia, or other medical illness. Environmental behavior management techniques are often effective and should be tried first to reduce aggressive behavior. Techniques such as decreasing stimulation, reorienting the resident to time and place, use of simple one-step directions, staff vigilance, and allowing extra time for tasks to be completed are helpful. If aggressive and violent behaviors persist after environmental techniques have been tried, neuroleptic medication and restraints may be needed to protect the resident, staff, visitors, and other residents. Restraint and neuroleptic use should not be undertaken lightly.

Physicians need to consider the risk/benefit ratio and communicate it clearly to the resident and his/her family/guardian where appropriate. Starting a neuroleptic medication can be a difficult decision since data show a higher mortality rate associated with its use. Conversely, neuroleptic medications such as risperidone have been shown to reduce aggressive and violent behavior in individuals with severe dementia. At this time, other classes of medication have not convincingly and consistently shown the same level of improvement as the neuroleptics have.

The authors report no relevant financial relationships.