Disruptive Behaviour by Residents in Long Term Care

By Michael Stones, Stacey Stewart, Wendy Kirkpatrick

Introduction

The drunken woman, in pain from a fall, screamed curses at the Emergency room admitting clerk. It took six staff to subdue the former boxer, now a resident in a psychogeriatric unit of a nursing home, when he started swinging his fists. The born-again religious fanatic regularly reduced nursing home staff to tears by calling them whores. Along with socially improper conduct and refusal to comply with care, they are examples of disruptive behaviour that make it harder for staff to perform at their best. The aim of this article is to explore reasons for disruptive behaviour by those under care.

Although a frequent problem in long-term care, issues raised by disruptive residents seem almost taboo given their neglect in the literature. A chapter on abuse in institutions by Beaulieu and Bélanger hinted in only one sentence that actions by residents might affect care: the relationship between an older person and caregiver gives rise to various types of stress, creating an environment in which neither elders nor caregivers are able to flourish. Nelson’s analysis of power imbalance in nursing homes failed to consider the use of aggression by residents as a way to influence staff. The notion that abuse involves a breach of trust defines away the possibility of abuse by residents. However, ignoring a problem does not make it go away. Only by facing the fact that disruptive residents can affect staff behaviours can we begin to seek solutions.

The aims of this study were to develop a measure of disruptive behaviour in long-term care and to examine correlates of such behaviour. Although a few prior studies viewed the topic through a narrow lens (e.g. studies of aggression in residents with dementia), this study used the broader focus provided by the Minimum Data Set 2.0 (MDS 2.0).

Research Methodology

Participants

The MDS 2.0 data were from 399 non-comatose residents of five long-term care units in Thunder Bay in 2000-1. Their mean age was 81.7 years (ranging from 52-105 years), with 31.7% male and 68.3% female.

Measures

The items on disruptive behaviour were E4bA-E4eB. These items measure the frequency and alterability of verbal and physical abuse, socially inappropriate behaviour and resistance to care.

The correlates included sex, age, psychiatric diagnosis (anxiety, dementia, and depression), medication (antianxiety, antidepressive, and antipsychotic), behaviour management (mood therapy, use of restraint), functional status (activities of daily living, cognitive impairment, recent and lasting delirium, and urinary continence), depressed affect (the MDS Depression Rating Scale and withdrawal – E1o-p), and involvement in the life of the facility (F1a-f).

Results

A Measure of Disruptive Behaviour

A statistical procedure called factor analysis showed that all eight disruptive behaviour items belonged to a single factor. In other words, all the items measured the same concept of disruptive behaviour. This finding allowed us to sum the scores on the items to form a Disruptive Behaviour Scale, which had a high reliability of .835.

The scores on the scale can range from 0-16, where 0 indicates no disruptive behaviour. Figure 1 shows that half the residents showed no disruptive behaviour and half showed varying degrees of that behaviour.

Correlates of Disruptive Behaviour

Measures correlated significantly with the Disruptive Behaviour Scale were diagnosis of dementia, use of antipsychotic medication, daily use of restraint, low activities of daily living, low cognitive competence, lasting delirium, bladder incontinence, depressed mood, withdrawal, and low involvement. These correlates probably reflect both risks and responses to disruptive behaviour. Risks are traits that put residents at risk of behaving in disruptive ways whereas responses include treatments meant to control such behaviour (e.g., antipsychotic medication, restraint).

The use of a statistical procedure called linear regression showed that strongest independent predictors of disruptive behaviour (p.01) to be risks rather than responses. These were

Figure 2 illustrates these findings with residents grouped by delirium and depressed affect. [available in the print edition of Stride]

The delirium groups were those with no or any sign of lasting delirium. The affect groups were residents with any daily signs of (1) withdrawal, (2) depressed mood (on the Depression Rating Scale), (3) both the preceding conditions, and (4) neither condition. The graph shows additive effects of delirium, withdrawal, and depressed mood. The group with all three symptoms had disruptive behaviour significantly higher than any other group.

The findings are important because these risks relate to treatable conditions. Although often a lasting condition in residents with dementia, delirium is reversible unlike progressive cognitive impairment. Because delirium was a stronger independent correlate of the Disruptive Behaviour Scale than either cognitive impairment or diagnosed dementia, it is conceivable that the discovery and successful treatment of the sources of delirium might reduce the occurrence of disruptive behaviour. Similarly, withdrawal and depressed mood relate to the anhedonic and dysphoria aspects of treatable depression. ix Major depression usually involves both anhedonia and dysphoria, with anhedonia a frequent but often unrecognized condition in late life.

The data present no obvious way to compare the effects on disruptive behaviour of treating delirium but depression is a different story. A grouping of residents based on diagnosed depression and treatment by antidepressive medication enables comparison of the effects of treated versus untreated depression. Although neither diagnosis nor treatment was a significant correlate of disruptive behaviour, Figure 3 shows that significantly more residents with untreated depression displayed disruptive behaviour compared with the other groups. This finding suggests that a failure to treat depression makes disruptive behaviour more likely or, conversely, that treatment of depression may reduce the occurrence of disruptive behaviour.

Discussion

The aims of this study were to develop a measure of disruptive behaviour based on the MDS 2.0 and to explore its correlates. The main findings were that:

The findings offer hopes that effective diagnosis and treatment of conditions common in long-term care may help to reduce disruptive behaviour. At present, the main responses to disruptive behaviours include antipsychotic medication, restraint and behavioural regimens. If greater success in the diagnosis and treatment of delirium and depression were to result in a reduction of disruptive behaviour throughout an institution, both residents and staff would benefit. (See the case study below.) That is the positive conclusion from this study.

Case Study

Mrs. C. was a tiny, frail, woman of ethnic background. She had been in an abusive relationship all her married life, with her eight adult offspring living elsewhere. When her husband died, it was the first time she had lived alone. Unfortunately, she had a fall that fractured her hip and she was unable live independently. After admission to a nursing home, her behaviour changed almost immediately. This formerly quiet woman became very abusive to family, staff and residents. Treatment by chemical restraint impaired her mobility and distressed her family to the point where a son decided to take her to his home. Her behaviour there remained intolerable, resulting in admission to a different facility. She occupied a shared room in the new facility but the other bed remained vacant for several weeks. Although her behaviour there was less disruptive than in the recent past, she did have a number of violent episodes notably in a communal space. She also began take sheets and towels from the clean linen area at night and spent hours in her room cutting them into strips and sewing them back together. Staff tried many ways to involve her in activity programs but she would not participate. Following the introduction of a roommate, her disruptive behaviour escalated. She became paranoid and physically abusive towards the roommate. After referral for psychogeriatric assessment, the specialist diagnosed a longstanding depression and admitted her to hospital. After about three months of treatment, she returned to the nursing home. She now functioned at a much higher level, resulting in transfer to a residential unit with another roommate. Her disruptive and paranoid behaviours had all but disappeared and for the first time in many years – as noted by her family – she smiled a lot, seemed happy, and took part in social activities.