Psychotropic Drug Administration As a means of managing behaviour among long-term care residents with dementia

[Editor's Note: All tables referenced in this article are available in the print edition of STRIDE.]

Acknowledgement

Funding for this study was received from the National Health Research and Development Program, Health Canada.

Introduction

Psychotropic drugs (anxiolytics, antidepressants, and neuroleptics) are commonly administered to long-term care residents with dementia. Their use, however, is controversial, given that this category of drugs can be used as a form of chemical restraint. The risk that a resident will be subjected to such a restraint depends to some extent on his or her behaviour, as well as on the policy governing the use of such drugs of the facility in which they reside. However, while the vast majority of intermediate care facilities in British Columbia have a “no restraint” or “least restraint” policy in place, almost all report having recently used a number of psychotropic drugs for the specific purpose of management of behavioural difficulties. The line between legitimate medicinal use of a drug and its use as a chemical restraint may be influenced by the availability of adequate numbers of appropriately trained staff. The end result may be a shifting of the line, inadvertent or otherwise, from legitimate medical use to chemical restraint use. In the context of chronic under-funding of the health care system in general, and an increase in frailty and severity of dementia of the long-term institutionalized population, the question is a pressing one. This study examines the relationship between negative behavioural manifestations commonly observed among long-term care residents with advanced dementia and psychotropic drug administration, and seeks to determine whether staffing levels influence this relationship.

Background

Psychotropic drug prescriptions are common among long-term care residents with dementia. For example, Lasser and Sunderland (1998) report that 52% of 92 demented residents in 7 nursing homes in Massachusetts had orders for high potency antipsychotics. Other studies report between 43% and 72% of residents with dementia receive some type of psychotropic medication (Beers et al., 1988; Sloane et al., 1991). Psychotropic drug administration is not an exact science. Medical benefits acknowledged, side effects are always a concern. Existence and seriousness of side effects depend on drug type, frequency of administration, dosage, interactions with other medications, as well as individual predisposing biochemical and personality factors. Nonetheless, some studies have found that neuroleptics, for example, are effective in dealing with symptoms of dementia like agitated behaviours, hallucinations and suspiciousness (e.g. Risse & Barnes, 1986). Likewise, Barnes (1982) compared a group of disturbed dementia patients taking neuroleptic medication with a control group that was not. They found an improvement in level of confusion, impairment of recent memory, and disorientation in the experimental group but not in the control group.

However, neuroleptics have also been implicated as a common cause of extrapyramidal side effects in the elderly (Bassuk & Schoonover, 1977). Ebly (1997) found that falls, fractures, impaired cognition, impaired self-care abilities and general morbidity were associated with psychotropic drug use. A number of additional negative side effects can be expected from certain drugs, the most serious of which is tardive dyskinesia (Peabody, 1987), but also include depression, increased risk of seizures, and eye-lens and corneal deposits (Bassuk & Schoonover, 1977). Since elderly persons in institutions tend to be administered a greater variety of drugs than other segments of the population, the probability of negative drug interactions is also higher. Peabody (1987) also cites the increased likelihood of postural hypotension, anticholinergic effects, and over-sedation as more dangerous for the elderly. In general, adverse effects of psychotropic drugs are more frequent and of greater negative consequence in the elderly (Bassuk & Schoonover, 1977). In this regard, Kane and associates (1993: 568) argue that “The potential for harm from misuse or overzealous use of psychoactive medications is greater than that for physical restraints.”

Predictors of psychotropic drug use include physically abusive behaviour by residents, severe mental status impairment and frequent family visitations (Sloane et al., 1991). Among facility characteristics, a positive relationship between facility size and antipsychotic medication prescriptions has been reported (Ray et al., 1980), although Sloane and associates (1991) found a negative relationship, with residents in larger facilities being less likely to be administered psychotropics. Of the few studies that have linked staff-to-resident ratios specifically with psychotropic drug use, mixed results are reported. For example, Sloane and associates (1991) found that staffpatient ratio was not related to pharmacologic restraint use. On the other hand, Shorr, Fought, and Ray (1994) report an overall decline in antipsychotic drug prescriptions among 9,432 Tennessee nursing home residents over a 30 month period, with greater “third-shift” staffing levels contributing to greater observed decline. Some studies have gone a step further, looking at the effect of staffing quality on appropriateness of psychotropic drug prescription. For example, both Schmidt et al. (1998) and Svarstad & Mount (1991) found that better nurse staffing was related to less frequent occurrences of inappropriate use. Thus, the more salient consideration may be RN-toresident ratios rather than overall ratios. This has been shown in the relationship between RN ratios and physical restraints (e.g. Zinn, 1994).

Data and Methods

Data for this study are drawn from the Intermediate Care Facility (ICF) Project, conducted by the Centre on Aging, University of Victoria between 1995 and 1999. In Phase 1, data were collected on facility characteristics and care routines, including chemical restraint use for all ICFs in British Columbia that care for persons with dementia. Phase 2 consisted of a longitudinal (data collected at admission and 12 months later) study of newly admitted residents with suspected moderate or severe dementia in 77 ICFs. Information was obtained on actual drug administration, frequency and type of agitated behaviour, level of dementia, and a wide range of additional facility and resident characteristics. A special data collection procedure was implemented to gather administrative data, including staffing information, for 50 special care units. Staff ratios were calculated as number of FTE (7.5 hours) staff (RNs, care-aides, and activity aides) for each shift per 100 residents. Thus, information for this study was obtained from administrators, staff, medical charts, independent observations, and residents themselves. Data from Phase 1 were linked to the information gathered during Phase 2, in addition to the information gathered during the special administrative data collection sub-phase, for the 50 SCUs.

This study first establishes the prevalence of facility-level psychotropic drug use as a means of managing resident behaviour among 47 SCUs (information on psychotropic drug use was not available for 3 facilities). Next, the relationship between frequency of negative resident behaviours and psychotropic drug administration at admission and 12 months later is examined for 160 SCU residents. Finally, we control for the possible confounding effect of staff-to-resident ratios, age and sex and cognitive impairment of resident, facility size, and percentage of resident population with dementia, on the relationship between behaviour and drug administration.

Results

SCU administrators clearly indicated their overriding concern for resident behaviour and by implication the need to implement effective management strategies. The most common responses given by the 50 SCU administrators in an open-ended question asking them to specify admission criteria were behaviour-based: disruptive resident behaviour; elopement risk; and wandering behaviour (See Table 1). So great is the concern about resident behaviours in SCUs that a similar pattern is found among reasons given for refusing admission to prospective residents: aggressive resident behaviour; disruptive behaviour; psychotic behaviour. “Effective management of difficult behaviours” also shows up as one of the top three desired resident outcomes. This outcome was mentioned more often than other outcomes such as “maintaining resident’s abilities”, “safety and security of resident” and “resident is happy and content”. Thus, the type and severity of a resident’s difficult behaviours can play an important part in determining not only whether he or she will be admitted, but will also help determine the care routine - which may include psychotropic drug administration - developed for that resident.

It is precisely because resident behaviour plays such a central role for all facets of care provision that psychotropic drug administration requires a closer look. An indication of the extent of chemical restraint use in BC’s SCUs is shown in Table 2. In response to a specific question about the reason for the administration of any of the 17 listed psychotropic drugs, directors of nursing indicated clearly that these drugs had recently and commonly been used for controlling resident behaviour. Anxiolytics appear to be the favoured choice for such purposes, but antidepressants and neuroleptics (especially Loxapine) had also been administered within the previous year. A breakdown of percentage of facilities using each of the three drug types, and the mean number of all drugs administered within each drug type category is presented in Table 3. On average, facilities had used 7.3 of the 17 listed drugs during the previous year specifically for the management of behavioural difficulties among SCU residents with dementia. Only one facility reported not using psychotropic drugs for behaviour management. In the majority of facilities, then, agitated resident behaviour and psychotropic drug administration are intimately linked.

Turning to the resident-level data, resident behaviour tends to be consistent over time, that is, residents displaying agitated behaviours at admission are also often agitated 12 months later. Additionally, residents displaying one type of agitated behaviour (physical aggression; verbal aggression; or physical non-aggression) are often observed displaying one or both of the other types of agitation. Thus, agitated behaviours can be seen as resident characteristics that are enduring and in need of effective long-term supervision and management.

Regular (i.e. not PRN) psychotropic drug prescription is related to agitated behaviours. In particular, increased verbal agitation is associated with regular prescriptions for neuroleptics and (especially) antidepressants. This relationship remains 12 months after admission. On the other hand, PRN psychotropic drug administration at admission is unrelated to agitated behaviours, but is related 12 months later. That is, residents showing verbal agitation were more likely than others to have been administered anxiolytics within the previous 30 days; residents displaying non-aggressive behavioural problems such as wandering, neuroleptics; and residents behaving in a physically aggressive manner, neuroleptics and anxiolytics.

It is plausible that lower staff ratios account for the likelihood that staff will resort to pharmacologic behaviour management, since approximately half of the 50 administrators told us that RN and care aide ratios were either “less than adequate” or “completely inadequate”. A full 70% told us that activity aide ratios were not adequate. To test this possibility, we statistically controlled for staff ratios, for each shift (day, evening, and night). The drug-behaviour relationship remained virtually unchanged. We then controlled for resident age and sex, cognitive impairment, number of facility beds and percent of resident population with severe dementia, in addition to the staff ratio variables. The drug-behaviour relationship proved durable.

Discussion

This study found that many residents with dementia who reside in special care units in British Columbia regularly receive psychotropic medications for the specific purpose of behavioural management. While this is not in itself unexpected, the fact that the relationship appears to be uninfluenced by staff ratios, and by resident and facility characteristics is. Of particular interest, the administration of psychotropic medications increases after admission in response to negative resident behavioural manifestations, but does not depend on staff ratios. Taken at face value, the results tell us that it is not of critical importance how many staff a facility employs, or whether it employs more or fewer RNs, care aides or activity aides per resident: psychotropic medication use will increase when residents display behavioural problems, regardless of the number and type of staff. However, such a conclusion would probably be premature, especially in the case of British Columbia. The variation in staff ratios between facilities is generally not large, and staffing levels are low (especially for RNs and activity aides) facts which limited our ability to detect statistical effects. Similar statistical limitations in long-term care research have been noted elsewhere (e.g. Chappell & Reid, 2000).

Concern for inadequate staff ratios is probably appropriate, however, since many studies show the beneficial effects of enhanced staff-to-resident ratios on other resident outcomes. For example, Reid and Chappell (2000) found that higher activity aide-to-resident ratios in the sample used in the present study resulted in less decline over 12 months in 3 important resident outcomes: expressive language skills, social skills, and cognitive function. They suggest that modest increases in activity aide ratios may be beneficial. In short, one must not throw the baby out with the bath water: staff ratios are important, and will have variable impact depending on the outcome studied. For example, in a study of over 15,000 U.S. nursing homes, Castle and Fogel (1998) linked facility-level probability of being (physical) restraint-free with higher rates of RNs per resident and lower average occupancy, arguing that these conditions in combination allow facilities to explore alternative care options, such as restraint-free care.

Few would dispute the need for effective management of institutionalized persons with dementia whose behaviours may be a danger to themselves and to others (including the nursing staff). However, there is controversy over the most desirable methods of management, and when the option to administer psychotropic drugs is left open, the issue becomes particularly complex. Research has shown that the potential for misuse of psychotropic drugs, whether well-intentioned or otherwise, is considerable (e.g. Sloane et al., 1991). This suggests that concern should be less for ratios - once they reach a particular threshold - than for staff education and training (e.g. Ray et al., 1993). Training that provides knowledge of and confidence to use behavioural management techniques that preclude the use of psychotropic drugs hold some promise.