Respondents were asked to indicate the frequency at which 20 agitated behaviours, as described in the CMAI, had occurred in their last working week. A Likert scale was used to record the frequency (not at all, once/week, several times/week, once/shift, several times/shift). A large proportion of staff (40 per cent) indicated that screaming occurred more than once/day, followed by offensive speech (20 per cent) and negative criticism (twelve per cent). The majority of staff (96 per cent) reported that stabbing (for example, with a pen, knife or fork) had not occurred in their last working week and was thus the least frequently occurring behaviour. These findings are outlined below in Table 7.1.
Table 7.1: Percentage of Staff Reporting Frequency of Each Behaviour (n=25) Behaviour Never (%) Once/week (%) 2-6 times/week (%) Once/day (%) More than once/day (%)
Never Infrequently Frequently
Screaming/yelling 20 20 12 8 40 Offensive speech 28 20 28 0 20 Negative criticism 36 28 16 8 12 Grabbing 56 12 12 12 8 Resistance to personal care 44 12 24 12 8 Threatening language 32 40 12 8 8 Insults/unkind speech 36 16 36 4 8 Verbal accusations 64 12 12 4 8 Scratching 80 8 0 4 8 Combativeness 56 28 4 8 4 Punching 88 0 4 8 0 Shoving 80 12 0 4 4 Slapping 80 4 8 4 4 Kicking 84 8 4 4 0 Pinching 92 0 4 4 0 Biting 92 0 4 4 0 Spitting 88 8 0 4 0 Throwing objects 84 12 0 0 4 Pulling hair 88 12 0 0 0
To facilitate the analysis in view of the small cell sizes, the time categories once/week and two to three times/week were collapsed into ‘infrequently’, and the categories once/shift and several times/shift are collapsed into ‘frequently’. Using this frequency system, screaming was the most frequent behaviour (48 per cent, n=12). Offensive speech, negative criticism, grabbing and resistance to care were the next four most frequently reported behaviour as perceived by 20 per cent (n=5) of staff. Stabbing is perceived by 96 per cent (n=24) of staff as the most infrequent behaviour followed by pinching and biting (92 per cent, n=23).
When the behaviours are grouped according to the categories, aggressive, physically non-aggressive, and verbally agitated (see Section 5.6), staff reported verbally agitated behaviours as occurring more frequently than aggressive behaviours. Of those behaviours considered aggressive, eight per cent of staff reported scratching and grabbing as occurring more than once/day. The majority of staff (80-96 per cent depending on the behaviour identified) considered that aggressive behaviours had not occurred during their last working week.
7.1.3 Incident Reporting
As a part of the survey, respondents were asked to identify what influenced their decision to complete an incident report following an episode of any of the 20 listed behaviours. The majority of staff stated they would complete an incident report if any of the behaviours listed on the survey occurred during their shift (Yes=15, No=10). The reasons given for completing an incident report varied, with many staff citing more than one reason. The majority of respondents indicated that the occurrence of injury to a staff member or resident influenced their decision (n=9). The legal need to complete documentation was noted by five staff as driving their decision to document incidents on an incident report. Other
reasons included: the severity of a BoC (n=3); whether the resident frequently exhibited a BoC/the BoC was known or accepted by staff (n=4); if the BoC is a new behaviour for that resident (n=1); the need to assess, monitor and plan management of BoC (n=4); and to assist in staff education (n=2). One staff member also suggested that there was a perceived pressure from management not to complete incident reports, which had discouraged staff from doing so.
Staff were more likely to verbally report an incident rather than complete a written incident report (verbally report=20, not verbally report=4, no response=1). Verbal reports of incidents were provided to a variety of people within the facility: the Registered Nurse (n=8), during handover (n=5), the NUM (n=5), fellow Personal Care Assistants (n=1), and the DON (n=1).
The final part of the survey asked staff to reflect on a situation involving an aggressive incident, and to consider what might have assisted them to have prevented or managed that situation better. Most staff considered that withdrawing from the resident and reapproaching him or her later would have resulted in a more positive outcome (n=7). Other comments indicated that other useful alternative approaches might be: anticipating the BoC (n=5); using a quiet or low stimulus area (n=4); and having more staff education (n=1). However, the unavoidable nature of some BoC (n=1), the lack of time and staff (n=2), and a lack of access to psychiatric services and diversional therapy (n=2) were also cited as realistic factors impeding the staff’s ability to prevent or manage BoC.
However, this is a general term that is open to many different interpretations by respondents. Thus, in the Facility X survey, respondents were asked about specific behaviours, such as biting and hitting. The findings of both the BoC survey and the Facility X survey indicated that aggressive behaviours only occurred infrequently (i.e. less than once/week) in the residential aged care setting sampled.
Van der Beek and colleagues (1994) argue that objective observational data are considered more valid than data collected through self-reports. Therefore, with consideration of the purpose of the study, the conceptual framework and the results from the BoC survey and Facility X survey, an observational study was conducted in Facility X to observe resident behaviours directly.
7.2 Structured Observation
The BoC survey and the Facility X survey identified that physically aggressive behaviours occurred infrequently in Tasmanian RACFs. More specifically, 80 to 96 per cent of nursing and care staff in Facility X reported that specific aggressive behaviours, such as punching, shoving, kicking, pinching and biting had not occurred in their last working week. However, data generated by both these surveys relied on participant recall, which may be limited by retrospective bias. Therefore, to investigate the incidence of BoC within Facility X in much more detail, a third quantitative method, structured observation, was undertaken.
The purpose of the structured observation was to observe, define and record systematically the behaviour of the eight selected residents living in Facility X for whom consent had been given (see Section 5.8). Resident characteristics are reported in Table 5.3. A modified form of the Agitated
Behaviour Mapping Instrument (ABMI) was used to collect data (see Appendix F).
7.2.1 Time of Observations
Eight resident participants were observed undertaking activities according to their usual routine in Facility X. The observation times were spread to sample across the 24-hour day to sample a range of times and activities (see Table 7.2). The delivery of care to people with dementia is largely set in routine. There are key periods in the day during which particular activities occur. My experience as a Registered Nurse in Facility X ensured I was well aware of the normal daily routines of the resident participants. Most of the observations were conducted between 1500 and 1700hrs (37 per cent), which is referred to as the ‘sundowning period’ (Ryden, Bossenmaier and McLachlan 1991; Dewing 2003; Bachman and Rabins 2006), and between 0800 and 1000hrs (24.1 per cent), which is the hygiene care period (Ryden, Bossenmaier et al. 1991). These two periods cover the times of the day when most of the physically aggressive behaviours are expected to occur. Residents in Facility X were frequently being assisted to bed from 1600hrs onwards, with the majority of residents being in bed by 1800hrs. In addition to time sampling during the key periods (sundowning and hygiene care), sampling also occurred across the night. This was conducted because there was anecdotal evidence to suggest that this was a ‘quiet’ period and so it might be able to provide a contrast to the more active periods.
Table 7.2: Time of Observations
Time period Number of observations Per cent of observations 0001-0600 hrs 46 23.6 Hygiene 0801-0900 hrs 21 10.8 Care 0901-1000 hrs 26 13.3 1001-1500 hrs 19 9.7 Sundowning 1501-1600 hrs 29 14.9 period 1601-1700 hrs 43 22.1 1701-2400 hrs 11 5.6 Total number of observations 195 100.0 7.2.2 Frequency of Behaviours
The ABMI provides a means for observing and recording behaviours of residents in the institutional setting. Descriptions of 30 behaviours are included together with descriptors for location, activity, other persons present, body position, environment (hot, cold, noisy, quiet, light, dark), and the presence/absence of restraints. The data recorded under environment will not be discussed here as it is considered subjective rather than objective data.
Over the entire observation period, seven behaviours were manifest by the eight participants: request for attention; constant talk; babble; screaming; aimless walking; inappropriate disrobing; and verbal aggression, for example cursing. However, over 70 per cent of the observations revealed that the residents were not engaged in any BoC. The most frequently observed behaviours were babbling (8.7 per cent) and aimless walking (8.2 per cent). No physically aggressive behaviours were manifest during the observation periods (see Table 7.3).
Table 7.3: Frequency of Observed Behaviours
Behaviour Frequency of behaviour Per cent of
behaviour No unusual behaviour 138 70.8 Requests attention 5 2.6 Constant talk 10 5.1 Babble 17 8.7 Screaming 6 3.1 Aimless walking 16 8.2 Disrobe inappropriately 1 0.5 Verbal aggression 2 1.0 Total 195 100.0 7.2.3 Time of Behaviours
Interestingly, there were some patterns in the timing of behaviours. For those residents who engaged in aimless walking, 25 per cent of observed behaviours occurred between 1500 and 1600hrs with 50 per cent between 1600 and 1700hrs. These results also found that nineteen per cent of aimless wandering occurred between 0800 and 0900hrs. However, according to a Chi Square analysis of time
versus behaviour these results were not statistically significant (98, n=195) = 85.964, p≤0.05.
Only one of the residents engaged in screaming, 66.6 per cent of which occurred between 1500 and 1700hrs. The correlation of time of day and screaming was not statistically significant.