Of the 44 nursing homes, 27 were eligible for inclusion. To obtain balance across locations and size, 24 nursing homes were randomised into the intervention or control group (Figure 1). All 24 agreed to participate in the study, and all remained involved in the study throughout the project timeline.
Residents with dementia
At baseline, among all residents living in the 24 nursing homes 1304 (59.8%) had a diagnosis of dementia recorded; 608 in the intervention nursing homes and 696 in the control nursing homes. Demographic and clinical data of residents with dementia were similar for those living in the intervention and control homes, except more residents from the intervention group had an advance care directive (ACD) in place (Table 1).
Table 1: Baseline clinical and demographic characteristics of residents with dementia
Characteristic
|
Intervention nursing homes (n=608)
|
Control nursing homes (n=696)
|
Total (N=1,304)
|
Age in years, median (IQR)
|
86.4 (81.0-91.5)
|
87.3 (80.9-91.5)
|
86.9 (80.9-91.5)
|
Female, n (%)
|
409 (67.3)
|
463 (66.5)
|
872 (66.9)
|
Years living in care home, median (IQR)
|
2.2 (1.0-4.6)
|
2.3 (0.9-4.2)
|
2.3 (0.9-4.4)
|
ACFI scores - high, n (%)
Activities of Daily Living
Behaviour
Complex health care needs
|
342 (56.3)
468 (77.0)
416 (68.4)
|
425 (61.1)
556 (79.9)
502 (72.1)
|
767 (58.8)
1024 (78.5)
918 (70.4)
|
ACD in place, n (%)
|
198 (32.6)
|
190 (27.3)
|
388 (29.8)
|
Type of dementia
Alzheimer’s disease
Vascular dementia
Mixed dementia
Lewy body dementia
Fronto-temporal dementia
Other
Not specified
|
284 (46.4)
67 (11.0)
56 (9.2)
14 (2.3)
11 (1.8)
20 (3.3)
158 (26.0)
|
346 (49.7)
87 (12.5)
53 (7.6)
12 (1.7)
10 (1.4)
30 (4.3)
158 (22.7)
|
628 (48.2)
154 (11.8)
109 (8.4)
26 (2.0)
21 (1.6)
50 (3.8)
316 (24.2)
|
Comorbidities
CHF
Chronic lung disease
Chronic kidney failure
Chronic liver disease
Stroke
Cancer
|
68 (11.2)
76 (12.5)
51 (8.4)
6 (1.0)
102 (16.8)
109 (17.9)
|
92 (13.2)
92 (13.2)
69 (9.9)
4 (0.6)
122 (17.5)
115 (16.5)
|
160 (12.3)
168 (12.9)
120 (9.2)
10 (0.8)
224 (17.2)
224 (17.2)
|
Abbreviations: ACD advance care directive; ACFI Aged care funding instrument; CHF congestive heart failure; IQR inter quartile range
Unplanned hospital transfers and deaths in hospital are shown in Table 2. During 6-months follow-up, a total of 260 (19.9%) residents with dementia from all care homes had at least one unplanned hospital transfer; 128 (21.1%) in the intervention and 132 (19.0%) in the control group. Analysis of the primary outcome (proportion of hospital transfers or deaths in hospital among residents with dementia over 6 months) yielded an odds ratio of 1.14 (95% CI, 0.82-1.59), p=0.44. Overall, 154 residents died during 6-months, 75 (12.3%) from the intervention and 79 (11.4%) from the control group.
During 12-months follow-up, 407 (31.2%) residents with dementia had at least one unplanned transfer to hospital, 201 (33.1%) from the intervention and 206 (29.6%) from the control group (odds ratio 1.17; 95% CI, 0.84-1.63); and 310 (23.8%) residents died, 154 (25.3%) from the intervention and 156 (22.4%) from the control group.
Table 2: Number and proportion of residents with dementia with an unplanned hospital transfer and death in hospital at 6- and 12-months follow-up (N=1304)
Outcomes, n (%)
|
Intervention
(n= 608)
|
Control
(n= 696)
|
Odds ratio (95% CI), p value
|
Hospital transfers, 6 months
|
128 (21.1)
|
132 (19.0)
|
1.14 (0.82 – 1.59), p=0.44
|
Hospital transfers, 12 months
|
201 (33.1)
|
206 (29.6)
|
1.17 (0.84 – 1.63), p=0.34
|
Deaths in hospital, 6 months
|
14 (18.7)
|
14 (17.7)
|
1.07 (0.39 – 2.91), p=0.90
|
Deaths in hospital, 12 months
|
22 (3.6)
|
28 (4.0)
|
0.90 (0.44 – 1.83), p=0.76
|
Two of the core nursing modules focused on GOC discussions and documenting GOC medical treatment plans [22]. A new GOC form was introduced in the module and all senior nurses and general managers were sent a copy of the form. However, there was no uptake of the form which relied on a doctor for completion. This was despite general practitioners being sent information about the forms.
Staff knowledge and attitudes
A total of 330 staff completed the follow-up qPAD, 122 from the intervention and 208 from the control nursing homes. Staff from both groups were similar in terms of sex, age, position, years’ of aged care experience, hours worked per week, and highest level of education. (Appendix 1). Among them, 218 staff completed both baseline and follow-up qPAD, 96 from the intervention and 122 from control nursing homes.
There was no significant between-group difference in total qPAD scores; intervention group mean score=60.1 (95% CI 57.4-62.8) compared to the control group mean score=59.5 (95% CI 57.2-61.8), p=0.77. Most of the variability was contributed by individuals (93%) not nursing homes (7%) indicating clustering did not have a big effect on scores. Similarly, when baseline scores were accounted for, there was no significant between-group difference in total qPAD scores; adjusted intervention group mean score=61.1 and control group mean score=60.0 (p=0.53).
Bereaved carer survey
A total of 72 of 258 bereaved carer surveys were received over 12-months. At the request of the provider, surveys were not sent if nursing homes had experienced COVID outbreaks or if coroner’s investigation was underway. There were no significant between group differences in mean SWC-EOLD scores, with mean scores of 26.6 in the intervention and 27.0 in the control group, p=0.80. Most of the variability was contributed by individuals (87%) not nursing homes (13%) demonstrating that clustering did not have had a big effect on SWC-EOLD score.
Process evaluation findings
Due to low uptake of the online modules and feedback from staff, the training period was extended from 2 months as originally planned to 6 months, and face-to-face training sessions were held by members of the research team and the project consultant nurses at each of the intervention sites to boost training participation.
Reach and dose: During the 6-month training period 42 (4.3%) staff completed their core modules, 27 (3.7%) PCWs and 15 (6.0%) nurses. None of the nursing homes achieved the aim of 50% of nurses completing core modules. The three nursing homes with the highest proportion of nurses completing core modules were 24% (care home 9), 21% (care home 4) and 15% (care home 3) (Figure 2).
Overall, 251 (26%) staff completed at least one module during the training period, 79 nurses and 172 PCWs. 160 staff attended the face-to-face sessions, and 102 participated in the training online, with some staff completing modules both online and at face-to-face sessions. The proportion of staff from each care home who participated in the training ranged from 15% to 53%.
Staff satisfaction with the training: Among the 103 follow-up survey respondents who had done the training, 61 (59%) agreed that completing the modules increased their confidence in looking after residents with dementia, 12 (12%) did not agree or were unsure and 30 (29%) did not respond. The reasons staff gave for not participating in the training included: 25% were unable to attend the session, 17% were not aware of the training, 16% had technical issues accessing the training online, and 15% did not have time.
Barriers to participation: The Consolidated Framework for Implementation Research (CFIR) was used to design the process evaluation and analyse the findings from the qualitative interview and survey data [23]. One of the main barriers to staff participation was competing priorities that occurred at the same time as the IMPETUS-D program was implemented, specifically the introduction of the National Aged Care Standards and the Royal Commission into Aged Care Quality and Safety. Staff were expected to attend training in preparation for the new standards, and many had additional administrative work in relation to the new standards and the Royal Commission, and these activities took time and the focus away from IMPETUS-D. These barriers relate to the CFIR constructs of relative priority (standards took priority over IMPETUS-D training) and available resources (staff not having adequate or dedicated time and resources to do the training). In addition, there were organisational and nursing home level barriers, that affected implementation including variable leadership engagement, intra-organisational communication, turnover of care home management and senior nurse roles at 50% of intervention homes (stability of the team), and time constraints and disruption to schedule. Some characteristics of the training intervention itself were also reported as potential barriers to implementation including the online and individual learning approach, the use of a separate LMS rather than the organisation’s LMS that staff were familiar with, and that the training was not mandatory.