This study explored tailored interventions guided by an FBA for problematic person transfer situations in two dementia care dyads at an SCU. From a social cognitive theoretical perspective, the reciprocity between the individual, environmental and behavioural factors in the transfer situation was clearly illustrated in care dyad 1 but not as evident in care dyad 2.
In care dyad 1, nonverbal discomfort decreased (DIDTAS item 6) and a positive trend in both verbal and nonverbal discomfort (DIDTAS items 6 and 7) was observed. Caregivers’ transfer-related behaviour mirrored these changes by providing a clear verbal command (DIDTAS item 11) just before beginning the transfer (DIDTAS item 9), waiting for the PwD to respond (DIDTAS item 11) and adapting to the situation, such as keeping eye contact and holding the hands of the PwD during the transfer (DIDTAS item 14).
The high variability in DIDTAS item 7 in terms of addressing verbal discomfort in the baseline phase was stabilized in the intervention phase. It is well known that PwD behaviour can vary greatly (Wångblad et al., 2009), and stabilization, e.g., transfer-related behaviour, could imply an important improvement from a clinical perspective. Lack of time has been reported by caregivers as a barrier when providing palliative care for people with severe dementia (Midtbust, Alnes Einang, Gjengedal, & Lykkeslet, 2018); however, in care dyad 1, the time for completing the person transfer increased by only approximately one minute. Simultaneously, the caregivers perceived less physical strain during the person transfer. The PwD in care dyad 1 suffered from paratonia, which is a motor behaviour problem prevalent in 90–100% of all PwDs with severe dementia and increases the caregiver burden with time (Souren, Franssen, & Reisberg, 1997). Symptoms of paratonia in person transfers are crucial to address, and the intervention in care dyad 1, including caregivers’ verbal and nonverbal communication before initiating the transfer situation and waiting for the PwD to react, shows promising results for PwDs with paratonia.
In care dyad 2, high variability in PwD behaviour was seen in both the intervention and baseline phases. The transfer-related behaviour in the PwD was sustained despite the significant change in the caregivers’ behaviour, as reflected by their adaption of actions (DIDTAS item 14) and the inability of the PwD to remain attentive in the transfer situation (DIDTAS item 1). Possible reasons for the nonsignificant results could be that the tailored intervention did not meet the needs of the PwD. The PwD tended to be less resistive to care in the intervention phase than the baseline phase, although the change was not significant. Reduced resistive behaviour was noted after a behavioural intervention in a previous study (Thunborg et al., 2019), which may indicate that these behaviours can influence behavioural interventions.
In both care dyads 1 and 2, caregiver behaviour improved significantly but still varied in the intervention phase. The variability may be related to the varying level of adherence to the intervention by the individual caregivers, which might reflect the difficulties for caregivers working in SCU to change their accustomed behaviour in their daily work. Caregiver training should focus on more intense and in-depth training to increase the understanding of behavioural change across the development of dementia (Appleton & Pereira, 2019). Genuine professional knowledge concerning the fundamental needs in persons with dementia is recognized as crucial in caregivers for dyadic interactions in dementia care (Jakobsen & Sørlie, 2010; Manthorpe et al., 2010), which means that health care professionals need to recognize the perspectives of both parts since the PwD needs to rely on the other’s contributions in the interaction (Lichtner et al., 2016).
There are both strengths and limitations in this study. One limitation is that an AB design was used. Unfortunately, the health condition of the PwD made it impossible to add additional phases because the internal validity could have been strengthened by an ABAB design (Kazdin, 2017; Petursdottir & Carr, 2018); however, the design was not feasible, mainly for two reasons. First, it would not have been ethical to end a caregiver behaviour after the intervention phase if it was beneficial for the PwD. Second, it would have been difficult for the caregiver to return to their former behaviour, although one possible change would have been to apply a follow-up phase (Kazdin, 2017). In both care dyads, the intervention implemented was a combination of several components, and it might have been better to divide the intervention into several phases, such as an A-B1-B2-B3 sequence. Considering the need for interventions in both care dyads, such a division would have delayed the possibility for an improved transfer situation for the care dyad.
In care dyad 2, the scarce number of observations and high variability make it difficult to draw causal inferences about the impact of the intervention. More observations during a longer period might have better clarified the needs of the PwD and established a stable baseline. The FBA indicated that environmental factors affected the PwD behaviour in care dyad 2, but these factors were also influenced by personal factors, such as motor and social interaction behaviour (Bandura, 1986). The care dyad´s interaction is therefore complex, and the high variability complicated the analyses. It can take time for one factor to influence the others in the triad (Bandura, 1986), and the intervention phase for care dyad 2 may be too short for this to happen. In care dyad 1, the variability in DIDTAS item 7 was high and a stable baseline was not achieved. An extended baseline and additional repeated observations would have been preferable to rule out the possibility that history or maturation could have influenced the change in the dependent variable (Petursdottir & Carr, 2018; Shadish, Cook, & Campbell, 2002). However, the rapid change of the PwD behaviour in care dyad 1 in the intervention phase together with the short latency in caregivers´ behavioural change indicates the impact of the intervention. Thunborg et al. (2019) also demonstrated high variability and a short baseline in a care dyad, thus reflecting the challenges of conducting an intervention in this population. PwDs in the late phase of their dementia disease are frail, and their health condition can rapidly change. Such conditions increase the difficulty of following individuals over time, and in longitudinal dementia studies, dropouts due to death or other reasons commonly reduce the follow-up sample size (Anderson & Blair, 2020). For this reason, we tried to capture as many observations as possible during a short period of time. Unfortunately, only a portion of all the caregivers in the nursing home consented to participate in the study, and due to shift work, it was not possible to conduct and observe the care dyads every day in a row. On the other hand, this study reflects the reality of dementia care and the true conditions for the care dyad.
The results of the visual inspections were strengthened using NAP as a nonparametric statistical complement. Regarding the participants, especially the caregivers, we cannot rule out the possibility that testing could have influenced their actions during the patient transfer situations because they knew they were being filmed (Petursdottir & Carr, 2018; Shadish et al., 2002).
This study is a systematic replication of a former SCS by Thunborg et al. (2019). The present study was conducted in a new setting with regard to the nursing homes, caregivers and municipalities. Additionally, the constellation of the research team for this study was new. Although Thunborg’s participation in both studies could be considered a weakness, it aided in the precise operational definitions of DIDTAS.
Kazdin (2011) emphasized that the selected participants in an SCS should be typical clinical cases. A strength of this study is that the two single cases illustrate two different person transfer situations commonly present in an SCU for PwD. Furthermore, the study was conducted in the participants´ natural environment. Every transfer situation was assessed with DIDTAS, PAINAID and RTC-DAT. The combination of these instruments meant that a behaviour could be evaluated from different aspects. For example, shouting can be interpreted as pain, a way to show resistance to care or an expression of discomfort with words/sounds. By carefully observing each behaviour with all three assessment scales, it was possible to form an idea of what function the behaviour had, which is an essential part in developing an FBA hypothesis (Hanley, 2012).