Gentle Persuasive Approaches Dementia Education Improves Staff Self‐Efficacy and Knowledge in a Post‐Acute Care Hospital: A Quality Improvement Project

ABSTRACT

Introduction

The prevalence of dementia cases within post‐acute rehabilitation hospitals in Canada poses a unique challenge due to the distinctive vulnerabilities of the patient demographic in these settings. Healthcare staff in rehabilitation hospitals receive limited formal education about responsive behaviours.

Aim

To address a practice gap, a program evaluation of the Gentle Persuasive Approaches (GPA) in Dementia Care education program was implemented at an inpatient rehabilitation hospital in a large metropolitan centre.

Design

A multiple method, pre‐post intervention design was used to assess the impact of GPA classroom sessions on staff capacity for providing person‐centred dementia care.

Method

Eighty‐eight staff members attended GPA sessions and completed quantitative and qualitative evaluation measures of dementia care self‐efficacy and knowledge both pre‐ and post‐intervention. Paired t‐tests and reflexive thematic analysis were used to compare pre‐ and post‐intervention scores and responses.

Results

Quantitative findings revealed that after attending GPA sessions, participants showed significant improvements in both self‐efficacy and knowledge scores relative to baseline. At 6–8 weeks post‐intervention, significant improvements in self‐efficacy were sustained. Qualitative responses before the intervention showed that during situations involving escalating responsive behaviours, participants relied upon basic strategies to regulate their own emotions and provide person‐centred approaches with variable success. In contrast, 6–8 weeks after GPA education sessions, participants identified and described responsive behaviours by relating them to the unmet need that needed to be addressed. Participants then reported successful application of tailored non‐pharmacological approaches with confidence and skill at the point of care. GPA dementia education improved staff capacity to provide compassionate and effective care to individuals with dementia.

Patient or Public Contribution

No patient or public contribution.

1. Background

1.1. Dementia in Post‐Acute Settings

The growing number of dementia cases in Canada presents a pressing challenge for post‐acute care facilities (Prince et al. 2016). Patients within these facilities have an increased risk of dementia, attributable to factors such as old age, immobility, polypharmacy and surgery or intensive care (Oh‐Park et al. 2018). Over 500,000 Canadians currently live with dementia, with predictions rising to approximately 912,000 by 2030 (Alzheimer Society of Canada 2016), leading to an increased demand for healthcare services capable of effectively addressing the unique needs of this population. It is imperative to provide specialised care that is tailored to the needs of people living with dementia and address the existing barriers that hinder the provision of effective rehabilitation (Cumal et al. 2022).

1.2. Defining Responsive Behaviour

Dementia causes behavioural changes that can be distressing for both people living with dementia (PLwD) and their caregivers. Responsive and reactive behaviours are terms used to describe actions, words or gestures exhibited by PLwD in response to upsetting or confusing stimuli (Nucera et al. 2021). These behaviours can include agitation, wandering, restlessness, hallucinations, paranoia, repetitive vocalisations and withdrawal. Traditionally labelled diagnostically as behavioural and psychological symptoms of dementia (BPSD), the term responsive behaviour has been identified as a preferred term as these behaviours often occur in response to a person's unmet needs (Bourgeois and Hickey 2009; Cohen‐Mansfield et al. 2015; Dupuis and Luh 2005; Dupuis et al. 2012).

Guidelines for the assessment and management of BPSD released by the Canadian Coalition for Seniors' Mental Health (2024) recommend as good practice the provision of education and organisational systems of support for healthcare providers, allowing them to implement structured approaches to BPSD. Recommendations also include implementation of psychosocial interventions guided by an assessment of the PLwD and identification of contributing factors to the behaviours. These non‐pharmacological approaches as first‐line interventions can mitigate responsive behaviours and facilitate deprescribing of antipsychotic medications, which have harmful side effects in PLwD (Watt et al. 2020). Responsive behaviours qualifying as BPSD of risk may require a combination of non‐pharmacological and carefully monitored pharmacological approaches. However, staff must know respectful and suitable techniques for self‐protection in the workplace, given that if these combined approaches are not completely or immediately effective, both staff and PLwD remain vulnerable to injury.

1.3. Clinical Context

Runnymede Healthcare Centre is a 206‐bed post‐acute rehabilitation hospital in which approximately 40% of patients are older than 85 years, and 18% are between 65 and 74. The prevalence of dementia in this age group necessitates a specialised approach. Facilitation of the Gentle Persuasive Approaches (GPA) Basics education (Advanced Gerontological Education 2014) program at our organisation was, therefore, a strategic response to the specific needs of the patient population.

1.3.1. Problem Statement

Post‐acute care facilities play a pivotal role in the continuum of care for individuals recovering from various health conditions, including those living with dementia. However, staff within these specialised settings often do not receive specific training in dementia care and therefore lack knowledge, skills and confidence when it comes to the care requirements of PLwD (Bamford et al. 2019; Goodwin and Allan 2019). Post‐acute care rehabilitation for PLwD often results in suboptimal outcomes, underestimated prognoses, premature discharges and overall ineffective care (Flanagan et al. 2020; Loh et al. 2023; Semelka et al. 2024). Compounding these challenges are prevalent doubts and pessimism about the benefits of therapy for PLwD, stemming from biases or historical difficulties in providing this type of rehabilitation (Kitwood 1997; Laver et al. 2020; Sedney et al. 2019). In this program evaluation, our aim was to implement dementia‐specific education that would empower interdisciplinary point‐of‐care staff with the knowledge and skills necessary to enhance rehabilitation efforts for PLwD.

1.4. Intervention Selection

Gentle Persuasive Approaches (GPA) is an established dementia‐care curriculum first developed in 2004 and now implemented across Canada. It was initially designed to educate long‐term care staff on how to use a person‐centred approach to respond respectfully, with confidence and skill, to responsive behaviours. GPA encourages staff to reframe behaviour traditionally viewed as ‘aggressive’ as self‐protective behaviour reflecting unmet physical, psychological, spiritual or cultural needs. Staff learn to assess the meaning behind behaviour and work with the person and family to address unmet needs. The GPA program also includes gentle physical body containment techniques that are safe and respectful to use with older adults. These techniques are invaluable in some instances of escalating physical behaviours for momentarily redirecting the patient with dementia away from an altercation or situation involving risk of injury to the patient or staff.

GPA has been associated with a variety of positive outcomes in acute care settings. For example, findings from an inpatient geriatric psychiatry program showed that incidents of patients' aggressive behaviour declined by 50% after staff completed GPA, and this decline was sustained 3 months later (Speziale et al. 2009). Other studies have demonstrated significant improvements in acute care staff self‐efficacy for supporting PLwD and responsive behaviours after attending GPA sessions (Pizzacalla et al. 2015; Schindel Martin et al. 2016). A qualitative study (Hung et al. 2018) showed that after GPA, hospital staff learned to adopt a positive attitude towards dementia care and look for the meaning behind patients' behaviours they had previously perceived as challenging or aggressive. Finally, the online format of GPA significantly enhanced self‐efficacy, knowledge and competence in dementia care for staff on five in‐patient medicine units at a large Ontario hospital (Crandall et al. 2022). Qualitative findings showed GPA‐educated staff made efforts to get to know each PLwD, investigated modifiable reasons for responsive behaviours and began to adapt hospital routines to address unmet needs.

In this program evaluation, we aimed to determine the effect of GPA dementia education on staff confidence and knowledge in skills necessary to support PLwD and responsive behaviours. We anticipated that staff who attended GPA sessions would immediately show statistically significant increases, relative to baseline pre‐GPA scores, in self‐efficacy and knowledge in person‐centred dementia care. Moreover, we expected that these increases would be sustained 6–8 weeks later. Additionally, we expected that participants' post‐intervention qualitative responses would reflect their acquisition and application of new systematic, evidence‐informed and tailored non‐pharmacological approaches to responsive behaviours during everyday practice.

2. Methods

2.1. Evaluation Setting and Sample

Eighty‐eight interdisciplinary staff members attended GPA Basics education sessions and volunteered to complete evaluation measures. Participants had a range of professional roles, including Registered Practical Nurses, Registered Nurses, Advanced Practice Nurses (APN), Allied Health (Physiotherapist/Occupational Therapist/Registered Dietician/Speech‐Language Pathologists/Rehab Assistants), Patient Care Administrators/Clerical staff and staff from the Patient Flow and Discharge team.

2.2. Ethical Considerations

This project was designated as a program evaluation by the institutional review board and therefore ethical approval for research was not required. Nevertheless, standards as explicated by the Tri‐Council Policy Statement: Ethical Conduct for Research Involving Humans (2018) were followed. Participants provided informed consent and, although completion of GPA was required as part of their employment, participation in the evaluation was voluntary and anonymous.

2.3. GPA Intervention

GPA includes four interactive modules with content on (1) person‐centred care principles, (2) brain changes common in dementia and their manifestations in responsive behaviours, (3) communication and interpersonal strategies that can either escalate or defuse behavioural symptoms and (4) staff‐specific physical self‐protective skills and debriefing and reassurance techniques that are effective, safe and respectful to use with PLwD. The modules build upon each other in complexity, such that the physical techniques are scaffolded upon a solid understanding of the principles of person‐centred care.

Over the course of 7 months (April–October, 2021), pairs of in‐house certified GPA coaches facilitated 13 separate 7.5‐h day‐long education sessions, each with a group of 10–15 staff, and led the evaluation data collection. GPA is a scripted program including PowerPoint slides, videotaped vignettes, hands‐on practice, opportunities for reflection and dialogue, and a resource booklet and certificate of completion for each participant. Due to the COVID‐19 pandemic, some interactive components were modified for social distancing according to standardised guidelines.

2.4. Evaluation Design and Measurement

Evaluation measures were completed at three time points: immediately before, immediately after and 6–8 weeks after attending a GPA session:

1. The pre‐GPA socio‐demographic questionnaire included age, sex, years working with older adults, years with the organisation, job category, shifts worked, employment status, education and previous dementia education.

2. The Self‐Perceived Behavioural Management Self‐Efficacy Profile (SBMSEP); (Crandall et al. 2022; Schindel Martin et al. 2016) in which participants indicated their perceived confidence in accomplishing clinical behaviours and tasks necessary to competently support patients with responsive behaviours. Ten items were measured on seven‐point Likert‐type scales, followed by open‐ended questions about participants' past experiences, usual approaches and perceptions of best practices. The 10 quantitative ratings were summed to form a total self‐efficacy score out of 70, such that a higher total self‐efficacy score indicates greater overall participant confidence in supporting patients with responsive behaviours.

3. An eight‐item multiple‐choice knowledge test (Crandall et al. 2022) assessed knowledge of dementia and person‐centred care strategies. Two versions of the multiple‐choice questions (MCQs) were used to prevent copying responses. Each participant completed the same version across the three time points. The knowledge score consisted of the total number of correct responses out of a maximum possible score of eight.

3. Results

3.1. Quantitative Analyses and Results

Most participants were female and aged between 20 and 49 years (see Table 1). Most participants had either a college diploma or graduate degree as their highest level of completed education and had substantial experience working with older adults. Most were full‐time RPNs, but there were also significant numbers of RNs and Allied Health professionals. Only a small proportion of participants had completed prior dementia‐related education.

TABLE 1.

Demographic frequencies for the participant sample, n = 88.

Variable	Frequency	Percent	Missing
Age			2
20–29	22	25.3
30–39	27	31.0
40–49	23	26.4
50–59	13	14.9
60+	0	0
Gender			2
Male	11	12.6
Female	73	83.9
Other	0	0
Education			2
High School	0	0
Undergraduate Degree	14	16.1
Graduate Degree	31	35.6
College Diploma	40	46
Professional Certificate	0	0
Other	0	0
Years working with older adults			1
< 1	1	1.1
1–2	7	8
2–5	27	31
5–8	13	14.9
8–10	14	16.1
10–20	20	23
20+	4	4.6
Years with organisation			1
< 1	10	11.5
1–2	15	17.2
2–5	26	29.9
5–8	5	5.7
8–10	15	17.2
10+	15	17.2
Job role			2
RN	11	12.6
RPN	51	58.6
Allied Health	17	19.5
Administration	3	3.4
Non‐Clinical	3	3.4
Employment status			3
Part‐time	26	29.9
Full‐time	52	59.8
Casual	6	6.9
Shifts worked			1
Day	61	70.1
Evening	2	2.3
Night	3	3.4
Multiple	20	22.9
Previous dementia education			18
U‐First	0	0
GPA Basics	10	11.5
PIECES	5	5.7
Other	1	1.1
None	51	58.6

3.1.1. Self‐Efficacy Scores

A paired‐samples t‐test comparing mean immediate pre‐ and post‐GPA self‐efficacy scores (n = 84) revealed a significant post‐GPA increase in self‐efficacy (see Table 2 and Figure 1). A one‐way repeated measures ANOVA with Greenhouse–Geisser correction for the measure of self‐efficacy for all complete cases (n = 28) showed that self‐efficacy scores differed significantly across the three time points (see Table 3). Bonferroni‐corrected pairwise comparisons revealed a significant increase in self‐efficacy scores between the immediate pre‐ and immediate post‐GPA measures (p < 0.001) and this increase relative to pre‐GPA scores was sustained at 6–8 weeks (p = 0.006; see Figure 2).

TABLE 2.

Quantitative findings for all complete cases who supplied data for both the immediate pre‐ and immediate post‐GPA measures of self‐efficacy and knowledge in dementia care.

Measure	n	Immediate pre‐GPA		Immediate post‐GPA		t(df)	p
		M	SD	M	SD
Self‐efficacy (SBMSEP), /70	84	46.54	9.72	63.46	5.14	19.74 (83)	< 0.001
Knowledge (multiple choice), /8	82	4.76	1.57	5.85	1.58	7.81 (81)	< 0.001

FIGURE 1.

Mean total self‐efficacy (A) and knowledge (B) in dementia care at two time points: Immediately prior to GPA and immediately after GPA. Stars denote a significant difference (p < 0.05), error bars depict standard error of the mean. Significance is noted in figure captions as p < 0.05.

TABLE 3.

Quantitative findings for all complete cases who supplied data for the immediate pre‐, immediate post‐ and 6 to 8 weeks post‐GPA measures of self‐efficacy and knowledge in dementia care.

Measure	n	Immediate pre‐GPA		Immediate post‐GPA		6–8 weeks post‐GPA		F(df)	p
		M	SD	M	SD	M	SD
Self‐efficacy (SBMSEP), /70	28	42.71	9.72	62.14	5.66	59.50	7.17	110.69 (1.32, 35.75)	< 0.001
Knowledge (multiple choice), /8	29	5.00	1.49	6.14	1.30	5.66	1.50	9.86 (2, 56)	< 0.001

FIGURE 2.

Mean total self‐efficacy (A) and knowledge (B) in dementia care at three time points: Immediately prior to GPA, immediately after GPA and 6–8 weeks after GPA. Stars denote a significant difference (p < 0.05), error bars depict standard error of the mean. Significance is noted in figure captions as p < 0.05.

3.1.2. Knowledge Scores

The paired‐samples t‐test comparing immediate pre‐ and post‐GPA knowledge scores (n = 82) revealed a significant post‐GPA increase (see Table 2 and Figure 1). A one‐way repeated measures ANOVA for the measure of knowledge for all complete cases (n = 29) showed that knowledge scores at the three time points differed significantly from each other (see Table 3). Mauchly's test confirmed the assumption of sphericity was not violated. Bonferroni‐corrected pairwise comparisons showed a significant increase in knowledge scores between the pre‐ and post‐GPA measures (p < 0.001) and this significant increase relative to pre‐GPA scores was not quite sustained 6–8 weeks post‐intervention (p = 0.06). There was no significant change in knowledge scores from immediately post‐GPA to 6–8 weeks later (p = 0.224, see Figure 2).

3.1.3. Overall Summary of Quantitative Findings

Quantitative findings show that immediately after participants attended GPA, their mean scores on measures of knowledge and self‐perceived efficacy in person‐centred dementia care significantly increased relative to baseline. Increases in self‐efficacy were substantial and rapid, and although there was a significant decline after 6–8 weeks, the significant increase relative to baseline was maintained. Increases in knowledge showed a similar pattern, although knowledge scores at 6–8 weeks were not significantly lower than the immediate post‐GPA scores. Overall, results suggest that at baseline, there were gaps in participants' confidence and knowledge that were addressed through participation in GPA.

3.2. Qualitative Analyses and Results

Qualitative data from the self‐efficacy measure were subjected to reflexive thematic analysis (Braun and Clarke 2022; Braun and Clarke 2006) and were analysed using a step‐by‐step inductive approach to arrive at the final thematic categories. Participant responses were analysed as a group within each time point and not matched for each individual across time points. Compelling and vivid example responses for each theme were selected by the evaluation team and are reported below.

3.3. Immediate Pre‐GPA Open‐Ended Questions

Seventy‐nine participants answered the three baseline open‐ended questions about their current approaches, perceived need for practice change and knowledge of best practices when caring for PLwD and responsive behaviours. Three main themes emerged from the participants' responses: Basic Approaches to Responsive Behaviours, Existing Person‐Centred Approaches and Need for Additional Responsive Behaviour Strategies.

3.3.1. Basic Approaches to Responsive Behaviours

The most frequently reported approach to responsive behaviours was for participants to control their own emotional response through remaining calm (n = 35). Specific strategies to achieve or sustain emotional calmness were not explicated. Some respondents indicated that calmness required a conscious effort, for example, they ‘try to stay calm’ (#23), but might not achieve this.

Many participants noted that modifying their tone of voice and body language was essential to prevent further escalation of responsive behaviours when caring for PLwD. As one participant reported, ‘if a patient approaches me at the nursing station and they are being aggressive, I approach them in a calm and patient way, with a soft‐spoken voice’ (#2). At baseline, most participants had the basic insight that caring for PLwD involves a reciprocal relationship influenced by the emotional state of both the care provider and the care recipient. However, relatively few participants described specific and systematic person‐centred communication strategies that go beyond remaining calm and being polite, which are general techniques not specific to professional healthcare providers and not necessarily effective at mitigating the emotional response of the care provider or PLwD.

Another frequently mentioned basic strategy was leaving the PLwD alone and returning sometime later. For example, one participant stated, ‘I try to stay calm, if the situation is out of control/safety, leave the site and report. Give some time for person to calm down and approach later’ (#23). While participants recognised their presence might be contributing to the responsive behaviour, they did not identify the concurrent need to adapt care practices. A best practice approach would involve remaining with the person to support their de‐escalation with purposeful, non‐pharmacological interventions. Given that responsive behaviour could emerge with each subsequent attempt to initiate care, a brief rest period without care adaptation will not sustain calmness in the PLwD.

3.3.2. Existing Person‐Centred Approaches

A small number of participants were already using systematic person‐centred strategies at baseline. Some recognised their responsibility to know the person's history and preferences and how these might influence the likelihood of responsive behaviours. One participant said ‘I read, get to know the [patient's] holistic history and go from there. Knowing what triggers the [patient] and what makes [them] calm, specifically the old memories of them as they forget the very recent memories’ (#3). These participants were correctly interpreting responsive behaviours not simply as a component of their dementia diagnosis, but instead as attempts by the PLwD to communicate unmet needs. Two participants understood the importance of addressing the person's unmet needs before redirection was likely to be effective, for example: ‘I listen respectfully and try to address his/her needs. I use a calm voice to explain how I can try to help. Redirect and re‐engage the person in another activity/conversation’ (#24). Additional baseline person‐centred strategies included offering choices, providing care at an appropriate pace and involving the person's family.

3.3.3. Need for Additional Responsive Behaviour Strategies

Although some participants described person‐centred approaches they were already using at baseline, many others did not feel adequately prepared when faced with responsive behaviours and stated they would like to learn additional strategies. Many participants felt fear or anxiety when caring for a person showing responsive behaviours, and this fear prevented them from reacting with confidence and skill. As one participant stated, ‘I am scared, trying to de‐escalate the situation but not confident if the approach I know is right’ (#27). Several participants reflected that their fear would sometimes lead to the patient being prescribed medication, for example, ‘I was scared and sometimes freeze with their behaviour, I would look to give medication when receiving communication and help to alleviate the behaviour’ (#83). Others were aware that their own fear was likely to exacerbate the situation: ‘I escalate the situation as I feel I do not have the knowledge and skills currently to address the situation personally’ (#17) and ‘I try to stay at a distance. I am tempted to call Code White, but I don't always feel that is appropriate and [I'm] concern[ed] it will escalate the situation’ (#60). Importantly, this response indicates that organisational protocols such as Code White are considered as a non‐proportional or excessive approach to some responsive behaviour.

Participants hoped to expand their knowledge of evidence‐based best‐practice strategies for preventing and de‐escalating responsive behaviours. One participant wanted to ‘improve toolbox of skills/approaches, use a framework—purposeful/thoughtful approaches versus current instinctive approach to responding, improve documentation about responsive behaviours—accuracy, descriptiveness’ (#21). Others wanted to improve their ability to identify precursors to responsive behaviours and learn communication strategies that they could use once the behaviours were already occurring, for example, ‘[I would like to] enhance my knowledge so that I know the right things to say to de‐escalate the situation’ (#17). Respondents were looking for education about precise techniques to control negative emotions, improve the quality of verbal/physical interactions and document observations to mitigate interpretive bias in assessment procedures.

Overall, immediate pre‐GPA qualitative responses indicated most participants used general interpersonal strategies when faced with PLwD who were agitated or upset and recognised the importance of remaining calm without describing specific ways in which this could be accomplished. It is promising that several participants were already implementing some person‐centred approaches, although many felt fear and discomfort when faced with escalating responsive behaviours. Participants recognised that their ability to provide optimal care could be enhanced by specific education in non‐pharmacological interventions.

3.4. Immediate Post‐GPA Open‐Ended Questions

Immediately after GPA, 68 participants again answered the same three open‐ended questions as at baseline. Responses are categorised here under the emergent themes of Improved Self‐Awareness and Confidence and New Non‐Pharmacological Interventions.

3.4.1. Improved Self‐Awareness and Confidence

Immediately after GPA, many participants expressed understanding of how their previous approaches to care may have unknowingly provoked or exacerbated responsive behaviours. One participant realised that prior to GPA, they ‘haven't thought much about seeing from the patient's viewpoint’ (#25). Other participants acknowledged they had previously used approaches that might be considered punitive, stating ‘I usually react to them that they are old enough and should not do that. I would [like to] change on how I label them and outpace them’ (#18), and ‘I was aggressive and guarding bound[aries] when reacting to a client with dementia’ (#34).

Several participants reported increased confidence in caring for PLwD and responsive behaviours because of the GPA strategies they had learned. One said, ‘I was very intimidated because this was something I was never taught. I am still apprehensive but more confident as I now have some training and techniques I can apply’ (#17). Another expressed that knowledge of GPA self‐protective techniques would alleviate their anxiety when faced with physical responsive behaviours, stating that they would now ‘utilize self‐protective techniques instead of panicking’ (#75).

3.4.2. New Non‐Pharmacological Interventions

Immediately after GPA, participants could name a range of person‐centred approaches as best dementia care practices they planned to use in the future. Some of these included ‘consider[ing] past history of people with dementia and use to organize their care plan’ (#5) and identifying precursors of responsive behaviours and documenting them once identified. Participants stated that they would use ‘STOP & GO’ as a purposeful technique to provide small breaks during care while remaining present to support behavioural de‐escalation. Many respondents could also now identify the four crucial components of successful verbal redirection, understanding they are unlikely to be able to distract a PLwD without first validating their feelings and joining in their reality (Advanced Gerontological Education 2014). For example, one participant said ‘Listen to them, understand the trigger of the behaviour. Understand their past and needs. Validate to what they are saying instead of quickly diverting the topic’ (#1). Lastly, many participants noted the utility of the physical self‐protective techniques such as the thumb and reflex grab releases, suggesting that they are encountering these situations frequently in their clinical practice.

3.5. Six to Eight Weeks Post‐GPA Open‐Ended Questions

Six to eight weeks after GPA, 26 participants again answered the same three questions as previously and were asked to consider whether they had encountered any instances of responsive behaviour since attending GPA. They were asked whether they had implemented any GPA strategies and if so, to reflect on their effectiveness. Responses are presented under the single predominant theme of Growing Confidence in the Practical Application of Nonpharmacological Approaches to Responsive Behaviour. In their responses, participants mentioned as best practices many of the person‐centred GPA strategies they had described at the immediate post‐GPA time point, including STOP & GO, validation, successful verbal redirection, self‐protective techniques, identifying unmet needs, being mindful of invading personal space and adjusting the pace at which care is provided. Even after several weeks, participants could still recall and apply their new knowledge of specific GPA approaches. Other participants provided responses showing increases in confidence relative to baseline were still evident 6–8 weeks later, for example, ‘I tried to avoid them (people and situations) previously. I feel more confident that I can calmly offer assistance now that I have received this GPA training’ (#17).

Fourteen participants reported instances of responsive behaviours during the interval between their GPA session and the third evaluation time point. For example, one participant said:

Today also one of my patients was constantly screaming and yelling at the beginning of my shift. I used Stop and Go method, I went to her room, stopped whatever I was doing and gave her my undivided attention, observed her cues and planned my care towards her. She became calm and when [I] asked her what [she] really needs she told me she needs her adult brief to be changed. Thus, I did exactly what she asked for then she went back to sleep. I am so glad that I took this GPA course, now I have a new way of understanding my patients. I can give the best possible care for my patients [that] I always wanted to give. (#1)

This participant recognised that in combination with careful observation, PLwD can provide useful information about their unmet needs when asked. Another participant used a similar strategy to support a patient who wanted to be discharged earlier than planned. The participant engaged with the patient to find out the reason for their distress and help meet their needs. The participant (#11) said ‘… [the] patient was missing his family’ and the participant helped the patient to ‘connect with family members in a few ways’. The participant reported feeling ‘not comfortable at the beginning, but after analyzing the situation [I] got the problem resolved’.

Due to anxiety or time constraints, participants had not previously fully engaged with PLwD during episodes of responsive behaviours, therefore sacrificing opportunities to gain information about how the patients' needs might be accommodated. As one participant stated, before GPA they would ‘… just let them say what they want to say and never listen to them just to get my job done. I would [like to] change by communicating to them with respect and dignity. I would also give them my full attention and not just because I need to go home on time’ (#18).

Overall, the qualitative findings after 6–8 weeks show that increases in knowledge and confidence were sustained over time and that participants were able to apply GPA approaches to successfully de‐escalate responsive behaviours. They had investigated unmet needs underlying behaviours they encountered and had used the information they discovered to provide solutions tailored to the needs and preferences of the patient. Participants' responses indicate that GPA training had significantly enhanced their ability to provide respectful, person‐centred care using actionable core competencies.

4. Discussion

In this program evaluation, staff at a post‐acute care hospital completed the GPA dementia educational program as well as pre‐ and post‐intervention evaluation measures. GPA, which delivers best‐practice knowledge and competencies in caring for people living with dementia and responsive behaviours, had significant positive impacts on staff self‐efficacy and knowledge in dementia care, and these improvements were sustained over a period of at least 6–8 weeks. Before GPA, many participants reported feeling anxiety and fear when faced with escalating responsive behaviours, and participants hoped to learn new communication strategies and techniques they could use in their practice. After GPA, participants had gained knowledge of new person‐centred, actionable, non‐pharmacological interventions that could prevent and de‐escalate responsive behaviours. Moreover, in the 6–8 weeks following GPA, participants reported having been able to put these learned strategies to use and described scenarios in which they had demonstrated an enhanced ability to confidently and skillfully meet the medical and psychosocial needs of their patients living with dementia. Participants' descriptions included instances of competent application of psychological supports and inclusion of the PLwD as a care partner. These findings support the conclusion that a formal education program about responsive behaviour has potential to expand the competencies of nurses who work in the post‐acute care setting. Introduction of such programs would meet identified dementia‐specific knowledge gaps in the post‐acute context (Drake et al. 2019; Plys et al. 2022; Semelka et al. 2024).

Qualitative findings indicate that participants did learn practical and effective non‐pharmacological approaches that enabled them to successfully de‐escalate situations involving responsive behaviours. Before GPA, participants reported that they were more likely to utilise pharmacological interventions as a potential means to de‐escalate behaviours. After GPA, staff responses reflected an awareness of the importance of using non‐pharmacological interventions as a first‐line approach. The use of antipsychotics for responsive behaviour does not change the experience of fear for PLwD and does not address the person's underlying unmet needs (Bueckert et al. 2019; Seniors Health Strategic Clinical Network 2024). It is therefore critically important for organisations to provide staff with a range of education and tools, including programs like GPA, that will allow care providers to implement non‐pharmacological approaches for responsive behaviours and to only use pharmacological interventions when clinically indicated. The GPA dementia education program allowed us to assist staff to identify when antipsychotics were unwarranted and increase their confidence that they had the non‐pharmacological means to care for the patients that might otherwise be sedated unnecessarily.

4.1. Implications for Sustainability

Participant completion rates for the evaluation measures at 6–8 weeks post‐GPA were relatively low, but while attrition can contribute to bias in statistical analyses, it is a complex process that depends on both individual factors and contextual factors of the intervention (Amico 2009; Marcellus 2004). Attrition among healthcare staff during practice change program evaluations has been shown to be related to scheduling/workload difficulties and staff turnover and not necessarily due to dissatisfaction or disengagement with the intervention (Pascoe et al. 2021; Kreiter et al. 1999). Successful uptake and sustainability of GPA education within our organisation will depend in part on continuous facilitation and support for team members in applying best practice approaches (Kitson et al. 1998). This support will consist of recurrent advanced practice modelling and reinforcement, huddles with discussions about application, incentivisation and commendation of consistent best practice. Organisational leadership engagement is critical for effective implementation and uptake of person‐centred dementia care approaches (Casarez and Smith 2024; Rokstad et al. 2015). Within our organisation, nursing leaders are experiencing the same educational content as staff care providers, such that all team members can then ground their practice in the same best practice concepts and communicate using the same professional care language. GPA sessions continue to be provided to clinical staff at orientation and on an ongoing monthly basis. Including GPA as part of the staff onboarding process ensures they are equipped with the tools and resources to approach dementia care with knowledge and confidence, which is essential for retaining staff in a complex environment.

4.2. Strengths and Limitations

The primary strength of this project is the novel facilitation and evaluation of dementia education related to responsive behaviours in a post‐acute care hospital. While dementia education programs have been implemented and evaluated within acute care settings (for a systematic review, see Karrer et al. (2021)), there is very limited existing evidence for this type of education in post‐acute care, and our evaluation aligns with the need for additional research related to dementia‐related behavioural symptoms in post‐acute facilities (Drake et al. 2019). Within Ontario, 85% of residents in hospital‐based continuing care in 2023–2024 were over the age of 65; 18.2% of residents had a diagnosis of dementia and 37.1% had moderate to very severe cognitive impairment (Canadian Institute for Health Information 2024b). Moreover, in 2022–2023, 19% of residents in complex continuing care hospitals/units had behaviour classified as mild to moderately aggressive and 3% as severely aggressive (Canadian Institute for Health Information 2024a). In this population of older adults, there is a complex intersection of multiple medical comorbidities (including dementia and its behavioural and psychological symptoms), and education is required to equip nurses in post‐acute hospital settings with the skills necessary to support these patients' complex behavioural needs.

Limitations of this program evaluation include the focus on participants' self‐efficacy and knowledge outcomes without accompanying information on rates of reportable incidents of responsive behaviours of risk. Instead, participants' qualitative responses revealed the outcomes of their application of GPA approaches. Although it is conceivable that dementia education programs may impact rates of staff injury related to escalating responsive behaviours (Mohr et al. 2022), and data on staff injury rates were available to us, the data did not allow us to determine the relationship between the injuries and the presence of a patient diagnosis of dementia or expressions of responsive behaviour. Because of the intrinsic and sometimes unobservable nature of factors like pain, hunger, and discomfort, it is not possible to anticipate and prevent 100% of responsive behaviours in the dementia context. But when staff are equipped to identify escalating behaviour and use GPA strategies to protect themselves and patients from injury, it is reasonable to assume that there could be some impact on reported injury rates. Another limitation was participant attrition in completion of the evaluation measures at the third evaluation time point. Consequently, the stories of successful application of GPA principles and skills, while evocative, are limited. Incentivisation for completion of all measures would likely yield a greater response rate at this final follow‐up evaluation.

4.3. Implications for Future Research

Future research on responsive behaviours in the post‐acute care setting can include evaluation of the types of non‐pharmacological interventions that are most often effective in this context. This could involve documentation of specific care adaptations staff have made to meet the psychological safety of persons with dementia. A longer post‐intervention evaluation follow‐up period would also likely allow participants to report more instances of skilful application of tailored, non‐pharmacological approaches as participants build their competency over time. Research establishing a link between dementia education and rates of responsive behaviours and/or rates of injury will require substantial funding to tailor organisational reporting. For example, the Behavioural Supports Ontario Dementia Observation System (DOS Working Group 2019), which measures the frequency, duration and risk level associated with responsive behaviour aligned with the specific time sequence of an incident report, could be used to capture deeper information about the nature, quality and context of responsive behaviours.

5. Conclusion

This evaluation provides valuable information regarding the self‐efficacy, knowledge acquisition and perceptions of skill application of professional caregivers supporting PLwD and responsive behaviours who have been admitted to post‐acute care. Our findings have practical implications for clinical practice educators, healthcare decision‐makers and policymakers involved in strategic improvement planning for PLwD. Our experience provides evidence that post‐acute hospitals require gerontological nursing education, in particular dementia care nursing competencies specific to responsive behaviours. Without it, post‐acute care staff will be unprepared to care for a population of progressively older people with complex chronic illnesses, including dementia, who may react to the care environment in responsive ways. Continued quality dementia education, of which GPA is one form, will assist post‐acute hospitals in continuing to provide the highest levels of compassionate, respectful and patient‐centred care.

Author Contributions

Karimah Alidina: conceptualisation (lead), qualitative analysis (equal), writing – original draft (equal), review and editing (equal). Victoria McLelland: conceptualisation (supporting), quantitative methodology and analysis (lead), qualitative methodology and analysis (equal), writing – original draft (equal), writing – review and editing (equal). Lori Schindel Martin: conceptualisation (supporting), qualitative methodology and analysis (equal), writing – original draft (equal), writing – review and editing (equal).

Funding

The authors have nothing to report.

Ethics Statement

This project was designated as a program evaluation by the institutional review board and therefore ethical approval for research was not required. Standards as explicated by the Tri‐Council Policy Statement: Ethical Conduct for Research Involving Humans (2018) were followed.

Conflicts of Interest

The authors declare no conflicts of interest related to this manuscript. Two of the authors, K.A. and L.S.M., currently serve on the volunteer governance Board of Directors for Advanced Gerontological Education (AGE), a not‐for‐profit organisation that develops and ensures fidelity of the Gentle Persuasive Approaches in Dementia Care program. These authors were not board members at the time of intervention selection, data collection or analysis. Additionally, the evaluation was conducted outside the influence of any board members, being in the jurisdiction of AGE's independent Research & Development Committee. The purpose of this evaluation is to disseminate best practices in responsive behaviour care. All authors have thoroughly reviewed and approved the final version of this manuscript.

Data Availability Statement

Data available on request from the authors.