The Effectiveness and Feasibility of Non-Pharmacological Interventions for Reducing Behavioural and Psychosocial Symptoms of Dementia and Improving Patient Experience in Acute Care Settings: A Systematic Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

 

 

 The Effectiveness and feasibility of non-pharmacological interventions for reducing behavioural and psychosocial symptoms of dementia and improving patient experience in acute care settings: A Systematic Review

An interesting study on the the effects of non pharmacological interventions for reducing behavioural and psychosocial symptoms of dementia and in acute care settings -- assessed by reviewing literature in published in the last 11 years in 7 relevant databases.

Introduction

Well written. In line with the study purpose. Relevant references.

Methodology -well detailed. Insures study replicability. an appropriate design, conducted by PRISMA guidelines. PICO framework used

Results

Clearly described and well organized on categories

Discussions

Great analysis  - objective interpretation.

 

Minor recommendations:

• Do not use the PWD acronym
• Remove (or correct) references on page 28
• Avoid using meaningfully (conclusion section) as no statistics are provided to support the use of this term

Author Response

Comment: Do not use the PWD acronym.

Response: Thank you for this comment. We appreciate the importance of respectful terminology when referring to people with dementia. In response, we have removed the acronym “PWD” throughout the manuscript and now use the term “people with dementia” in full.

Comment: Remove (or correct) references on page 28.

Response: Thank you, we have looked at the references on page 28 and made adjustments.

Comment: 

Avoid using meaningfully (conclusion section) as no statistics are provided to support the use of this term.

Response: Thank you, in response to this comment, we have revised the wording in the discussion to adopt a more cautious interpretation of the findings. Specifically, we now emphasise that while non-pharmacological approaches may show promise for reducing agitation and distress in people with dementia in acute care settings, the evidence is variable and effects are not always sustained. The revised text also highlights the need for further well-designed studies with consistent outcome measures and longer follow-up periods to better determine the durability of these interventions.

Reviewer 2 Report

Comments and Suggestions for Authors

I congratulate the authors to their well-executed and well-written systematic review. I have only few, minor comments:

 

1) Clarify/be specific throughout the manuscript (including in the title) if you looked at BPSD in general or only at anxiety and/or agitation.

2) In the abstract, you also claim to aim at looking at improvement of patient experience. Do you really do that?

3) Please provide data on the severity/stage of dementia in the included studies. This is very important to understand if findings can be transferred to other populations.

Author Response

Comment 1: Clarify/be specific throughout the manuscript (including in the title) if you looked at BPSD in general or only at anxiety and/or agitation.

Response: Thank you for this helpful comment. The focus of the review was on behavioural and psychological symptoms of dementia more broadly, rather than exclusively on anxiety or agitation. We recognise that the wording in parts of the manuscript may have implied a narrower focus on specific symptoms. As agitation and anxiety are commonly reported components of BPSD, several of the included studies measured these outcomes specifically using validated assessment tools. We have clarified this from line 105. Furthermore we have revised the manuscript to clarify that the review examines non-pharmacological interventions targeting BPSD in people with dementia in acute care settings, with agitation, anxiety, emotional distress, and related behavioural symptoms considered within this broader symptom cluster. We have also adjusted the wording throughout the manuscript to ensure that the scope of the review is consistently described.

Comment 2: In the abstract, you also claim to aim at looking at improvement of patient experience. Do you really do that?

Response: Thank you for highlighting this point. Patient experience was not assessed using direct patient-reported outcome measures in most of the included studies due to cognitive impairment and the acute care context. However, we believe that aspects of patient experience were explored indirectly through observational measures of wellbeing and engagement (e.g., pleasure, alertness, responsiveness), behavioural indicators of distress, and qualitative feedback from staff and caregivers regarding patient responses to interventions.

For clarity, we have revised the wording in the abstract to specify that the review considers indicators of patient experience such as wellbeing, engagement and observable responses, rather than direct patient-reported experience measures. From line 12.

Comment 3: Please provide data on the severity/stage of dementia in the included studies. This is very important to understand if findings can be transferred to other populations.

Response: Thank you we agree about the importance of dementia severity when considering the transferability of findings and that the stage of dementia is an important factor influencing behavioural symptoms such as agitation and the potential responsiveness to non-pharmacological interventions. However, the included studies did not consistently report the stage or severity of dementia using standardised measures. While some studies indicated that participants had a diagnosis of dementia no clear categorisation into mild, moderate, or severe stages was specified.

Due to this inconsistency in reporting, it was not possible to reliably synthesise or present comparable data on dementia stage across the included studies. We have clarified this limitation in the manuscript and now explicitly note that the lack of consistent reporting of dementia severity limits the ability to determine how transferable the findings are to specific dementia populations. Future studies would benefit from clearly reporting dementia stage using standardised measures to improve comparability and applicability of results. 

Reviewer 3 Report

Comments and Suggestions for Authors

General Feedback

This manuscript addresses a relevant clinical issue and contributes to the growing literature on dementia-friendly acute care practices. However, several methodological and reporting limitations currently diminish the scientific rigor and clarity of the review. Specifically, issues related to methodological transparency, synthesis depth, bias assessment, and evidence interpretation must be addressed before the manuscript can be considered for publication.

Major Comments

1️. Descriptive Evidence Synthesis Lacks Analytical Depth: On page 16, the authors categorize interventions into three groups (music-based, psychosocial, and sensory), yet the comparative strength of the evidence is not critically examined. A more robust synthesis should incorporate a deeper analytical layer by comparing study designs, such as randomized controlled trials versus observational studies, while also evaluating intervention intensity and the validity of various outcome measurements. Furthermore, the synthesis should include a critical discussion regarding the consistency of effects across the evidence base; without such a rigorous approach, the conclusions risk appearing over-generalized.

2. Insufficient Reporting of Risk of Bias Assessment : The manuscript notes that the Mixed Methods Appraisal Tool (MMAT) was used for quality assessment. While the authors claim studies were of "medium to high quality" (page 17), this assertion is not well-justified. For a systematic review, the authors should provide a clear risk-of-bias table and a critical discussion of bias sources, such as small sample sizes, lack of randomization, observational designs, and short follow-up periods.

3️. Overinterpretation of Effectiveness: The discussion states that non-pharmacological interventions can “meaningfully reduce agitation and distress” in acute care. However, several included studies showed no sustained effects beyond the intervention periods, inconsistent reductions in agitation, or mixed findings. The conclusions should be more cautious and better reflect these inconsistencies.

4️. Inconsistency in Outcome Measures : A wide range of outcome measures—such as the PAS, CMAI, NPI-Q, OERS, MPES, and MiDAS—were used across the studies. These tools capture distinct constructs, including agitation, emotional engagement, behavioral disturbances, and mood. Consequently, comparing intervention effects is difficult due to this heterogeneity. Specifically, the conclusion regarding “anxiety reduction” may be misleading given the lack of standardized measurement for that construct.

Minor Comments

1. The manuscript states: 747 participants in the abstract, inconsistencies numbers that 749 participants showed in the results section. This discrepancy should be corrected.
2. 2. Figure 2 conceptual model: The conceptual pathway (Figure 2) is interesting but appears not directly derived from systematic analysis. The authors should clarify that whether this model emerged from data synthesis or represents a conceptual framework proposed by the authors.

Author Response

Major Comment 1: Descriptive Evidence Synthesis Lacks Analytical Depth: On page 16, the authors categorize interventions into three groups (music-based, psychosocial, and sensory), yet the comparative strength of the evidence is not critically examined. A more robust synthesis should incorporate a deeper analytical layer by comparing study designs, such as randomized controlled trials versus observational studies, while also evaluating intervention intensity and the validity of various outcome measurements. Furthermore, the synthesis should include a critical discussion regarding the consistency of effects across the evidence base; without such a rigorous approach, the conclusions risk appearing over-generalized.

Response: Thank you for this helpful suggestion. We agree that a critical synthesis of the strength and consistency of the evidence is important. In response, we have expanded the discussion to more explicitly consider differences in study design, intervention intensity and outcome measurement across the included studies. This provides a clearer appraisal of the comparative strength and consistency of the evidence and help contextualise the conclusions of the review. 

Comment 2: Insufficient Reporting of Risk of Bias Assessment : The manuscript notes that the Mixed Methods Appraisal Tool (MMAT) was used for quality assessment. While the authors claim studies were of "medium to high quality" (page 17), this assertion is not well-justified. For a systematic review, the authors should provide a clear risk-of-bias table and a critical discussion of bias sources, such as small sample sizes, lack of randomization, observational designs, and short follow-up periods.

Response: Thank you for raising this point regarding the reporting of the quality appraisal. The methodological quality of included studies was assessed using the Mixed Methods Appraisal Tool (MMAT). At least two members of the review team independently appraised the studies, and any discrepancies were resolved through discussion to reach consensus. Based on the MMAT criteria, the majority of studies met most of the methodological quality indicators and were therefore categorised as being of medium to high quality.

We acknowledge that several methodological limitations were present across the included studies, including small sample sizes, limited randomisation, observational or quasi-experimental designs, and short follow-up periods. These sources of potential bias are now more explicitly discussed in the manuscript to provide greater transparency regarding the strength of the evidence. As the number of included studies was relatively small, we have incorporated this appraisal within the narrative synthesis rather than presenting an additional risk-of-bias table. This is clarified in Section 2.9 and in the limitations section.

Comment 3:

Overinterpretation of Effectiveness: The discussion states that non-pharmacological interventions can “meaningfully reduce agitation and distress” in acute care. However, several included studies showed no sustained effects beyond the intervention periods, inconsistent reductions in agitation, or mixed findings. The conclusions should be more cautious and better reflect these inconsistencies.

Response: Thank you for this important observation. We agree that the evidence across the included studies is heterogeneous and that the effectiveness of non-pharmacological interventions was not consistently demonstrated.

In response to this comment, we have revised the wording in the discussion to adopt a more cautious interpretation of the findings. Specifically, we now emphasise that while non-pharmacological approaches may show promise for reducing agitation and distress in people with dementia in acute care settings, the evidence is variable and effects are not always sustained. The revised text also highlights the need for further well-designed studies with consistent outcome measures and longer follow-up periods to better determine the durability of these interventions. 

Comment 4: Inconsistency in Outcome Measures : A wide range of outcome measures—such as the PAS, CMAI, NPI-Q, OERS, MPES, and MiDAS—were used across the studies. These tools capture distinct constructs, including agitation, emotional engagement, behavioral disturbances, and mood. Consequently, comparing intervention effects is difficult due to this heterogeneity. Specifically, the conclusion regarding “anxiety reduction” may be misleading given the lack of standardized measurement for that construct.

Response: Thank you for highlighting the heterogeneity in outcome measures used across the included studies. We agree that a range of instruments were employed. We believe that these measures assess related but distinct constructs, including agitation severity, behavioural and psychological symptoms, emotional expression, and engagement. This reflects the multidimensional nature of agitation and behavioural disturbance in dementia, which are often conceptualised as overlapping clusters of behavioural, affective, and engagement-related symptoms rather than a single construct.

However, we acknowledge that the use of different outcome measures introduces heterogeneity and limits the direct comparison of intervention effects across studies. We have clarified this limitation in the manuscript and have moderated the interpretation regarding anxiety reduction, noting that anxiety was not consistently measured using a dedicated or standardised instrument across studies.

Minor Comments 1:The manuscript states: 747 participants in the abstract, inconsistencies numbers that 749 participants showed in the results section. This discrepancy should be corrected.

Response: Thank you this has been corrected to 749.

Comment 2: Figure 2 conceptual model: The conceptual pathway (Figure 2) is interesting but appears not directly derived from systematic analysis. The authors should clarify that whether this model emerged from data synthesis or represents a conceptual framework proposed by the authors.

Response: Thank you this has been clarified to a conceptual framework proposed by the authors.