PROTOCOL: Factors influencing the implementation of non-pharmacological interventions for behaviours and psychological symptoms of dementia in residential aged care homes: A systematic review and qualitative evidence synthesis

Types of studies

Only qualitative data will be included for evidence synthesis to answer the research question of the planned systematic review. The sources of qualitative data include studies that employ various data collection methods including, but are not limited to, focus group discussions, interviews, qualitative surveys, ethnographies, and data analysis methods that encompass, but are not limited to, framework analysis, content analysis, and thematic analysis. Additionally, the qualitative components of the mixed methods studies will be included. However, qualitative studies with unsupported findings (i.e., findings not accompanied by illustrations or supporting evidence) will be excluded from the final synthesis. Studies will be selected if they report at least one implementation factor whether it is a barrier, facilitator, enabler, or related term irrespective of whether the word ‘implementation’ is present in the article or not. Implementation factors pertaining to the staff, families/relatives, volunteers, managers/leaders from their experience, that is, studies reporting the views and experiences of RACH staff, people with dementia, their families, and volunteers will be included. Only studies published in the English language and available in full text from the date of the inception of databases up until 31 December 2023 will be considered for inclusion. Articles published in a language other than English will be excluded because of resource constraints to retrieve these types of articles.

Types of participants

The people living with dementia in RACHs, who are exhibiting BPSD will be included. Any study focusing on the general elderly population or a mixed population of people with dementia and the general elderly where there is no separate result presented for BPSD, and people with dementia who have no BPSD will be excluded.

Types of interventions

The planned systematic review will include any NPI relevant to BPSD, neuropsychiatric symptoms, changed behaviours, responsive behaviours, or other synonyms for these terms in people with dementia. Examples of these interventions include music therapy, reminiscence therapy, validation therapy, meaningful activities, light therapy, multisensory stimulation, exercise therapy, a training programme for the care staff, person-centred care, and Namaste care, to name a few. Other non-pharmacological procedures that are not specifically used for BPSD such as surgery, radiation therapy, and imaging will not be included.

Types of outcome measures

The outcomes of this systematic review will include a summary of findings regarding factors influencing the implementation of NPIs, drawing from the views and experiences of various stakeholders. These stakeholders include the RACHs staff (caregivers or personal care workers, nurses, nurse aides or assistants, allied health professionals, e.g., physiotherapists), physicians, volunteers, dementia residents' families, and the dementia residents themselves. The primary outcome will be a summary of the factors influencing the implementation of NPIs for BPSD at RACHs.

Types of settings

In this review protocol, the setting RACHs which refers to residential facilities or long-term care or nursing homes or housing for the elderly or residential homes or assisted living facilities or homes for the aged or aged care homes or halfway house(s) or group home(s) or retirement communities or sanatoriums or housing for older persons will be included. Any study focusing on other settings like hospitals, adult day care centres, homes, or community settings will be excluded. Overall, the type of studies, population, settings, and interventions discussed above are used as inclusion criteria.

Electronic searches

The search terms were developed with the assistance of the academic librarian at the University of South Australia. A preliminary search was conducted by entering the title of the systematic review into Google Scholar, and the Google search engine. The development of search terms for the research question focuses on four key aspects: the population (residents living with dementia), the phenomenon of interest (types of NPIs specifically used for managing BPSD), the context (RACHs), and the study type (qualitative studies).

Both Medical subject heading (MeSH) and keywords were used for each of these aspects.

An initial search was conducted on MEDLINE to locate articles about the research question. The keywords and index terms that appear in the titles, abstracts, and indexes of relevant articles were used to create a thorough search plan for MEDLINE. A set of articles that meet the inclusion criteria and are unambiguously eligible were chosen to authenticate the MEDLINE search plan. The search plan was translated, along with all the relevant MeSH terms, from MEDLINE to other databases such as EMCARE, Embase, CINAHL Complete, and APA Psych Info. The search in the selected databases has already been conducted up until 14 March 2023. Subsequent search updates will be performed up until 31 December 2023. A detailed listing of each database searched, and a line-by-line search is included in Supporting Information: Appendix 1. The OVID platform was used to search for articles in MEDLINE, EMCARE, Embase and APA Psych Info, while EBSCOhost was used to search for articles in CINAHL.

Searching other resources

The relevant sources for grey literature including, Trove, and ProQuest dissertations, theses, and conference abstracts such as Alzheimer's Association International Conference as well as reference lists of included studies and any relevant systematic reviews will be identified. Due to time and resources constraints, contacting individuals and organisations for information about unpublished or ongoing studies will not be performed. Since any relevant published studies may be captured by the databases selected such as MEDLINE, Embase, PsycINFO, EMCARE, CINAHL, there is no plan for hand searching of a specific journal.

Description of examples of methods used in primary research

♦

Interviews are a popular qualitative method of data collection that involves asking questions to a single study participant or a group of participants in more natural settings. It can be conducted face-to-face, on the telephone, or through a video chat (Hancock, 2009).

Selection of studies

The articles identified from the searches will be uploaded by the primary author (HA) into EndNote (version 20.2.1) to remove duplicates. After removing duplicates, the remaining articles will be imported into the JBI Database of Systematic Reviews and Implementation Reports (JBI SUMARI) software (Aromataris, 2020). The JBI SUMARI is primarily designed for qualitative systematic review and employs a meta-aggregative method to synthesise qualitative evidence. One key characteristic of this approach is the ability to generate generalisable recommendations presented as a set of statements, which guide practitioners and policymakers (Aromataris, 2020). The initial stage will involve two reviewers (HA and GK) screening the title and abstract of the articles to exclude articles that are irrelevant using the JBI SUMARI tool. In the second stage, the full text of the remaining articles will be reviewed to include articles for final analysis. All disagreements in determining eligibility will be discussed between the two reviewers to reach an agreement. In case of ongoing disagreement, we will then discuss the disputed article(s) with a third author (TN or MA) to reach a consensus.

Data extraction and management

A standardised tool will be used to extract information from the included qualitative studies (see Supporting Information: Appendix 2) within JBI SUMARI (Aromataris, 2020). Data extraction involves extracting details on the setting, participants, intervention, study methodology, and conclusions of the authors for each qualitative study. Data extraction will also include the list of factors influencing the NPI implementation along with their corresponding illustration. The primary author HA will conduct the data extraction process.

Assessment of risk of bias in included studies

The methodological quality of the included articles will be assessed independently by two reviewers (HA and GK), using the standardised JBI SUMARI critical appraisal tool (Aromataris, 2020). The critical appraisal result will be presented in both narrative and tabular formats. The extraction of the list of factors influencing the implementation of NPIs, and their aggregation will be performed on all included articles, regardless of the outcomes of their quality assessment. The cut-off point will be decided by the reviewers (GK and HA) before conducting a risk of bias assessment. All disagreements in determining the quality assessment scoring will be discussed between the two reviewers, to reach an agreement. In case of ongoing disagreement, we will then discuss the disputed article(s) with a third author (TN or MA) to reach a consensus.

Data synthesis

A meta-aggregation, a pragmatic synthesis method for qualitative studies that was developed at JBI, will be used to synthesise and summarise the practicalities and usefulness of the findings (Korhonen, 2013) in which each verbatim extracted finding will be accompanied by a participant's voice-quoted directly with an assigned level of credibility. The degree of credibility will be determined based on the coherence between the finding and the corresponding illustration, classifying it as unequivocal, equivocal, and unsupported (Lockwood, 2015). The findings from the planned qualitative systematic review will be meta-aggregated and presented with a set of statements in a table (Aromataris, 2020).

The findings will be organised into categories based on the similarity of their meanings. To establish evidence-based practices, the categories will be merged to form a set of consolidated synthesised findings. The synthesis will include only unequivocal and credible findings (Aromataris, 2020). The TDF domains will be listed as categories in the JBI and each extracted finding (i.e., finding related to barrier or enabler or related term) will then be matched with these domains. Next, the meta-aggregation diagram will be exported, and the resultant domains will be mapped to the COM-B model.

Summary of findings and assessment of the certainty of the evidence

To build assurance in the results of synthesis, the final synthesised findings will be assessed using the confidence of synthesised qualitative findings (ConQual) score and presented as a summary of the findings (Munn, 2014). The ConQual score is defined as the rating of confidence in the synthesised qualitative findings (Munn, 2014). The scoring process involves an initial ranking of the studies from high to very low based on their type. In this instance, qualitative studies are given a high rank, whereas expert opinions are ranked very low. Then, the scoring is given to the five yes/no questions (three congruity questions, one question related to researcher culture, and one question related to researcher influence). These questions measure how dependable the finding is. If the scoring has four to five yes responses, the study rank does not change (Aromataris, 2020). Otherwise, the study rank is downgraded by one level for two to three yes responses and by two levels for zero to one yes responses (Aromataris, 2020).

This initial ranking is followed by credibility scoring, in which the synthesised findings are categorised as unequivocal (finding followed by an illustration that has an undoubted association with it and cannot be challenged), credible (finding followed by illustration, but its association with the finding is unclear and open to criticism), and unsupported (finding is not supported by the accompanying illustration) (Munn, 2014). If the finding is unequivocal, the initial ranking of the study remains unchanged. However, the study is downgraded by one level for a mix of unequivocal/credible findings. For credible findings, the study is downgraded by two levels, and for a mix of credible/unsupported findings, the study is downgraded by three levels. Finally, for unsupported findings, the study is downgraded by four levels. After the findings are examined by applying the dependability and credibility rules, the total score is called the ConQual score (Munn, 2014).

The essential components of the summary will comprise the title, population, context, and phenomena of interest. Each synthesised finding will be presented in conjunction with the research type that informs it, as well as the dependability, credibility, and ConQual score (see Supporting Information: Appendix 3).

The summary of findings of the intended systematic review will include a description of the process of article inclusion using a PRISMA flowchart. The flowchart will provide a summary of the number of articles identified, screened, selected for retrieval, and included/excluded along with their reasons for exclusion. It will also detail the numbers appraised and included/excluded, as well as the numbers included in the qualitative synthesis. In addition, the results of the methodological quality of each included article will be presented using the JBI SUMARI critical appraisal checklist.

Furthermore, the description of the included articles based on the objective of the systematic review, which includes setting, country, methodology, study participant, and the phenomenon of interest will also be presented in a table. The last part of the summary of findings of the planned systematic review will include a meta-aggregative overview flowchart. This flowchart will present four major components. The first component will consist of a list of a specified number of unequivocal and/or credible findings extracted from a specified number of the included studies, excluding unsupported findings as they will not be included in the meta-synthesis. The second component will include a list of categories, each of which comprises one or more unequivocal and/or credible finding(s) extracted from the included studies. The third component will include a list of synthesised findings consisting of one or more categories along with a summary statement. Finally, the fourth component will include a list of recommendations that will inform the practitioners and policymakers. The findings of the review will be used as a basis for generating recommendations by synthesising and interpreting the collective insights derived from the evidence synthesis. In summary, the findings will be meta-aggregated and presented in a table and diagram form.