Reducing refusals of care through improved personal care interactions between caregivers and people with dementia: protocol for a realist synthesis

Assessing video-recorded personal care interactions and interview transcripts

To ensure that caregivers and people with dementia’s experiences about refusals of care and personal care feed into our initial programme theories, we will conduct a secondary data analysis of interviews and observations conducted for the Pro-CARE Study. This study was a fellowship for the first author funded by the Alzheimer’s Society in England. The programme of research examined refusals of care and personal care interactions between caregivers and people with advanced dementia. Interviews were conducted with 20 family carers, 12 care-home staff,28 17 home-care workers29 and 13 people with dementia receiving assistance with their personal care. Interviews focused on experiences of personal care assistance and refusals. Observations consisted of 26 video-recorded personal care interactions between 14 family carer or care-home staff/people with dementia dyads and covered activities where the person with dementia was clothed such as teeth cleaning, hair washing and shaving (3 hours of footage in total).7

The first author will analyse these datasets with a realist lens examining a range of complexities that potentially interact. The analysis will seek to explain the factors involved in actions for care, reactions to care interactions and potentially helpful strategies for working with refusals of care. Data will be categorised into tentative programme theories and possible underlying causal mechanisms of actions will be developed about how approaches worked and if possible, what needed to be in place to support their use. Comparison within and between participant categories and observation or interview-derived data will enable identification of cross-cutting causal insights. Causal insights will be taken forward and evidence of data retained to enable an audit trail of theoretical development and decision-making. Team discussions will assist in theory building, interrogate initial ideas and add further insight to initial theories.

Preliminary review

An initial exploratory review of key literature will take place27 aiming to identify the breadth of interventions and intervention components to feed into the initial programme theories with the aim to find cross-cutting mechanisms. Targeted searching will include key articles, for example, those included in a recent systematic review on strategies to reduce or manage refusals of care.14 In addition, articles and interventions known by the study team involving caregiver interactions with people with dementia such as namaste care,30 The Voice Study dementia communication skills training31 and Humanitude Care32 will be assessed.

Team discussions

The study team has expertise in occupational therapy, personal care assistance, dementia care research, and care home, family carer, and hospital care settings. Using theoretical insights gleaned from the secondary data analysis and preliminary review, we will develop initial IF, THEN, BECAUSE statements. Project team theory building will involve considering insights from data, initial literature and their own expertise to generate causal statements, which can be developed into context-mechanism-outcome configurations at the end of stage 2. From these initial programme theories, the team will agree on the scope and focus of the realist synthesis, if necessary, prioritising initial programme theories to take forward based on their relevance, plausibility and preliminary evidence levels. The team will design a theoretically based framework to be populated with evidence.27 Existing mid-range theories will be sought through targeted searches led both from the existing knowledge of the research team and from the developing programme theories. These will be examined and, if relevant, incorporated to guide our theory building.

Searches

The search strategies have been developed by the study team (see online supplemental file), aiming to find literature that will generate, evidence and develop programme theories to, if possible, specify context, mechanism and outcome configurations.33 34 Further searches will be conducted iteratively to explore individual programme theories as necessary as the synthesis progresses.

The searches have three strands: dementia, refusals of care and personal care/ADLs. We will purposively search through the following databases: MEDLINE, EMBASE, PsycINFO, CINAHL, CENTRAL—Cochrane Central Register of Controlled Trials and the Social Sciences Citation Index. Searches will be adapted depending on the database used. We will search grey literature sources such as practice-focused publications and organisational websites such as the Alzheimer’s Society. The searches will be restricted to the English language and from 2000 onwards to reflect the time when Kitwood’s person-centred care ethos35 36 started to influence dementia care. EndNote reference manager will be used to collate citations and identify duplicates, from our searches.

Selection

Selection will be carried out by one reviewer with independent screening of a subset of documents by a second reviewer to support quality assurance. Selection will be based on each publication’s relevance and contribution to theory building. Included studies will focus on refusals of personal care or personal care interactions with people living with dementia, be related to any of four care settings (care home, family home, hospital, home care), include an intervention or examine a strategy to reduce refusals of care or improve personal care for people with dementia and be in the English language. Publications not focused on personal care interactions between people with dementia and caregivers or those focused on end-of-life care will be excluded. To maximise potential understanding any type of peer-reviewed study will be eligible to be included in the review. Therefore, no restrictions will be placed on the study design.

We are interested in whether interventions or components of interventions employed to reduce or manage refusals of personal care in dementia are successful or not, how they may work, for whom and under what circumstances. Depending on retrieved literature and time resources, we will use a quota system, initially assessing 10 studies for each setting: hospital, care home, home care and family, prioritising those with the most relevance to theory building. If limited literature is found, we will examine theoretically relatable practices such as interactions with people with dementia for other purposes, and/or personal care interactions in other groups, for example, learning disability or other neurological or psychiatric conditions. Contextual factors may include the value ascribed to personal care, suitability of the environment, espoused ethos of care, time available, caregiver skills, and symptoms and stage of dementia.

Quality appraisal

Quality appraisals will focus on assessments of relevance, richness and rigour,27 37 not appraising for methodological quality but assessing all parts of papers including the introduction and discussion for causal thinking and insights.38 Relevance will be judged on whether or not the resource contains data applicable to the topic area or programme theories, and richness on whether the resource can contribute sufficiently to theory building or testing.37 39 Rigour will be judged on the extent to which the method generating the insight is credible and trustworthy39 and whether the theory is coherent.37 Where there are difficulties interpreting quality or uncertainties surrounding judgements, articles will be discussed between two reviewers, three if no consensus is reached. Depending on whether other evidence is available for that programme theory, resources with low relevance and richness will be excluded. Where concepts such as refusals of care are not portrayed accurately or the meaning authors are ascribing mismatches our conceptualisation (such as those centred on non-adherence or refusal of medical procedures), we will exclude articles. However, articles offering alternative explanations within our conceptualisation will be included and used to challenge biases and team interpretations.

Extraction

A bespoke form aimed at populating the theoretically based framework with evidence linking to initial IF/THEN/BECAUSE and context-mechanism-outcome configurations for testing initial programme theories will be developed. Data extracted are likely to include:

• Study information: authors, title, journal, country of origin, publication date.

• Setting and participants: care setting, sample size and composition including stage of dementia and caregiver type.

• Methods: design, analysis

• Context description: contextual factors including refusal information.

• Intervention/interaction description: inputs, components, processes, mechanisms resource, mechanisms reasons/reactions, outcomes.

• Evidence relating to initial programme theories.

The form will be piloted and refined to make sure it supports detailed information about studies, interventions and intervention components. Data will be extracted from included studies and grey literature by one reviewer (TB) with independent screening of a subset of documents and extraction by a second reviewer to support quality assurance. Data extraction will initially be kept separate for each care context.

Synthesis

Synthesis will be exploratory, revising the initial programme theories considering the data to develop evidence-based context-mechanism-outcome configurations. The aim will be to evidence theories related to strategies and mechanisms of interventions between caregivers and people with dementia that contribute to reducing refusals of care and determine how they work in which contexts, why and for whom. One researcher (TB) will read extracted data and relevant texts again, focusing on the theoretical underpinnings of the programme theories considering meanings, theories and rival theories, making notes, establishing connections between concepts. Consideration will be given to both intended and unintended causal insights. Prioritised initial programme theories will be populated with relevant extracted data generally and if possible, under context-mechanism-outcome configurations, including mechanism resources from the intervention and how this changes the reasoning of actors.40 Synthesis will occur through ongoing examination of each of the theories and relevant data considering causal insights. Team meetings will examine and interrogate developing theories and context-mechanism-outcome configurations alongside the evidence. If specific evidence is available and only relevant for similar types of interventions or strategies and/or care settings synthesis will occur for subcategories. For example, evidence relevant only to certain settings such as hospitals will be kept separate in our synthesis from programme theories relevant for all settings.