Methods

Claire Goodman; Christine Norton; Marina Buswell; Bridget Russell; Danielle Harari; Rowan Harwood; Brenda Roe; Jo Rycroft-Malone; Vari M Drennan; Mandy Fader; Michelle Maden; Karen Cummings; Frances Bunn

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Goodman C, Norton C, Buswell M, et al. Managing Faecal INcontinence in people with advanced dementia resident in Care Homes (FINCH) study: a realist synthesis of the evidence. Southampton (UK): NIHR Journals Library; 2017 Aug. (Health Technology Assessment, No. 21.42.)

Managing Faecal INcontinence in people with advanced dementia resident in Care Homes (FINCH) study: a realist synthesis of the evidence.

Show details

Contents

< Prev Next >

Chapter 2Methods

Rationale for using a realist review

The rationale for using a realist synthesis approach is twofold: (1) the absence of evidence of what is effective in the reduction and management of FI in care homes and (2) a recognition that future interventions designed to address FI in PLWD will need to be multicomponent, depending on the behaviours and choices of those delivering and receiving the care. Realist review is a theory-driven interpretive approach to evidence synthesis.⁶⁰^–⁶² It assumes that there is more to reality than the way in which it is socially constructed. There is an external reality or world that can be observed and measured; however, how this reality is articulated and responded to is constantly being shaped by individuals’ perceptions and reasoning and/or dominant social and cultural mores. It is this constant interaction that creates particular responses that lead to observed outcomes.⁶³ Realist synthesis, therefore, endeavours to go beyond lists of barriers to and enablers of care to unpack the ‘black box’ of how interventions to reduce and manage FI work. The often repeated statement used to explain the focus and purpose of realist synthesis is that it makes explicit ‘what works, for whom, in what circumstances?’. It uses a theory-driven approach to articulate how particular contexts (C) or resources, have prompted certain mechanisms (M) or responses by those providing and receiving care to lead to the observed outcomes (O). The underlying premise is that the observed ‘demi-regular patterns’ of interactions between the components that make up complex interventions in the evidence reviewed can be explained through theoretical propositions. The iterative process of the review tests those theories that are thought to work against the observations reported in the evidence included in the syntheses.⁶⁴ This enables us to take account of a broad evidence base as well as the experiential and clinical knowledge that relates to the physiology and management of FI in older people, and specifically in older people with advanced dementia living in long-term care. It will also help us to take account of the heterogeneity of care home provision in the UK (Box 1).

BOX 1

Definitions of realist terms and how they have been applied throughout the review

A realist review does not follow a neat linear process, although it does have distinct phases summarised in Figure 1. This process was specified in the review protocol.⁶⁵

FIGURE 1

Four-phased approach to the synthesis.

It is an assumption that mechanisms will only activate in the right conditions, providing a context + mechanism = outcome formula. Explicit disaggregation of resources as a component of a context helps to be more explicit about the aspects of an intervention that trigger different responses. We have characterised this as context (intervention)–mechanism–outcome (C–M–O) and used this approach in our analysis of the evidence.

Changes from the submitted protocol in the review process

A realist review is an iterative process, and adjustments are made to the review protocol in the light of emerging or new lines of enquiry. Changes to the protocol submitted to the funder are outlined in Table 1.

TABLE 1

Changes to the protocol that was submitted to the funder

Phase 1: defining the scope of the realist synthesis – concept mining and theory development

Concept and theory development by the team

A realist synthesis often uses input from content experts to help develop programme theories. In this project our Research Management Team (RMT) and Study Steering Committee (SSC) included many of the national experts in the field of continence care and care home research in the UK (see Appendix 1). The first RMT meeting in September 2014 included an open discussion by the team in which they were asked to draw on their expertise to articulate:

the dominant approaches and assumptions that informed current thinking about what supported (and how) the reduction and management of FI
important outcomes and how impact was measured.

The full notes from this meeting are included in Appendix 2. After the meeting all RMT members were asked to send relevant reviews, papers or guidelines that had been referred to at the meeting to the core team. The early progress was reported to the SSC in November 2014.

Scoping interviews

To complement the expertise provided by the team we interviewed key stakeholders. They were identified through the clinical and research networks of the RMT, and five key groups were identified and approached to inform the scoping process (Table 2).

TABLE 2

Rationale for selection of stakeholder groups

Interviews

Interviews were conducted either one to one or in a group by Claire Goodman, Marina Buswell and Bridget Russell. All interviews were face to face. Interviews were recorded, when possible, and transcribed. A set of interview prompts was developed and tailored to reflect the focus and expertise of each group. We wanted to explore:

participants’ experience of working with and for people with dementia and FI in care homes
estimates of how severe FI is in care home residents
why they think FI occurs
what good continence care looks like and what is necessary to achieve that and why
who needs to be involved in providing and reviewing care
how having dementia specifically affects continence care, assessment and decision-making, and what needs to be in place to address the effects of dementia
how they would define success from the perspective of the resident, the care home staff, visiting clinicians and family carers
when and why the decision would be made to use pads.

All participants highlighted the importance of person-centred care. The interviewers sought to understand how this approach to care was articulated in the assessment, care and review of continence care provided in care homes. In many of the interviews the prompts led to narratives of success and failure with particular residents; these examples demonstrated the participant’s understanding of what needed to be done and what good care looks like.

Continence specialists

Interviews were completed at the continence specialists’ place of work. We approached three continence specialist teams but were able to access only one team and interviewed three nurses who visited care homes. This was supplemented with a group discussion with members of the British Geriatrics Society (BGS) continence special interest group, which included physicians and nurses.

Interviews with care home managers and care home staff

We recruited care home managers through two care home provider organisations, one not for profit and one commercial chain. Meetings were organised by the two care home organisations on our behalf and held on their premises.

Care home staff involved in providing direct care to residents were recruited through contacts with local care homes (which were not part of the care home chain that managers were recruited from) and via the networks of the RMT. We recruited seven care home staff, one of whom was a senior care worker. Two of the participants worked nightshifts in separate care homes and were interviewed face to face. The remaining five worked in the same care home and were able to participate in a focus group discussion. Questions focused on their experiences of providing care, how they defined good care, what they thought about how a person’s dementia affected their ability to provide continence care and what was important to enable them to achieve that.

Interviews with family members of residents with dementia and experienced faecal incontinence

In phase 1, we attempted to recruit family members who had experience of supporting a relative with FI and dementia through relevant charities [a local branch of the Alzheimer’s Society and the Bladder and Bowel Foundation (BBF)]. Despite several attempts, these approaches were unsuccessful. An invitation through a local care home provider to invite participants yielded three expressions of interest; two of these people participated in a group interview. An analysis of the phase 1 interviews was undertaken thematically using the prompts as the organising framework, and the findings from this were fed into the RMT deliberations and workshop.

In phase 3, to identify people with personal experience of FI and dementia through caring, we renewed our request for help to the BBF. This time the BBF supported a mail-out to their members that elicited responses from interested relatives (see Appendix 3).

To ensure transparency of approach and an audit trail, focus group discussions and interviews were recorded and transcribed. Structured field notes were kept about which sources of evidence and experiential knowledge may be linked to which strands of theoretical development.

Literature scoping: phase 1

Search methods

We conducted four separate searches. The initial search built on literature identified as relevant at the first RMT meeting. This focused on evidence on continence-related research in care homes, dementia and continence, older people and continence, implementation research in care homes and person-centred care. A second search was then conducted that was designed to identify papers on FI and care homes and incontinence pads (see Appendix 4).

The scoping interviews with geriatricians at the BGS highlighted the link between diet and hydration and good bowel health, and a further search was conducted to capture literature on interventions to promote nutrition and hydration (eating and drinking) for PLWD in care homes. This search was conducted to test whether or not this body of work included outcomes related to continence and FI. After the second SSC meeting in November 2014, a decision was made to scope the learning disability literature for continence-related research.

Selection and appraisal of documents: phase 1

Frances Bunn, Claire Goodman and Bridget Russell carried out the first searches and selected studies for initial inclusion based on the broad themes outlined by the RMT. A Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram was produced, although as the review progressed it did not capture the iterative process of searching, selecting and appraisal of documents for this process (ultimately replaced by a flow diagram; Figure 2). A summary of this scoping search was presented to the RMT in January 2015.

FIGURE 2

Document flow and review of progress: conceptual diagram. CH, care home; PCC, patient-centred care.

Data extraction: phase 1

A narrative process rather than a formal data extraction was used to produce the first scoping summary (see Appendix 2). In phase 1, however, we began to develop a data extraction process that captured our thinking about the putative theories and context–mechanism–outcomes (C–M–Os) of interest (see Phase 1: defining the scope of the realist synthesis – concept mining and theory development, and Prototype data extraction form).

Analysis and synthesis processes: phase 1 (concept mining and theory development)

Meetings

Team meetings formed an essential and productive part of the analysis and synthesis process. Formal meetings of the SSC and RMT were recorded and detailed meeting notes were made. Informal team meetings of the smaller core team (FB, MB, CG and BRu) were not recorded but notes and actions of these meetings were kept in log books (see Log books).

Log books

Log books were used as project notebooks to record when and how important decisions, thinking, sourcing of papers or documents and actions occurred. They provided a useful source of ‘raw’ data and reference points for analysis, and included notes about key papers that informed a shift in thinking and synthesis. These became the structured field notes on suggestions and decision-making processes about which sources of evidence were linked to which strands of theoretical development.

‘If . . . then’ statements

At the end of the RMT meeting in January 2015, a series of ‘if . . . then’ statements were tabled that drew on the interview and scoping data. ‘If . . . then’ statements are the identification of an intervention/activity linked to outcome(s) and contain reference to contexts and mechanisms (although these may not be very explicit at this stage) and/or barriers and enablers (which can be both mechanism and context).⁶⁶ Although we did not use logic models, the ‘if . . . then’ statement route of thinking logically and working back from outcomes provided a helpful way of structuring our thinking. It also helped to focus the process of taking ideas and assumptions about how interventions work and testing them against the evidence we found (see Appendix 5).

Prototype data extraction form

Early data extraction was carried out in parallel with development of ‘if . . . then’ statements and developing a framework using diagrams and mind mapping. At this point, the data extraction process was more a matter of concept mining. We developed a table in Microsoft Excel^® 2013 (Microsoft Corporation, Redmond, WA, USA) and used headings that drew on findings from the early scoping and discussions within the team. This form was piloted on five initial papers recognised as important in FI and care home research and then refined further.

Diagrams and concept mapping

The diagrams and concept mapping captured a summary of evidence and thinking, overarching themes, likely barriers, enablers, contexts and outcomes. This was the basis for conversations about possible interactions and the identification of areas for which there was consensus about what ‘good FI care’ may look like for PLWD in care homes. We also used flip charts to map possible C–M–O configurations and these helped in the development of the data extraction form.

Narratives

Narratives were an important tool for making thought processes explicit and facilitating theory development. We used two main formats for narratives.

In the first, we created detailed narratives around specific research areas, for example a historical narrative of prompted voiding (PV) studies and care home research, or a summary of guidelines and reviews relevant to continence care in older people. While we were ‘data extracting’ individual sources, our core team discussions started to cluster around themes and particular areas and programmes of research.

The second format was ‘working paper’ narratives, which were presented to the wider team to ensure that everyone could keep up to date with methods, results, idea flows and theory development between meetings. In a realist review it is not possible for the process to be completely transparent. However, capturing our narrative thinking in this way should help others to understand and critically evaluate how we have come to our conclusions. It also meant that the process for developing theory could be understood, and challenged, by the research team, the SSC and our stakeholders. Pearson et al.⁶⁶^,⁶⁷ have commented on how little information is available about how the realist review process is enacted; these working papers helped to make explicit our thinking and decision-making.

Phase 2: retrieval, review and synthesis

Searching processes: phase 2

In March 2015 the scoping searches conducted in phase 1 were refined and expanded to include additional evidence sources (see Appendix 4). The RMT agreed that literature regarding facilities other than care homes was unlikely to add to what had already been obtained; however, relevant UI literature and care home studies with transferable methods and similar aims were to be included.

We included studies of any design including RCTs, controlled studies, effectiveness studies, uncontrolled studies, interrupted time-series studies, cost-effectiveness studies, process evaluations, surveys and qualitative studies of participants’ views and experiences of interventions. We also included unpublished and grey literature, such as policy documents and information about locally implemented continence programmes or proceedings, which could provide a model for future practice or merit future evaluation.

The search was limited to papers published between 1990 and 2014, although we did include seminal papers from earlier, such as the work of Tobin and Brocklehurst,⁶⁸ and those identified from lateral searches. There were several reasons for this time limit. Health-care research in care homes is a relatively recent phenomenon. Gordon et al.⁶⁹ identified that of 292 RCTs conducted in care homes between 1974 and 2009, half had been published after 2003. Dementia research has been significantly influenced by the work of Kitwood,⁷⁰ whose seminal work was first published in 1990. Furthermore, the organisation and funding of care homes was radically altered in 1993 by the implementation of the 1990 National Health Service and Community Care Act.⁷¹ This led to progressive changes in the overall size, ownership and structure of the sector. The increased emphasis on domiciliary care has also meant that the level of dependency and frailty of older people now admitted to long-term care has increased.⁷²

Previous dementia reviews undertaken by members of the project team⁷³^,⁷⁴ have highlighted the importance of lateral searching for identifying studies for dementia-related reviews. Therefore, in addition to the above electronic database searches, we undertook the following lateral searches:

checking of reference lists from primary studies and relevant systematic reviews (snowballing)⁷⁵
citation searches using the ‘cited by’ option on Web of Science, Google Scholar and Scopus and the ‘related articles’ option on PubMed and Web of Science (‘lateral searching’)
contact with experts and those with an interest in dementia, care homes and FI to uncover grey literature (e.g. National Library for Health Later Life Specialist Library, Alzheimer’s Society, Royal College of Physicians, Royal College of Nursing and The Queen’s Nursing Institute)
contact with disease-specific charities and user groups, residents and relatives associations
internet searches for grey literature and searches for continence-related evaluations or intervention research that makes specific reference to FI and/or people with a dementia diagnosis, national inspection and regulation quality reports by the UK regulator (Care Quality Commission and predecessors).

Four overlapping searches were completed that built on and expanded the search completed in phase 1 (see Appendix 4).

Searches 1a and 1b searched for evidence on care home research, either continence or FI related, that included PLWD, and care home research covering implementation or patient-centred care (PCC) that included people with dementia.
Search 2 focused on continence literature in care homes that may be about factors associated with FI, such as the use of incontinence pads or constipation.
Search 3 focused on research in care homes for people with dementia that concerned nutrition and/or hydration in the care home population. We were interested both in whether or not this research mentioned outcomes relevant to FI or continence generally and in discussion about how eating and drinking interventions may be implemented.
Search 4 focused on literature on continence care for people with learning disabilities living in residential care.

As the review progressed, and particularly the idea of needing to fit into care home routines, wider literature that discussed the impact of care home routines and working practices was consulted by team members (see note on bias in Strengths and limitations of the review). Further papers, professional journals and books were identified through lateral searches, alerts from professional networks, recommendations from the SSC and conference abstracts.

Theory development: phase 2

In parallel with the second phase of literature searching just described, mid-range theories were developed. This process is represented in Figure 2 and the results of theory development are described in the results. Six mid-range theories or ‘theory areas’ were developed and assessed against the selected literature using the data extraction process described in Data extraction: phase 2.

Selection and appraisal of documents: phase 2

Management of references

Search results from electronic databases were imported into EndNote bibliographic reference management software (Thomson Reuters, CA, USA) and, when possible, duplicates were deleted. Documents from other sources were manually recorded in the same EndNote library. At least two reviewers (BRu, MB, CG and FB) independently screened titles and abstracts to identify potentially relevant documents to be retrieved and assessed. Disagreements were resolved by discussion. To enable us to keep track of the large body of literature and the changing ‘inclusion’ criteria as the mid-range theories emerged, decisions made at different points in time were recorded in the EndNote record for each paper.

Although literature was considered for inclusion in terms of its relevance to the UK care home sector, the nature of a realist review meant a linear included/excluded paradigm was not wholly appropriate. Instead the team regularly revisited the literature to assess whether or not, in the light of evolving theory development, previously ‘excluded’ sources may in fact be relevant.

Data extraction: phase 2

The second phase of data extraction took place from July 2015 to October 2015.

Development of data analysis and synthesis form

We developed a prototype data analysis and synthesis form (DASF) tailored to reflect the phase 1 findings; this reflected the six mid-range theories which were to be tested. All RMT members completed data extraction on at least two of three papers, with all papers and other sources of evidence (e.g. books and professional papers) being reviewed by at least two members of the team. Most of the reviewing was completed by the core team (MB, BRu, CG and FB).

It should be noted that ‘data’ in a realist sense are not just restricted to the study results or outcomes measured. Author explanations and discussions can provide a rich source of ‘data’, and so these were included in the DASF. Box 2 summarises the six sections of the DASF.

BOX 2

Data analysis and synthesis form sections

Three exemplar DASFs are given in Appendix 6.

Synthesis

The process of synthesising the data from the individual DASFs included:

the organisation of extracted information into evidence tables representing the different groups of literature (i.e. care home interventions, FI and UI and person-centred care)
theming across the evidence tables in relation to the C–M–Os and emerging patterns of association seeking confirming and disconfirming evidence
linking these observed demi-regularities (patterns) to develop hypotheses.

Integral to this process were cross-team discussions about the integrity of each theory and the competing accounts of why interventions did or did not work. This included a consideration of the unintended consequences of some approaches to continence care: what appeared to be true regardless of setting (e.g. by national care system or presence or absence of clinical staff), what was a setting-specific outcome, what were intermediate outcomes (e.g. changes in staff behaviour and knowledge) and what was a FI-specific outcome (e.g. reduced episodes of FI).

The findings were discussed by the core team and with some members of the wider team. In addition, a full draft was circulated to the team for detailed comment. Time constraints meant that it was not possible to hold a second workshop with the RMT as originally planned. Moreover, the team considered that at this stage it was more important to focus on presenting and discussing the findings and resultant hypotheses with the stakeholder groups. Initially, we had proposed that these would be presented to the stakeholders as a series of C–M–O links and the characteristics of the evidence underpinning them. The nature of the evidence (and lack thereof) ultimately led the team to focus on the development of programme theory that posited the likely C–M–O configurations that would lead to improved continence care.

Phases 3 and 4: test and refine programme theories (validation) and develop actionable recommendations and evidence-informed framework

To enhance the trustworthiness of the resultant hypotheses and to develop a final review narrative to address what is necessary for the effective implementation of programmes to manage FI in care homes, we reviewed the hypotheses and supporting evidence through consultation with the SSC and interviews with stakeholder groups, some of whom had participated in the scoping. Findings were presented to:

the continence specialist interest group at the BGS who had participated in the phase 1 interviews; approximately 60 delegates attended the conference on 6 November 2015 – 11 provided active feedback and comment during the discussion and six did so during the lunch break, and comments from four other speakers provided support for aspects of the emergent findings;
members of the BBF who had personal experience through their caring roles of living with FI and dementia;
a meeting of the National Care home Research and Development Forum (1 January 2015);
organised meetings in two separate care homes with staff involved in providing direct care from one of the care home organisations that had participated in phase 1 (9 December 2015).

Discussion focused on stakeholders’ views on the resonance and significance of the programme theory and the suggested C–M–O threads from both a practice and a service user perspective. In addition, there was an invited presentation at an Evidence Synthesis Network [hosted by National Institute for Health and Care Excellence (NICE) and the University of Manchester] workshop on realist methodology held on 17 November 2015. The presentation using the Faecal INcontinence in people with advanced dementia resident in Care Homes (FINCH) study and its findings was an exemplar of how the programme theory was developed and tested, focusing on how we differentiated between contexts and outcomes in the evidence reviewed.

Ongoing dissemination work on the FINCH study was planned through seminars and a plan for publications and briefings in association with the care home manager member of the RMT and the SSC.

Public and patient involvement

The proposal was circulated and discussed with members of the public and patient involvement group at the University of Hertfordshire and amended accordingly. In addition to public and patient involvement representation on the SSC (Paul Millac), engagement with user and patient representative groups was key to stages 1 and 3 of the synthesis. There was public and patient involvement in the National Care Home Research and Development Forum meeting, during which emergent findings were discussed.

It was not possible to involve residents with dementia and experience of FI in the study. Karen Cummings and Victoria Elliot of the Orders of St John Care Trust care home group provided input to the RMT and SSC as individuals with direct experience of providing and managing continence care in care homes.

Copyright © Queen’s Printer and Controller of HMSO 2017. This work was produced by Goodman et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Included under terms of UK Non-commercial Government License.

Bookshelf ID: NBK447301

Contents