This chapter describes the second work package (WP2) in the CAREDEM study, in which the previously developed case management programme (WP1) was tested in a feasibility trial. The primary objective for this pilot phase was to ensure that case management skills and the collaborative care model would be easy to acquire and implement in routine practice. The secondary objectives were to determine whether or not practices could recruit 11 patients into the study (depending on practice size), that nine could be contacted at 6 months and that stakeholders would find the brief intervention procedures acceptable and feasible within routine NHS practice. The researchers intended to check assumptions about practice and patient recruitment and retention and to ensure the feasibility (data yield and quality) of outcome measures; if necessary, the sample size calculation for WP3 would be adjusted based on these data. Practices that took part in the pilot study would not be able to participate in the main trial. Approval for WP2 was obtained from Wandsworth NRES (11/LO/1555).
Recruitment of practices and case managers
Practices were recruited from each site as detailed in the following sections. Data were obtained on practice population size, number of GPs and deprivation score. The average deprivation score in England is 21.5, with a higher value indicating greater deprivation.
London
In December 2011, the researchers wrote to 26 GP practices in Camden and Islington, inviting them to take part in the CAREDEM pilot study. Eight practices contacted the research team requesting more information; however, seven of these practices declined participation. Reasons given for declining participation were concerns about time, resources and current commitments, meaning that they were already very busy. One of these eight practices invited the researchers to present the proposed study at its practice meeting. After this meeting the practice agreed to participate in the study on the condition that the research team could guarantee Service Support Costs for a practice nurse, to backfill their time whilst working on CAREDEM.
Additionally, in January 2012 the researchers made contact with the North Central London Research Consortium (NoCLoR), which put them in touch with two further GP practices from NoCLoR’s local research clusters. One of these practices agreed to the researchers presenting the proposed study at its practice meeting in January 2012 and this practice subsequently agreed to take part in the research. At this point, as it was not possible to guarantee the Service Support funding, the practice that wished to take part unconditionally was recruited. During this process the researchers were further assisted by NoCLoR, whose staff negotiated with local commissioners on the study’s behalf to obtain Service Support funding for the CAREDEM study in London. This process facilitated engagement with the local research network and the researchers have retained the details of four GP practices who were interested in taking part in the main trial or in further research in this area but who could not commit to the pilot study. The list size of the recruited practice is 15,510 patients and the practice is served by 8.5 whole-time equivalent (WTE) GPs. The practice serves a population spread across two London boroughs, with deprivation scores of 21.5 and 27.
Norfolk
In Norfolk, 30 practices were contacted through the primary care research network and the DeNDRoN local research network and 12 expressed an interest in participating in the study. Practices were visited and the first to confirm involvement was recruited. The practice recruited covers a mainly rural setting with one large market town at the main practice and two satellite practices. The main practice was located adjacent to a community hospital. The practice had a list size of 14,400 patients and is served by 4.5 WTE GPs. Its deprivation score is 18.77.
Newcastle
Two GP practices were approached by the site lead and agreed to participate in the study. Adult services in one local authority were approached to assist with recruitment of the case manager and offered a seconded role of a full-time case manager to provide the intervention across the two GP practices. Funding for the role was brokered by the locality PCT. A social worker with considerable experience of working with people with dementia and their families was initially recruited but accepted another post at the beginning of the recruitment screening. A second experienced social worker with knowledge of dementia took over the role. Recipients of case management were unaffected by the change in personnel. The social worker was primarily located at the larger practice but also spent a significant proportion of time in adult services, which afforded access to social work systems.
Patients were recruited across the two sites. One was a large, two-centred practice with a central and a satellite practice covering the city centre plus a broad radius of suburbs; the other was a smaller practice covering one area of the city. The larger practice had a list size of 28,396 patients served by 15 WTE GPs, and the practice’s deprivation score was 27.8 (putting this practice in the fourth most deprived centile). The smaller practice had a list size of 6501 patients served by 4.25 WTE GPs, and its deprivation score was 29.7. Recruitment was more successful in the larger practice.
Recruitment of patient and carer dyads
Inclusion and exclusion criteria
The target patient population for the CAREDEM WP2 study was people of any age with a diagnosis of any type of dementia (confirmed by secondary care assessment). Potential participants had to be living in the community at the point of recruitment and to have a carer (spouse, close relative, friend or other informal carer) who maintained regular contact. Those resident in care homes or being seen regularly by specialist dementia services were deemed to be being case managed already and were therefore excluded. Those unable to read English language information sheets about the study were also excluded.
Delays in achieving approval from local PCT research and development departments meant that recruitment of patients could not begin before July 2012. Practices recruited people with dementia and their carers over a 6-month period between July and December 2012 and followed them up for 5 months, the last interview being on 31 May 2013.
The WP2 protocol stated that patients would be invited to take part by their GP; however, in the London and Norfolk practices this was not realistic and eligible patients were screened and invited to take part by the practice nurse working in the case manager role, with the agreement of their GP. In the Newcastle practices the invitations were offered by the patients’ GPs and followed up by the research team.
The case manager or GP sent the patient and carer an information sheet, an opt-in form and a prepaid envelope and followed this up with a telephone call. Once the opt-in form or verbal consent form had been received, the research team was informed and contacted potential participants to answer any questions that they may have had. If the patient and carer were happy to proceed, the researcher arranged a visit at home or in another place specified by them to obtain informed consent and collect the baseline data. At the end of the baseline interview the researcher explained the next stages of the process and agreed with the participants that they would contact them for a follow-up appointment.
Directly after the baseline appointment the researcher informed the local case manager that the baseline assessment had been completed so that they could commence the intervention and completed a reflection log on the informed consent form process. Mentoring was provided by an Admiral nurse seconded to the project from Dementia UK, who visited the practices, carried out joint assessments with the case managers, communicated by telephone and e-mail with case managers on a regular basis and used the educational needs assessment and task matrix as the framework for discussing the case management role. Patients were followed up at 5 months (instead of 6 months as specified in the protocol) to allow completion of the study.
Quantitative data collection methods
Following the process of informed consent, interviews were conducted with the carer and the person with dementia individually, unless either person preferred the interview to take place with their relative in the room.
Demographic details, including date of birth, gender, marital status, level of education, employment status, ethnicity and relationship to care recipient, were obtained from participants at the baseline interview.
The carer completed the following assessments:
The Neuropsychiatric Inventory (NPI)
59 (on which the current sample size calculations are based). This is a validated instrument with 12 domains, completed in an interview with a carer to assess the prevalence of behavioural and psychological symptoms experienced by the person with dementia he or she is supporting.
The Bristol Activities of Daily Living Scale (BADLS)
60 was used to assess the functional impairment of the person with dementia.
The 28-item General Health Questionnaire (GHQ-28)
61 was used to provide a measure of the carers’ mood and quality of life.
The European Quality of Life-5 Dimensions (EQ-5D) (see
www.euroqol.org/about-eq-5d.html; accessed 23 June 2014) was used to assess carer quality of life and to generate quality-adjusted life-years (QALYs).
The Client Service Receipt Inventory (CSRI)
62 captures service utilisation data for the carer and the patient (including institutionalisation, extra patient care during therapy), unpaid carer support and other aspects relevant to health economics. The rates and dates of entry into institutional care are recorded.
The person with dementia was assessed in the following ways:
The Mini Mental State Examination (MMSE)
63 was used as a measure of cognitive function.
The Dementia Quality of Life (DEMQOL) scale
33 was completed if the patient scored ≥ 11 on the MMSE, in line with guidance. The DEMQOL scale is a generic measure from which it is possible to generate QALYs.
Discontinued measures
The decision was taken before recruitment of participants to remove two of the measures included in the original protocol. The scale measuring quality of life in Alzheimer’s disease (QoL-AD)64 was removed as we were already measuring quality of life with the EQ-5D and the DEMQOL scale; the researchers felt that the extra measure was excessive and therefore unnecessary. Similarly, the 12-item Health Status Questionnaire (HSQ-12)65 was removed and the GHQ-28 was retained.
Adverse events
The numbers and details of adverse events (e.g. emergency admission to hospital) or serious adverse events (e.g. deaths) were recorded.
Qualitative data collection
In-depth qualitative interviews were conducted with members of a range of stakeholder groups to explore different perspectives on case management. Those interviewed included people with dementia (patients), carers, case managers, the case manager mentor (hereafter ‘mentor’), health and social care professionals and members of the research team. Separate interview topic guides were developed for each stakeholder group to reflect different levels of engagement with case management. During data collection, interview topic guides for all stakeholder groups were adapted through an iterative process in light of low levels of intervention delivered in practice and emerging themes.
Interviews were conducted by members of the research team at various stages of the feasibility study to capture processes and experiences of case management at different time points and to capture key events such as case manager training. The majority of interviews were conducted face to face with individuals; however, when this was not practically feasible, a small number of interviews were conducted with two or three participants in a group or by telephone.
Patient and carer interviews
A purposive sample of patients and carers was invited to participate in a single in-depth qualitative interview to explore their experiences and views of case management (). Some interviews with patients and carers were carried out jointly at the request of the participants; however, when possible, interviews were conducted separately to enable exploration of potentially differing carer and patient experiences.
Patient and carer interviews by area
Case manager and mentor interviews
In total, there were four case managers throughout the pilot study. The views of the mentor were also sought through interview. Case managers and the mentor participated in several interviews throughout the study to explore changing expectations, the development of the role, training and supervision, the implementation of case management in practice and their views on the value of the case management approach. In total, 13 interviews were conducted (four with the mentor, two with each of three case managers and three with one case manager).
Research team interviews
Two members of the research team were interviewed and informal discussions were conducted with a further member of the team. These interviews/discussions explored the induction process, expectations of case management, barriers to implementation and their views on the value of the case management approach.
Health and social care professional interviews
In total, 18 in-depth interviews were conducted with a range of health and social care professionals (). Some practitioners reported direct interactions with the case managers whereas others gave a more theoretical perspective on case management (e.g. one of the commissioners). These are reported in Chapter 7.
Role of participating health and social care professionals (n = 18)
Qualitative data management and analysis
All interviews were audio recorded, transcribed, checked and anonymised. In the initial stage of analysis individual team members (CB, KB, MP and LR) read and reread a number of transcripts to familiarise themselves with the data and identify preliminary themes. A series of data workshops were then held in which team members discussed these preliminary themes and developed a draft coding frame. This was then applied to a small number of transcripts and the findings were discussed in subsequent data workshops. Following a series of iterations, and informed by the theoretical domains framework,66 a final coding frame was agreed. All transcripts were coded in NVivo (version 9; QSR International, Warrington, UK) to facilitate data management using the final coding frame. In the next stage of analysis, the output for different stakeholder groups was reviewed. This led to the combination of some codes, the identification of new subcodes and the production of a narrative summarising the key themes for each group. These narratives were then compared and overarching themes identified across the stakeholder groups.
Confidentiality
Although we recruited both male and female case managers, to avoid identification we have referred to all case managers as female and we have not attributed quotes to individual case managers. By virtue of her role, the mentor was identifiable and she has reviewed and agreed to the use of all quotes attributed to her. Quotes from patient and carer interviews are identified by site (A, B or C) and unique identifier within each site.
The findings from WP2 are presented in the following four chapters:
Chapter 4 – recruitment to the study of practices and patient–carer dyads
Chapter 5 – implementation of the study including recruitment processes, acceptability and feasibility of outcome measures, patient and carer views on study participation and case manager and mentor views on study procedures
Chapter 6 – capturing what the case managers did during case management
Chapter 7 – stakeholder perceptions of case management for people with dementia.
