Some material within this chapter has been reproduced with permission from Walton et al.1,47 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
The protocol was informed by and developed with input from the evaluation advisory group, other national and local stakeholders and RSET patient and public involvement and engagement (PPIE) group.
Setting
This research took place in England and Wales.
Design
This multisite study combined qualitative and quantitative approaches to explore peer support initiatives for social care in adult prisons in England and Wales. These methods are consistent with previous research in prisons which have used a range of methodologies, including observation, documentary research, self-report questionnaires, interviews and multiple methods.48
The design and research questions outlined in this study were informed by scoping meetings that took place prior to the study starting. These meetings included a range of stakeholders: individuals working in the prison service; in peer support initiatives; in social care organisations; and academics undertaking related work.
Ethical approval
This study has been reviewed and given favourable opinion by London – South East Research Ethics Committee (REC reference: 22/LO/0592), and approval from the National Research Committee (NRC reference: 2022-224).
Procedure
The evaluation comprised five workstreams outlined below (see Chapter 1 for research questions for each workstream):
- Review of prison peer support services (workstream 1).
- Documentary analysis of adult social care in prisons (workstream 2).
- Implementation and experience study (workstream 3).
- Outcomes study (workstream 4).
- Cost study (workstream 5).
Due to the COVID-19 global pandemic, the study took place in two stages. The documentary analysis (workstream 2) took place during the COVID-19 pandemic lockdown (2020), while the primary empirical study was put on pause.
Once lockdowns were lifted, planning for the empirical study began. It was during the empirical study planning phase that the review (workstream 1) took place (June 2022–March 2023). The empirical study (workstreams 3–5) was a rapid study which took place between September 2022 and September 2023 (data collection lasted 6 months, December 2022–June 2023).
Workstream 1: Review of prison peer support services
Protocol and registration
This review was registered on PROSPERO (registration number: CRD42022351592).
Design
We conducted a ‘rapid systematic scoping review’ on prison peer support services.
We followed the rapid systematic review method proposed by Tricco et al.49 This method is systematic, but with adaptations to reduce the time required to carry out the review. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement50 was used.
Given the broad focus of the review, and the need for accumulation of evidence from different study designs, the review was also underpinned by elements of scoping review guidance.51 The combination of rapid systematic review methodology49 and scoping methodology51 was chosen to enable researchers to rapidly scope what literature was available in the broad area of peer support, while also systematically evaluating current evidence on peer support in prisons, within the time constraints of needing to inform the wider empirical study. This was appropriate given that the review covered both methodologies, but also integration of findings relating to effectiveness, cost, implementation and experience.
Eligibility criteria
The review was based on prison populations who deliver peer support services. We looked at all outcomes associated with peer support services relating to implementation, staff or prisoner experience, effectiveness, and cost. Inclusion and exclusion criteria are specified in Report Supplementary Material 1.
Search strategy
Information sources
To identify relevant peer-reviewed papers, we searched the following databases in June 2022: MEDLINE/Ovid (interface), CINAHL Plus/EBSCOhost (interface), EMBASE/Ovid (interface), PsycInfo/Ovid (interface), ASSIA/ProQuest (interface), Web of Science core collection. ASSIA was chosen as other systematic reviews on health and social care in prison populations have included this.52 We also hand searched the preprint server medRxiv. To identify policy documents and grey literature, we conducted a Google Scholar search and also searched relevant policy websites, and grey literature databases (Google Scholar, grey literature reports, King’s fund library, Social Care online, Social Policy and Practice). We also identified additional papers through the reference lists of identified review papers.39,53
One researcher conducted the search and inputted records into EndNote [Clarivate Analytics (formerly Thomson Reuters), Philadelphia, PA, USA]. Duplicates were removed.
Search terms
We included search terms for prisons and peer support services. We based terms for ‘prison’39,52 and ‘peer support’39,54 on search terms used in previous research. We iteratively developed search terms with support from the university librarian (Veronica Parisi). We limited searches to English only, but searches included international literature. Searches were carried out on subject headings and individual text words (see Report Supplementary Material 2).
Study selection
Identified studies were screened in three phases: (1) title; (2) abstract/executive summary; and (3) full text. One researcher (HW) screened all titles and abstracts. A total of 10% of excluded articles were reviewed by 1 of 2 additional researchers (CSJ, EM) at the title (n = 233) and abstract (n = 49) stages. Full texts were then screened by 1 of 3 researchers (HW, CSJ, EM), and 10% of excluded articles were reviewed by an additional researcher (n = 16). Additional researchers checked 10% of excluded articles in the title, abstract and full-text stages. Disagreements were discussed until consensus was reached. This is consistent with recommendations for rapid reviews.49 Studies were screened against the inclusion and exclusion criteria.
Data charting process
Data extraction was carried out using a data extraction form developed in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA). The data extraction form included details relating to study characteristics (e.g. the study title, year, setting, aim, design, population and analysis method) and review outcomes (e.g. findings relating to implementation, prisoner experience, staff experience, outcomes measured and findings relating to these outcomes, the data used to measure outcomes, findings relating to cost, and data used to measure cost/benefits).
The form was iteratively developed following initial screening of articles. It was piloted independently by three researchers (HW, CSJ, EM) using a sample of 5% of articles (n = 4). Disagreements were discussed until consensus was agreed. Data extraction for remaining articles was completed by one researcher. Data extraction forms were checked by one of three researchers (HW, EM, CSJ) when synthesising findings.
The review focused on the following topics: (1) types of peer support services in prisons; (2) effects/impacts of peer support services in prison on outcomes (e.g. health and social care outcomes, psychological outcomes and behavioural outcomes); (3) cost and cost-effectiveness of peer support services in prison; (4) approaches for implementing peer support services in prison and factors influencing successful implementation; (5) prisoner (peer and recipient) and staff experience; and (6) measures used to quantify impact of peer support.
Analysis
We used narrative synthesis55 to analyse study characteristics and analyse types of peer support used. Findings relating to the research questions were grouped and analysed in three sections: (1) findings relating to implementation and prisoner and staff experience; (2) findings relating to effectiveness and the data that have been used to measure it; and (3) findings relating to cost and the data used. Studies that included more than one of these aspects were included multiple times.
The papers from the studies that included quantitative measures of effectiveness or cost were divided equally between two of the researchers to scrutinise in more detail against the criteria mentioned above. Findings relating to the effectiveness of peer support interventions were then grouped into themes and the outcomes relating to each theme were then pooled together to better observe agreements and differences between studies.
We extracted all passages from the included studies that related to findings regarding implementation, staff experience, peer experience and recipient experience. Extracts were coded line by line and grouped into themes and subthemes by one researcher (HW).
Critical appraisal of individual sources of evidence
We used the Mixed Methods Appraisal Tool56 to evaluate the quality of studies. This tool was deemed to be appropriate due to its applicability to quantitative, qualitative and mixed-methods studies. The tool was applied to quantitative studies by one of two researchers (CSJ/EM) and qualitative studies by one researcher (HW). Mixed-methods studies were assessed jointly by researchers with qualitative (HW) and quantitative expertise (CSJ/EM).
Workstream 2: Documentary analysis of adult social care in prisons
A summary of the methods is below, for detailed methods please see Walton et al.1
Design
We conducted a documentary analysis of His Majesty’s Inspectorate of Prisons (HMIP) prison reports57 which report findings from prison inspections conducted by Ofsted or Estyn (Wales), the Care Quality Commission, the General Pharmaceutical Council and HMIP. These reports also present the results of a survey of randomly selected prisoners that was conducted at the start of every inspection. We chose to include a documentary analysis as it enabled the research team to explore the wider context and status of social care provision in the UK through the lens of published inspection reports. The documentary analysis was also an important part of the scoping work for the empirical study, as it enabled researchers to systematically explore whether, and where, peer support services for social care were currently being used in England and Wales, and whether an empirical study evaluating peer support for social care would be feasible within a rapid timeframe.
Sample
There are 123 prisons in England and Wales. This research focused on social care delivered in prisons for adults (n = 115). Young Offenders Institutions (YOIs) for those under the age of 18 years (n = 5), secure training centres (n = 3) and prisons without inspection reports (2017–20) were excluded (n = 13). A total of 102 prison inspection reports were included (Figure 1).
FIGURE 1 A flow diagram explaining the sample for this study. a, The reports for two prisons were combined as one. b, One prison had two reports (one for male prison and one for women’s prison). (Figure not reproduced here; see the original article.)
Procedure
His Majesty’s Inspectorate of Prisons reports (published between January 2017 and June 2020) were downloaded from the Justice Inspectorate website57 between April and June 2020.
Data were extracted on the following topics: descriptions of social care (quotes from reports); any other information on social care provided in the reports; and information on social care indicators. Reports (n = 102) were allocated to 1 of 2 researchers (HW/SMT). Researchers extracted qualitative and quantitative data from these reports.
Once all data were extracted, one researcher (HW) checked, coded and summarised the qualitative data (descriptions of social care and other information on social care) and one researcher (SMT) synthesised the quantitative data (indicators of social care) from collated survey data provided in individual reports (see Walton et al.1 for details). Initial findings were developed, discussed and agreed by both researchers and the wider team.
Analysis
Descriptive statistics were used to analyse the data, including how many prisons delivered aspects of social care, the different providers of social care provision and types of peer support initiatives. Due to the low numbers of prisons in some categories, we combined male prisons into higher-risk (categories A, B and A/B) and lower-risk groups (categories C, D and C/D) and conducted chi-squared tests (one-sided p-values, statistical significance threshold: p < 0.05) to identify any notable differences between them.
To explore social care indicators in more detail, we summarised survey findings relating to the support received by prisoners who considered themselves to have a disability and expected social care needs and support of prisoners on release. We grouped findings according to categories of prisons and gender. Chi-squared tests were conducted to see if there were any differences between categories of prison in the degree of support provided for prisoners reported to have a disability and if there were any differences in expectations for social care support after release.
Workstream 3: Implementation and experience
This workstream aimed to: (1) describe the context for peer support services in England and Wales at national and local levels and explore stakeholders’ views of these services; (2) analyse how peer support initiatives are implemented and the factors influencing implementation for adult social care in prisons in England and Wales; and (3) study the experiences of those delivering and receiving peer support initiatives for social care and whether these differ across models of peer support initiatives (including those who do not currently have peer support initiatives for social care) in prisons in England and Wales.
Sample and recruitment
Selection of sites
There are seven categories of prisons in England and Wales: category A – high-security prisons; category B – either local or training prisons; category C ‒ training and resettlement prisons; category D ‒ prisons that have minimal security and allow eligible prisoners to spend most of their day away from the prison; female prisons ‒ closed conditions; female prisons – open conditions; and YOIs which are intended for offenders aged between 15 and 21 years. This study did not include YOIs as it focused on adult social care.
We aimed to conduct prison lead interviews in approximately 20 prisons out of 123 prisons in England and Wales.58 To achieve maximum variation, we purposively selected prisons using a range of criteria, including whether or not they had peer support initiatives for social care, type of peer support initiative (drawing on information from the documentary analysis1 and scoping work), the area of the country, categorisation of prison and gender of prison (e.g. male or women prisons). Once a shortlist of sites was selected using these criteria, research leads and governors for each prison were contacted by researchers from University College London (UCL) and/or EP:IC consultants to find out if they would be interested in taking part in the study.
From these 20 prisons, we selected 5 case study sites (variety of prisons and social care peer support schemes). Case study sites were selected using a range of criteria, including whether or not they have peer support initiatives for social care, the type of peer support initiative [if applicable, and using information from (1) national stakeholder interviews and (2) national prison interviews], the area of the country, categorisation of prison, and gender of prisoners. We aimed to select a range of prisons with different models of peer-supported social care (prisons using different models of peer support social care, and one prison which did not provide peer-supported social care).
National and local stakeholder interviews
To describe the context for peer support services in England and Wales at national and local levels and explore stakeholders’ views of these services, we conducted interviews with national and local representatives (n = up to 10), including service commissioners, representatives from local authorities, representatives from HMPPS, representatives from national social care organisations and representatives from lived experience charities. As part of these interviews, we collated learning from the scoping stage and asked national stakeholders to provide further details on any gaps.
National and local stakeholders were recruited by contacting relevant organisations [e.g. HMPPS, Association of Directors of Adult Social Services (ADASS) Care and Justice Network, local authorities, and lived experience charities, such as the Prison Reform Trust] and inviting relevant individuals to participate. Additionally, the scoping work informed the purposive selection of national stakeholders.
Prison lead interviews
To study the implementation of peer support initiatives for adult social care in prisons in England and Wales, we conducted a telephone or video interview with nominated individuals who lead or run social care peer support initiatives in 20 prisons across England and Wales. We were flexible in the sampling approach (i.e. whether we interviewed prison staff, local authorities or provider organisations) as the scoping findings indicated that leads best placed to talk about implementation of the peer support social care initiative differ across different prisons (due to the different set-up of services).
Once the study received ethical approvals, we worked with HMPPS regional research leads and governors to identify relevant contacts in each of the identified prisons (e.g. prison research lead or other relevant individuals) to establish a nominated contact. Nominated individuals for each prison were identified from initial discussions with each prison. Nominated individuals were contacted by researchers from UCL and/or EP:IC consultants to take part in a telephone interview. We asked each prison to nominate someone to take part on the prison’s behalf (i.e. those leading or running the social care peer support service). Nominated individuals could be from the prison, local authority, or provider organisations (i.e. those who deliver training). At times, where appropriate, multiple leads were present on the call.
Stakeholders delivering and receiving peer support social care (case studies only)
To explore experiences of those delivering and receiving the peer support initiatives in prisons in England and Wales, and whether these experiences differed between different models of peer support initiatives (including those who do not currently have peer support initiatives for social care), we selected case study sites to explore experience in more detail. The study aimed to originally include four case study sites but ended up including five case study sites (to ensure maximum variation of models of peer support and prison characteristics). The researchers aimed to interview up to three to five staff involved in training/supporting buddies to deliver peer support and/or social care, up to three to five buddies (where available) and up to three to five recipients from each of the case study sites.
To recruit staff interviewees, the nominated contact at each prison supported the researcher to identify potential staff for interview and link potential staff with researchers from EP:IC consultants. Staff may also have cascaded the recruitment e-mail to other members of their team who were interested and appropriate. The researchers from EP:IC consultants contacted potential participants via e-mail and sent them a participant information sheet. Participants were given at least 48 hours to review the information and ask questions about the study. If the participant agreed to take part in the study, they were asked to sign the consent form. The researcher arranged a time to carry out the interview in person, over the phone, or via an online platform [Zoom (Zoom Video Communications, San Jose, CA, USA) or Microsoft Teams], depending on COVID-19 restrictions at the time of data collection, and participants’ preferences. An informed consent process using participant information sheets and written consent (scanned forms or typewritten/electronic signature) or audio-recorded verbal consent was used for recruitment to ensure informed and voluntary participation.
For interviews with buddies and adults receiving social care support in prisons, the prison governor (or a nominated individual) selected potential interviewees. To ensure a range of views, we asked services to help purposively identify potential interviewees with a range of experiences and demographic characteristics. In addition to the purposive sampling, we also circulated a research advert in the form of a printed leaflet to potential participants to reach a wider pool. As not everyone in the prison would have been involved in the delivery of peer support social care or the receipt of peer support social care, and due to restrictions relating to COVID-19, we asked the nominated contact at each prison to arrange for posters to be distributed directly under the doors of all those who have been involved in delivering and receiving peer support social care (in those prisons with such a scheme), or receiving social care (in the prison without a peer support scheme). The leaflet provided details of two ways to respond if interested, including speaking with a nominated individual and a tear-off slip for potential participants to return. Names were collated by the contact in the prison who liaised with researchers from EP:IC consultants to arrange interviews. We also requested for a printed poster detailing the same as the leaflet (minus the tear-off slip) to be placed on the wing notice board in case the advert was lost for any reason, to trigger memories, or to catch those who had not been selected to receive the leaflet.
Rationale for sample selected
In this study, the target number of interviewees (up to 10 national stakeholders, prison leads in each of the 20 prisons, up to 20 staff, up to 15 buddies, up to 20 recipients and up to 20 workshop participants) were selected to ensure that the study was feasible to complete within the rapid timeframe, while also providing enough in-depth data to answer the research questions and thoroughly evaluate peer-supported social care from different perspectives.
Measures
Interview topic guides were developed for each type of stakeholder (national and local stakeholder interviews, prison lead interviews, staff, buddy and recipient interviews) (see Report Supplementary Material 3). The interview questions for buddies and adults receiving peer social care support were reviewed by the service user involvement group prior to interviews (during a PPIE meeting).
Data collection
National and local stakeholder interviews, prison lead interviews and staff interviews (case studies)
For national and local stakeholder interviews and prison lead interviews, data collection processes were as follows: Potential participants were sent an information sheet and given at least 48 hours to review the information and ask questions about the study or discuss the study with the researchers. If the participant agreed to take part in the study, they were asked to sign the consent form and return it via e-mail. Alternatively, audio-recorded verbal consent was used in some cases to obtain consent at the start of the interview. Interviewees were told that they were free to withdraw at any time up to 2 weeks after the interview had taken place, and that if they withdrew all interview data provided would be removed from the data set and destroyed securely.
National lead interviews were conducted online/over the telephone by one researcher (HW), prison lead interviews were conducted online/over the telephone by three researchers (HW, LW, DG) and case study staff interviews were conducted online/over the telephone/face to face by two researchers (LW, DG) using the relevant topic guide (see Report Supplementary Material 3). The interviews aimed to last between 30 minutes and 45 minutes, depending on how much stakeholders had to say.
Stakeholders delivering and receiving peer support social care (case studies only)
Potential participants were given a participant information sheet to read and think about taking part, verbally informed about the study, and asked if they would be happy to speak to a researcher. If they agreed, the researcher from EP:IC consultants arranged a time for the interview. The researchers from EP:IC consultants conducted all interviews with buddies and recipients in person, on a pre-arranged date with the prison. We were flexible with each of the prisons to ensure that the mode of data collection was appropriate for them, in line with COVID-19 restrictions and prison preferences.
The logistics of face-to-face interviews with prisoners was discussed with a special point of contact (SPOC) in each prison, to understand the prison’s regime and staffing resources prior to arranging the interview dates and times. This extended to the interview times for each prisoner, which could not clash with any external visits and, as much as possible, with any gym or association sessions. With the SPOCs support, we ensured that the interviews were undertaken in a safe and confidential space, and staff were aware the interviews were taking place. We provided information to the prisoner in advance about the purpose and content of the interview, and ensured the prisoner understood this at the start of the interview, when we read through the consent form. The interviewers took time to build rapport and were experienced in interviewing in this setting, and drew on this experience throughout. Participants were asked to sign a consent form or provide recorded verbal consent. Interviewees were told that they were free to withdraw at any time up to 2 weeks after the interview had taken place. Participants were told that if they withdrew, all interview data provided would be removed from the data set and destroyed securely.
While this was not needed in the study, it was planned to offer translation services or translate materials as required for those who did not speak English or who had communication difficulties.
Interviews were conducted by one of four researchers from EP:IC consultants (LW, DG, SR, PH) using the appropriate topic guides (see Report Supplementary Material 3). Members of the data collection team met regularly to ensure consistency in interviewing approach and data collection processes. Buddy and recipient interviews aimed to last between 30 minutes and 45 minutes.
All interviews
All interviews were semistructured, audio-recorded (subject to consent), transcribed verbatim by a professional transcription service, anonymised and kept in compliance with the General Data Protection Regulation 2018 and Data Protection Act 2018.
Data management
Interviews and workshops (qualitative data) were recorded on an encrypted, password-protected digital recorder (only the researcher knew the password). Data were collected by researchers from UCL and EP:IC consultants. Consent forms and audio recordings from interviews and workshops were securely transferred from researchers at EP:IC consultants to researchers at UCL using the Data Transfer portal on to the UCL Data Safe Haven (DSH – a secure electronic environment, certified to ISO27001 information security standard and conforms to the NHS Information Governance Toolkit).
Digital audio recordings of interviews and workshops were sent to a UCL-approved contractor for transcription (TP Transcription Limited). Once transcripts were received, a researcher from UCL uploaded them on to the DSH and fully anonymised them (names and places) and organised them by participant codes. Anonymised transcripts and other relevant data were stored in a secure folder to which only the named researchers had access. Only the research team (UCL and EP:IC consultant researchers) had access to participants’ personal data (i.e. name and contact details). A password-protected spreadsheet of interviewees and their details was held on the DSH. Participant identifier codes were stored in the DSH and kept separate from study data.
Analysis
Data collection and analysis were carried out in parallel and facilitated through the use of rapid assessment procedure (RAP) sheets as set out in Vindrola-Padros et al.59 Data analysis was conducted by RSET researchers and EP:IC consultants.
Rapid assessment procedure sheets are a research tool that has been developed to facilitate data analysis for rapid qualitative research projects.59 RAP sheets are a template document that include headings relevant to the research questions that can be used to summarise key findings from interviews in a rapid way. The categories used in the RAP sheets were based on the interview topic guides, and we maintained flexibility to add categories as the study progressed. For example, the RAP sheet included details relating to the research questions, such as subheadings for: what social care provision looks like in the prison, what peer support schemes look like, experiences of those delivering the service, experiences of those receiving the service, barriers to implementation, facilitators to implementation, etc.
We developed one RAP sheet per prison to facilitate cross-comparisons, and on this RAP sheet included findings from each group of participants (staff, buddies and those receiving peer support). Researchers added notes and summaries of findings to the RAP sheet following each interview. The data inputted into the RAP sheets were inductively coded using thematic analysis. Themes and subthemes were developed, discussed and agreed by the research team. We held an analysis meeting with the EP:IC researchers involved in data collection to ensure that the developed themes were representative of their interviews.
Following this initial inductive thematic analysis, we conducted additional in-depth coding of transcripts to gain a more thorough analysis of key themes. To do this, a coding framework was developed, drawing on the initial themes and subthemes identified. All transcripts were then deductively coded using this coding framework by one researcher (HW). Quotes for each theme and subtheme were extracted, to illustrate key themes and subthemes. Findings were discussed and agreed by the research team. This process is consistent with analysis methods used in previous research.60
Meetings with the evaluation advisory group and PPIE group identified topics to explore in more detail (e.g. payment and differences across different types of prisons). The final findings were discussed and agreed by all researchers involved in the project.
Workstream 4: Outcomes
This workstream aimed to investigate the available data, specifically, potential markers of success and challenges associated with peer support programmes and what options there may be for quantifying them. Evidence from initial scoping interviews suggested that available data for a quantitative evaluation were likely to be scarce. So, rather than aim to undertake primary analysis as part of this project, the intention was towards guiding future data collection and analysis and analysing theories of change to support decision makers.
Design
We used a five-stage process:
- Investigate where the benefits of social support services are likely to be realised, including those providing and receiving peer support, staff, health and social care services, and the wider prison community.
- Investigate how any of these could be measured (regardless of whether or not the data exist).
- Understand what data are available and their quality/completeness.
- Make recommendations for data collection, which might mean new data collection or making use of existing data.
- Analyse theories of change.
To understand the measurement of impact and availability of data, we gathered and assimilated information from the interviews with national leads and prison leads conducted in workstream 3 and conducted a stakeholder workshop with specific follow-up interviews. We also aimed to view any relevant data sources to help us understand how they were being used and their potential.
The stakeholder workshop was designed to enable participants to engage in group discussion, while also engaging in a series of activities that would enable them to work together to discuss and agree on outcomes, their measurement and potential gaps.
Sample
We invited 23 participants to the workshop, including service commissioners, researchers and representatives from HMPPS, social care organisations, local prisons, local authorities and lived experience organisations. We purposively sampled participants to ensure that a range of experiences were represented, including their role, their region, and experience of peer support initiatives for social care. Thirteen invitees attended on the day. Further interviews were conducted with three researchers who were not able to attend and a further individuals who had specific knowledge of certain data.
Procedure
The workshop was held online using Zoom and lasted for 120 minutes. It was structured into three sessions each with three separate breakout groups, facilitated by one member of the research team (CSJ, HW, EM), followed by feedback from all attendees. The topics for the different sessions were:
- benefits, shortcomings and challenges of prison social care peer support initiatives
- existing data that could be used to measure impact and costs
- identifying gaps between potential impact and existing data. Recommendations for future data collection.
These topics also formed the basis for the further interviews with researchers.
Where there were gaps between proposed measures and data availability, we explored the feasibility of new data collections to capture this information. New collections could cover routine data or bespoke data for future prospective studies. We also discussed how progress with data collection could be facilitated.
Notes from each breakout session were taken by another member of the research team (LW, NJF, DG) and each session was recorded.
Analysis
Information gathered from the workshop was collated and structured by one member of the team (CSJ) and shared with the other members. This was combined with further information, including the benefits and shortcomings that were identified in the follow-up and workstream 3 interviews. These findings, together with information obtained from interviews, led to further enquiries looking into possible data sources with those who either hold or collect the data. These data sets were assessed against a number of criteria, for example:
- data quality
- data completeness
- collection period (e.g. routinely collected or collected once)
- how many prisons they covered or how many prisoners in participating facilities
- how well they matched the population we were interested in.
Workstream 5: Costs
First, we sought to identify the cost components of these programmes and their sources. We sought also to investigate how such sources and costs may vary between the different programmes delivered and identify all sources of information for collecting these data on costs. Given that mode of delivery for such peer support programmes differ (e.g. they may be led by a local authority, led in-house by prisons, privately provided, or supported by the charity sector), we approached the main actors to get a better picture of the information on costs and their availability.
Given the non-systematic data collection and their limited availability, we adjusted the main intention of the analysis to the development of a cost template that might guide future data collections, suitable for cost analyses. This approach is similar to that used in workstream 3 and, combining findings from these two workstreams, enables the economic evaluation and cost-effectiveness of the peer support programmes in prisons. Theories of change were analysed in conjunction with workstream 3.
In workstream 5, we proposed a staged approach that sought to identify and investigate the following research questions:
- What sources of data on costs of peer support programmes are available, what is the quality and completeness of these data and who can provide this information?
- What cost-related data we will need for a future economic evaluation and cost-effectiveness of the peer support programmes in prison? What is the current availability of these data?
The methods used are described below.
Investigating what sources of data on costs of peer support programmes are available, what is the quality and completeness of these data and who can provide this information
Design
This part of the study drew on data collected from the review (workstream 1) and stakeholder workshops (workstream 4).
Procedure
We used the workshops conducted under workstream 3 to collect information about the existing cost data on the peer support programme. First, we sought to understand all the activities linked with the peer support programme, such as training, workshops, and meetings, and the costs involved in providing these activities. These costs included working costs (i.e. costs due to paid working hours) as well as other operating costs (i.e. uniforms, handouts, etc.) used for the peer support programme. Second, we sought to understand who was involved in these programmes, including prisoners, prison staff, and external assistants, and what the cost was of their potential involvement. The workshops were used to also clarify other aspects of the existing cost data, such as the quality of the data provided, the time period the data covered, the number of beneficiaries (providers and receivers). Additionally, we asked relevant questions in the interviews conducted in workstream 3.
Analysis
We assessed the information on the cost data that were available, and we clarified the sufficiency of the information and the quality of the data.
Investigating what cost-related data need to be collected for a future economic evaluation and cost-effectiveness of the peer support programmes in prison and what is the current availability of these data
Design
This part of the study drew on data obtained from the review (workstream 1), the workshops (workstream 3) and qualitative interviews with national stakeholders, prison leads and staff (workstream 2). This information allowed us to develop a cost template that we shared with the contacts at the 18 prisons included in this study and which we asked them to complete and return.
Procedure
First, we asked the recipients of the cost template whether their prison collects any data relating to the peer support programmes and, if so, whether they would be willing to share these data with us. Those who answered positively to both questions were asked to provide information on a series of questions about the participation, the involvement, and the costs related to these programmes.
Analysis
As with workstream 3, we explored the feasibility of new data collections to capture the information of interest. Having received back the cost templates from the prisons, we first assessed the current data availability and the quality of the available data. We calculated the proportion of missing data per question of the cost template. Where doable, we used descriptive statistics to summarise the collected data.
Integration
All three workstreams informed each other and were mutually supportive. Data collection processes, data collected and findings were triangulated throughout.
Patient and public involvement
We developed this study based on stakeholder input, including from representatives involved in peer support initiatives, representatives from the prison service, representatives from social care organisations, academics and peer reviewers.
We worked with EP:IC consultants to develop a service user involvement group of individuals with lived experience. This group were involved throughout the project and provided feedback on the protocol, study methods and aims, data analysis and dissemination. We held four meetings with this group throughout the study with the service user involvement group to discuss these topics. The service user involvement group included people who have previous experience of being in prison. Additionally, we sought input from the RSET PPIE group throughout the study (e.g. when developing protocol, data collection measures and disseminating findings).
Evaluation advisory group
In this study, we developed an evaluation advisory group of key stakeholders with an interest in peer support scheme for adult social care, including representatives from HMPPS, representatives from adult social care organisations, representatives from local authorities, representatives from the UK Health Security Agency (HSA), and researchers from universities and think tanks who work in this area.