Sections of this chapter have been reproduced from Holloway et al.1 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.
For individuals in contact with the criminal justice system, the prevalence of at-risk drinking is far higher than in the general population. There is very little evidence of efficacy or effectiveness of alcohol interventions in reducing risky drinking amongst those in the criminal justice system and in particular men on remand in prison. This is compounded by the limited evidence for the optimum timing of delivery, recommended length, content, implementation and economic benefit of an extended alcohol intervention in the prison setting. APPRAISE aimed to provide vital evidence to inform a future definitive RCT of an extended alcohol intervention for men on remand in prison.
APPRAISE specifically aimed to undertake a two-arm, parallel-group, individually randomised pilot study of a self-efficacy-enhancing psychosocial alcohol intervention for men on remand in prison.
Results relating to objectives are detailed in the individual chapters and summarised below.
Objective 1
To pilot the study measures and evaluation methods to assess the feasibility of conducting a future definitive multicentre, pragmatic, parallel-group RCT.
- 1a. Is it feasible to conduct a future multicentre RCT of a self-efficacy-enhancing psychosocial alcohol intervention for men on remand?
We cannot say that it is feasible to conduct a future definitive trial as we were unable to follow up participants – of course COVID-19 played a part in the low response rates but, at this present stage, with the results we have we are not yet in a position to move this work forward to a definitive study.
- 1b. Can we obtain reasonable estimates of the parameters necessary to inform the design and sample size calculation for a future definitive multicentre RCT? This includes standard deviations of potential continuous primary outcomes and estimates of recruitment, retention and follow-up rates.
Yes, but only partially. We can obtain reasonable estimates of standard deviations of outcomes based on the baseline data collected. However, the amount of valid data collected at follow-up was too little to be able to provide us with reliable estimates of parameters for outcomes at follow-up.
The pre-pandemic recruitment data give us reasonable estimates of recruitment rates in a future trial. However, the retention and follow-up rates we observed in the APPRAISE trial were likely to have been heavily influenced by the COVID-19 pandemic and associated lockdown measures, and therefore these are not reliable indicators of what they might be in a future trial.
- 1c. How well do participants complete the questionnaires necessary for a future definitive RCT?
At baseline there was a very small amount of missing data for a few questionnaires/outcomes, for example, ‘Readiness to Change Ruler: The people who are important to me support me in reducing my drinking’, and for the ‘Drinking Refusal Self-Efficacy Questionnaire’ and for the ‘Negative Alcohol Expectancy Questionnaire’. The amount missing was very similar between sites at baseline. At follow-up, the amount of missing data was very substantial in both sites. We were able to do slightly better at the Scotland site because ‘we were able to have some access to the prison site and identify where participants were located’. In addition, the same researcher conducted the majority of the follow-ups in Scotland at 12 months as conducted the recruitment and baseline measures, which may have had an effect on participants’ willingness to respond. However, numbers are small so we must be cautious with this assertion. In terms of data completeness of the questionnaires at each site, this was similar between the sites (Report Supplementary Material 1 S1.1 and S1.2). It is, however, really difficult to read too much into the differences because the numbers completing data at each site were so small. Similarly, the different questionnaires had similar amounts of missing data at follow-up (see Table 28) and we cannot really say too much about differences of ±1 or ±2. In summary, we did not find any notable differences across the different sites/instruments. However, it could be that self-completion was a factor; at baseline the questionnaire was not self-completed, but RA-delivered. In addition, we also observed and reflected on the suitability of certain questions, particularly if the questionnaire was to be self-completed (which was reflected in later amendments) and this would warrant further consideration in future research for this population.
- 1d. Can we collect economic data needed for a future definitive RCT?
We were able to design and implement an ABC study that provided detailed estimates of the implementation costs of the APPRIASE intervention. That we could do this despite COVID-19 complications suggests that our methodology will be appropriate for a definitive RCT. We nonetheless note that additional attention needs to be devoted to some aspects of our ABC study. In particular, more attention should be devoted to assessing the prison space needed to deliver the intervention and more attention should be given to separating research costs from true intervention costs when it comes to scheduling follow-up intervention visits. We were also able to modify the Economic Form 90 for use with a UK prison population and found no evidence that it was difficult for study subjects to complete. However, were unable to validate the data-collection instrument and recommend that such validation be part of a definitive trial.
- 1e. Can we access recidivism data from the PNC databases for trial participants?
- 1f. Can we access health data from routine NHS data sources for trial participants?
Although it is possible that COVID-19 played a role in our inability to obtain these data, we do not think that it was solely responsible. Two key issues are at play: prisoner confidentiality and data systems. We found that gaining approval to obtain the PNC or NHS identifiers was difficult to impossible due to a combination of prison and ethics board concerns. Although these concerns may fade as more evidence of the benefits of the APPRAISE intervention mounts, we conclude that it is very likely that these concerns will still exist for a definitive trial, and possibly be even more salient for some stakeholders in that context. Exacerbating the confidentiality issues were data-system issues that prevented data access on a large scale. Although this issue depends on the jurisdiction, in each of our two jurisdictions at least one of the data sources required case-by-case data extraction at some step of step of the process. We could not extract either PNC or NHS data on a large scale in either England or Scotland, making the use of such data prohibitively costly for a definitive RCT.
Objective 2
To assess intervention fidelity
- 2a. What proportion of the interventions are delivered as per protocol?
We recorded three of the in-prison interventions. They all scored highly and had maximum scores on the BECCI Scale.
- 2b. Is there any evidence of contamination between the two conditions and/or between those workers delivering the intervention?
We did not find any evidence of contamination between workers delivering the intervention. At the Scotland study site, the intervention was only delivered by one interventionist. In England two interventionists were delivering the intervention. There was mention of discussion of the study between one of the remand participants and another (see sub-section ‘Coherence: acceptability of the APPRAISE intervention’ in Remand participants) but the participant stated that he didn’t know if the other participant got the intervention or not. No other evidence was obtained of contamination; however, it would be prudent to explore this again in any future work in the absence of a global pandemic, lockdown and restrictions across the prison estate.
- 2c. To what extent was the intervention changing process variables consistent with the underpinning theory?
Table 28 shows the results for changes in process variables (secondary outcomes). The available data do not allow answering the question if the change in process variables from baseline to 6 and 12 months was greater in the intervention group than in the control group in central process variables (goals, self-efficacy, outcome expectations, according to social cognitive theory), as we were able to obtain only limited data for change in mediators (especially at 6 months) and only a small number of participants were released, which precludes interpretation of change in mediators after release from prison. We found small and inconsistent changes in process variables with considerable variation. Therefore, we cannot answer this research question.
Objective 3
To qualitatively explore the feasibility and acceptability of a self-efficacy-enhancing psychosocial alcohol intervention and study measures to staff and for men on remand and on liberation.
- 3a. How acceptable are the trial and intervention procedures (including context and any barriers and facilitators) to the following key stakeholders: men on remand in prison and on liberation; prison staff (including healthcare staff); commissioners; policy-makers and third-sector partners?
The qualitative process evaluation showed that the trial and intervention procedures were acceptable to staff and men on remand. We found that it was important for an individual to feel comfortable and trusting of an interventionist before sharing their personal experiences and consumption. Staff did report that the prison setting and resources could affect the work due to the lack of available space and the limited staff capacity. However, in general participants believed that delivering the APPRAISE intervention did not significantly increase their workload. The intervention training and content appeared to be largely acceptable, although the length and content could have been condensed. The APPRAISE intervention delivery benefited from buying into the intervention and being motivated and engaged with the content.
Similar findings were found around the prison’s logistical issues affecting the APPRAISE intervention delivery feasibility, such as space and staff capacity, the importance of engagement with prisoners and the lack of currently available support for prisoners on remand from the stakeholder interviews. Additional themes and subthemes that emerged were around the intricacies of the prison environment and how it affected the study and intervention delivery, the individual differences across the setting and across the participants and also further barriers and facilitators to delivery of the ABI.
A key finding, observed in all three participant groups, was that the in-prison APPRAISE sessions were deemed to be more acceptable and feasible than the post-liberation intervention sessions. This was observed due to a plethora of different factors, such as being unable to re-establish contact with males upon their release and individuals relapsing to previous risky alcohol consumption when alcohol becomes more freely available.
Objective 4
To assess whether operational progression criteria for conducting a future definitive RCT are met across trial arms and study sites and, if so, to develop a protocol for a future definitive trial. Operational progression criteria are based on previous research results.6
A large proportion of this study was carried out during the COVID-19 pandemic and as such we cannot demonstrate the full extent to how this affected the study. However, we did not meet the criteria for developing a protocol for a future definitive study.
- 4a. Do the two prisons invited to the study agree to take part?
Yes. Both Scotland and England prisons agreed to take part in the study and stayed involved until COVID-19 restrictions closed access to the project down.
- 4b. Based on knowledge from previous data, do at least 90 eligible participants consent to take part and be randomised across the trial arms?
No. Although in the England site, where recruitment began earlier than the Scotland site, we recruited 90 participants, the COVID-19 pandemic and associated lockdown restrictions meant that we were not able to enter the prison and therefore could not achieve the target of 90 participants in the Scotland site (n = 42). However, in a future study without lockdown restrictions we are confident that the required number of participants can be recruited.
- 4c. Do at least 70% of participants who consent to the trial receive the intervention?
Yes, partially. Of the 68 consented and randomised to the trial 53/68 (78% of participants) received the in-prison intervention, although much lower numbers received the post-prison component of the intervention. Only one participant received the day-3 post-liberation intervention, none received the day-7 post-liberation intervention, and one person received the 21-day post-liberation intervention, at least in part because of the COVID-19 context. A post-prison-liberation component of the intervention may not be viable in a future trial.
- 4d. Are at least 60% of those who received the intervention followed up at 12 months across trial arms and study sites?
No. Only 13% (18/132) of those who received the intervention were followed up at 12 months. Although the COVID-19 pandemic was undoubtedly a factor in the low percentage followed up, we observed high percentages of participants who were not released at 12 months; this was worsened by COVID-19 because of the number of those on remand awaiting trial due to suspended and closed courts; or of those who were released, many could not be contacted or could not be located. We consider this to be the greatest barrier in designing a future trial that will achieve satisfactory follow-up rates.
Objective 5
To ascertain what alcohol services are available in male remand prisons and how COVID-19 has affected services.
From 59 prisons in Scotland and England, successful e-mails were sent to 55 prison governors. Seventeen (31%) were completed. A limitation of the study was the barriers to freely accessing governor e-mails. While this was handled by a gatekeeper who worked for SPS for the Scottish prisons, no such facilitation took place in England, and while most e-mails were freely available, many were not. Two-thirds of the sample were interested in hearing more about ABIs.
- 5a. What alcohol services are currently provided within male remand prisons?
There are a wide range of services from a wide variety of providers. The most common were AA sessions (59%), NHS services (35%) and psychotherapy, counselling or 1--1 support (29%).
- 5b. To explore the prison governors’ understanding of brief interventions.
Two-thirds of the governors understood what ABIs are and that they were being delivered.
- 5c. To understand how the COVID-19 pandemic has impacted the services available in male remand prisons.
We found that the COVID-19 pandemic had a devastating affect across all prisons in relation to the services they provide. It was also useful to establish what the prison governors deemed important learning from the pandemic, in this instance mostly around PPE and a better use of technology.
- 5d. To identify whether the impact of the COVID-19 pandemic upon services could be avoided in the future.
There were major impacts in relation to services during COVID-19 which included all sessions being suspended and there being staffing issues and a lack of resources.
Patient and public involvement
Patient and public involvement (PPI) has informed and influenced the development of the study through various stages. PPI representatives involved in the research design include: those with experience of being in prison and the criminal justice system, a social enterprise group, prison employees and NHS staff. At the time of the study conception and development, our PPI co-applicant, Sharon Mercado (SM), was a community justice mentor (CJM) at the Wise Group. As part of our PPI work, she brought together service user representatives and held a participatory workshop with men who had experience of being in prison on remand. The aim of the workshop was to discuss and provide suggestions and feedback regarding development of the study design. Their views informed and influenced the study design: need for pictorial information for study material to address literacy/readability; strategies to maximise follow-up by obtaining contact details of family/significant others, homeless shelters and community justice support workers. SM was a member of the PRISM-A study Advisory Board and has continued to work with us as a co-applicant on the APPRAISE study, where she informed the research design and follow-up strategy using experience of her own follow-up work as a CJM with prisoners on liberation. SM continued to provide input throughout the duration of the study as a member of the PMG and in the recommendations for a future definitive trial. Wider PPI was present through TSC membership where we also had wider representation.
The PPI was an extremely positive experience for the team and the study benefited hugely from all their input. This was particularly the case when we were reviewing the impact of COVID on follow-up and considering the mitigations we could take moving to self-completion questionnaires, change of primary outcome measure and adaptations to some of the secondary outcome measures. The input from SM was key. The impact of COVID and the subsequent restrictions on face-to-face meetings did have a negative effect on PPI engagement. The face-to-face meetings were key for the relationships we had been building and meeting face-to-face with PPI colleagues was always a positive experience for everyone.
Reflections and lessons learned from patient and public involvement involvement
Our reflections and learning of our PPI engagement in the APPRAISE study was that we should in future look to strengthen the PPI presence within any future studies. It would have been useful to have created a PPI-led space for the PPI members for them to have the opportunity to meet away from the PMG and TSC. On reflection the TSC and PMG meetings could be quite technical at times, with language used that PPI colleagues felt they were less familiar with. As SM was the sole PPI member of the PMG, in future it would be useful to have a minimum of two members to have a more balanced group. If we created a separate PPI space where PPI members could meet outside of formal meetings to set the agenda, it could be used as a safe space for them to discuss and provide regular feedback to the TSC and PGM. It may be possible to also run workshops linked to the study but also to have informal activity too. With SM we would generally meet ahead of the PMG and have refreshments together; this worked well and has informed the idea of a separate PPI space for PPI colleagues.
SM has provided her own insightful reflections on the research study in relation to what went well and what could have been done differently to improve the PPI representation and involvement. These reflections and recommendations, in her own words, are presented here.
Engaging with offenders is not easy. It takes time and patience and recognising that sometimes, despite the willingness to engage, it doesn’t always happen the way you anticipated.
Holding a focus group needs careful planning especially as you need to make sure that those attending will feel safe, and anything said is confidential. Offenders’ lives, especially those that are involved with substance misuse/alcohol, are chaotic. Therefore, making sure those attending are not going to meet anyone else that they have issues with (enemies, debts etc.) requires input. Despite all the support to attend (incentives were included, food vouchers, travel expenses), the focus group was not attended by all those invited. However, even with a small number, input was good. Those still on remand were also on an individual basis asked for input regarding the information leaflets, and what works best in terms of follow-up appointments after liberation.
To summarise, the whole experience was challenging. Regardless of the research, working with offenders/ex-offenders can be difficult, especially with those who due to their alcoholism can have chaotic lives. This makes it hard to continue engagement with follow-up appointments. There are things that can be done which may assist in getting increased involvement.
- Building up more trust and spending more time with the offender can help. Rather than just a few engagements, have a few more or enlist support from other agencies working with the offender just to build up the momentum and remind the offender of the research project. This could prevent the offender from disengaging.
- Better offender tracking after liberation. Some people may have a mentor or a support worker: find out (hopefully before liberation) who is currently working and keep up communication with them, they may be able to inform you of a person’s whereabouts.
- Offering incentives may help. Whilst offering financial incentives (£20 per community engagement), it can help. It is recommended that any financial rewards are given in food/shop vouchers or pre-payment energy support.
- Invite those who are now in recovery and have recent relevant experience to come and talk to the research team at the start of the project. It can be a good opportunity to find out more about the lives of those affected by alcohol and why engagement or non-engagement happens.
Equality, diversity and inclusion
Equality, diversity and inclusion (EDI) was a core value of the purpose of our research by the very nature of the setting of the research, the focus on alcohol-related harm and the population we were focusing on. Conducting the study required EDI to be at the forefront of our work from inception with our awareness of the lack of evidence, to meeting ethical approval requirements and working within the criminal justice system. An overview of our experience and reflections are provided in relation to participants, the evidence base, the study design, the research team and PPI representation.
Population characteristics and optimising participation
We are aware that undertaking research within the prison setting is challenging and not commonplace and for trials research it is even less common. The challenge is increased even further when there is a focus on the remand population. The study was developed and conceptualised by drawing on the collective experiences of the research team, our previous research and work with those incarcerated and from the published literature. From this we had a unique understanding of our population and who should be in the study. In our previous research we had undertaken interviews with men in prison who were on remand and asked them specifically how we could optimise participation in a future trial. We had also undertaken interviews with prison service colleagues providing care in the prison setting. Their reflections and insights directly informed the study design and how we approached the research as described in Chapter 3 to be as inclusive as possible. We also purposely undertook the study in two different prison estates to explore how the impact of the variations in category of the prison population and the estate regimes impacted participation. This was a factor from our previous work we had become aware of and remains a key factor for researchers to consider in any future studies along with the use of digital technology from the use of digital recorders to use of mobile phones.
Gaps in the evidence base
To the best of our knowledge we are not aware of any other trials for men on remand focusing on the delivery of a self-efficacy-enhancing extended alcohol brief intervention, at least in the UK, hence the need to undertake the pilot trial. The key gap that remains for this particular population is how best to design a study that provides an optimum follow-up process. We have provided recommendations in Chapter 9 for future studies.
Encouraging under-represented groups
Our PPI engagement with service users during the conception and design phase of the study supported the development of the study material through use of images and the material layout. This was specifically with the aim of optimising accessibility and inclusion.
Research team inclusivity and development opportunities
Our research team comprised a range of under-represented stakeholder groups and individuals largely seen in our TSC membership. The research team also had a good gender balance with a larger percentage of women and early-career research representation evident as co-authors. The ethos and values of the team ensured development-supported opportunities were available to take on key elements of the study, for example, process evaluation interviews and data analyses and COVID survey. Similarly, we provided development opportunities for those coming from a non-research background, for example, our co-applicant service user (SM) and our researcher who collected the baseline data at the England study site.
Patient and public involvement representation
Our PPI representation is outlined in the section Patient and public involvement. The study participants were representative of those from our previous work with this population. The impact of COVID has been described elsewhere in the report. Our additional COVID impact survey provided some further data on how this population were affected in relation to the delivery of alcohol services during the pandemic.
Strengths and limitations
Strengths of the study
- The APPRAISE study, to the best of our knowledge, is the first pilot trial of an alcohol-focused self-efficacy-enhancing intervention for men on remand.
- The study provides significant insight into the feasibility and acceptability of pilot trials for this particular population.
- The ethical approval process for the prison setting across devolved countries provided useful insights into how to navigate the submission and approvals process across a range of ethics committees.
- The study provided good evidence of the feasibility of recruitment, training interventionists and subsequent in-prison delivery of the APPRAISE intervention to 73% of men on remand, randomised to the intervention, within a highly complex prison setting.
- The economic findings provide valuable insights, as we believe this to be the first of its kind for this type of intervention in this setting and that the methods discussed will likely be relevant and transferable across other settings.
- Useful insight regarding the access to PNC and NHS service use data was obtained.
- The process evaluation provided useful insights into the perspectives of the remand participants, intervention delivery team and wider stakeholders on the acceptability and feasibility of implementation of the intervention.
Limitations of the study
- Protracted multiple ethical approvals across devolved countries and processes meant a significant delay in recruitment commencing, particularly for the Scotland site.
- We were unable to deliver the post-liberation elements of the intervention.
- The onset of the COVID-19 pandemic resulted in prison restrictions and no access to the prison site, meaning we were unable to identify if participants had been released or not at 6 months, with no access still at the England site at 12 months and very limited access at the Scotland site at 12 months.
- Although the COVID-19 pandemic was undoubtedly a factor in the low percentage followed up, only 13% (18/132) of those who received the intervention were followed up at 12 months.
- Changes to the study protocol to include self-completion of follow-up data may have made it difficult for participants to complete the survey and resulted in larger proportions of missing data as opposed to RA survey completion at baseline, where missing data were minimal.
- We did not include probation services in the trial, which on reflection would have strengthened our post-liberation follow-up process.
Recommendations regarding a definitive randomised trial
From the study progression criteria for a full RCT, we have identified the following recommendations:
- Buy-in for a research trial of this nature in prison requires significant pre-study trust and relationship development, buy-in from the prison estate, governor, prison officers, peer mentors and embedded third-sector services.
- Recruitment and randomisation of men on remand to a future APPRAISE RCT are possible, with trust in the research team an important factor.
- Training team members of existing alcohol services to deliver APPRAISE as per protocol is possible.
- Delivery of the in-prison APPRAISE intervention is possible and would require appropriate space, time and team member capacity.
- Economic evaluation is possible.
- Post-liberation intervention delivery would only be possible if there was a robust follow-up process identified and in place.
- Six-month and 12-month follow-up would only be possible if there was a robust follow-up process identified and in place. What that might look like needs to be carefully considered, explored and understood in an already complex liberation and rehabilitation statutory and non-statutory agency landscape. This may be achieved through the use of in-depth individual case studies or a realist data-capture approach to cover this period.
- Further exploration of the inclusion of and collaboration with the probation service in the service delivery and implementation of the APPRAISE intervention may offer higher success rates of post-liberation intervention delivery and a more robust follow-up process.
- Further exploration of the feasibility of utilising routinely collected data to evaluate outcomes post prison release is also warranted, for example, presentation at community alcohol services, presentation to GPs, and/or liver function test readings on routinely collected blood-test investigations.
- Future research funding amongst prisoner populations must be made available to address the feasibility issues identified in this pilot trial so that a best-practice framework can be developed and utilised by research teams seeking to conduct much-needed research amongst this under-served group.
Conclusion
Addressing alcohol harm in prisons, at what can be considered a ‘teachable moment’, where there is an opportunity for reflection on their risky drinking and their current position, could potentially reduce the risk of re-offending and costs to society, while helping address health inequalities. For those on remand in prison, the prevalence of at-risk drinking is between 62% and 68%.7
Many men on remand do not have the opportunity to access mainstream prison health or public health services. An intervention such as APPRAISE offers an opportunity to provide an extended behavioural-based alcohol intervention to men on remand. The APPRAISE study has identified that despite the complexities of ethical approval and the time taken to build relationships and trust, it is possible to undertake key elements of a future RCT, but not all, namely follow-up. Further focused research needs to be undertaken to explore, identify and develop a robust process and system to optimise follow-up post liberation. A recent pilot trial in prisons has shown promise in this area, having identified satisfactory follow-up at 6 months, but not at 12 months.90
The evidence base to meet the needs of men on remand in relation to risky drinking remains weak. However, there are opportunities to build on the work of APPRAISE to ensure that those in the prison setting have equal access to interventions that have the potential to positively impact their relationship with and use of alcohol.