Trial timelines
The project was planned to be delivered in 36 months, with the first 2 months for recruitment and training of research staff, and to ensure that all approvals were in place. A full-time trial manager was appointed before the formal start of the trial, funded from other resources, to ensure that all procedures and permissions were in place; this took an additional 10 months. The recruitment period was planned for 32 months, with 2 months for analysis and preparation of the final report. The recruitment target was met within this timeline despite HMP YOI Isis being told it could no longer randomise participants in February 2019 because the pharmacy had trial kits left for only one treatment arm. Therefore, the final participants were randomised at HM YOI Polmont to ensure that recruitment from the sites was more even and to avoid the need to transfer trial medication kits from one site to the other.
A 5-month no-cost extension was requested in February 2019 for two main reasons:
- Extensive database checks were needed before the database lock and there was not sufficient time to complete these. The task was made more difficult because it was not possible to conduct all of the database checks outside of the prisons, and there were additional restrictions on access to medical and prison records dependent on whether a participant was still in prison, released, transferred to another prison or deported. These checks needed to be undertaken while research staff were still employed and on site.
- The SAP for the project was very detailed and required 2–3 months of analysis. In addition, more time was needed to write the report and obtain feedback from co-applicants during the writing process.
Following database lock and review of the preliminary SAP, there was a planned meeting in November 2019 to discuss the trial findings with the DMC, TSC and Trial Management Group (TMG). This meeting led to a recommendation to conduct additional post hoc analyses that might provide an explanation for the main findings. As these further analyses were proposed after database lock and review of the main trial results, they should be regarded as exploratory in nature. The aim was to identify potential factors that might explain the trial findings and that could be considered in future research. These further analyses are reported in Chapter 5. These further analyses were acknowledged, but not formally reviewed, by the DMC and TSC chairpersons. To allow for this work, an additional no-cost extension was approved until 30 June 2020.
Trial set-up
The CIAO-II project started on 1 August 2016; within the first 3 months, it had received a favourable opinion from the East of England–Essex Research Ethics Committee (REC) (reference number 16/EE/0117, Integrated Research Application System project identifier: 179456), as well as HRA approval. The ethics application was submitted prior to the introduction of the HRA and approval was given for the trial to be processed through pre-HRA approval systems. Research and development (R&D) approvals were obtained from Oxleas NHS Foundation Trust and NHS Forth Valley (reference number FV908). In addition, permission was required from HMPPS (reference number 2016-129) and the SPS.
The KCTU provided the randomisation service, IMP management, and the clinical trial database system with full audit trail (InferMed MACRO, London, UK). These were set up in the 10 months before the trial started.
All research and pharmacy staff undertook good clinical practice (GCP) training. This was arranged by King’s Health Partners Clinical Trials Office (KHP CTO) or by NHS Forth Valley. Some pharmacy staff attended a GCP pharmacy training tailored for pharmacy staff and facilitated by KHP CTO. In addition, a GCP light training was provided by KHP CTO on site for health-care staff. Those health-care staff who were unable to attend the training owing to shift patterns were able to review GCP slides and complete a self-certificate. The GCP training was reviewed continuously throughout the trial to ensure that all new staff were trained and that they were added to the delegation log. All staff were retrained by KHP CTO or by NHS Forth Valley after 2 years. Curricula vitae for all health-care staff were collected and updated every 2 years, as per guidelines.
In September 2016, training was provided to all research staff on the standard operating procedure for the project and for the following measures: DIVA, MINI, ZAN-BPD, CAARS-O, CGI scale and WRAADDS. Subsequent new staff received in-house training or attended training events.
The trial was monitored by the KHP CTO on behalf of the sponsor. KHP CTO deemed the sites ready for recruitment with an e-mail confirming ‘green light’, which was obtained for HMP YOI Isis on 11 October 2016 and for HM YOI Polmont on 3 November 2016. An initial site initiation visit for Polmont took place on 13 October 2016, but there was a delay in obtaining ‘green light’ status because of major information technology (IT) issues in King’s College London, which caused a delay in accessing certain records required for the project. Seven participants had to be randomised ‘manually’ with the assistance of the KCTU as a result of the IT issues.
At HM YOI Polmont, there was no routine service for identifying young prisoners with ADHD. For this reason, a two-stage consent process was developed to avoid the situation of asking participants to sign consent for a trial for which they would not be eligible. The first consent (consent I) was for screening using the Barkley ADHD self-assessment form, followed by a DIVA 2.0 assessment for those who screened positive with the self-assessment form. Consent II was the main trial consent, required prior to collection of further baseline data and randomisation. At HMP YOI Isis, there was already an established ADHD service set up by the chief investigator, Philip Asherson. This site also ran the pilot study and screening for ADHD had been established as part of routine clinical care. This meant that, for this site, there was a pre cohort that had already been screened prior to trial commencement. The pre cohort participants who had already completed the Barkley ADHD self-assessment and had screened positive were invited to sign consent I (for the screening and diagnostic stage of the trial) if they attended a DIVA 2.0 assessment for ADHD.
In June 2015, the Scottish Government published a youth justice strategy82 with the aim of reducing the number of young people sent to prison by establishing an early and effective intervention that provided more community options. This meant that the prison population at HM YOI Polmont was less than half of its capacity at the start of the trial, thereby affecting the expected recruitment pool for the trial. This resulted in uneven recruitment between sites.
In March 2017, we were informed that the potential third site, Her Majesty’s Prison (HMP) Rochester, would close in November 2017 for refurbishment. It was therefore decided that this site would not be launched. Owing to the recruitment targets being met, it was decided not to set up an alternative third site. The planned refurbishment at HMP Rochester was later postponed.
Accountability for trial medication
All aspects of treatment and accountability for managing the medication storage and delivery were managed locally by the prison pharmacies and health-care teams, as per standard practice for this medication in the prisons. IMP accountability was recorded and verified. All aspects of treatment compliance and recording of treatment administration/refusal were managed by the prison mental health teams and locally by health-care staff, as per standard practice for these sites. Participants were observed when they were given medication and were checked to ensure that the capsules had been swallowed. This information was then recorded (signed off) by the nursing staff who had delivered the medication on prison pharmacy record sheets or digital records. Despite precautions taken by nurses with regards to seeing prisoners swallow medication, there are documented cases when prescribed medication was diverted. Although there were no such cases documented for the trial medication in this trial, it cannot be ruled out.
Safety checks
Participants were monitored daily by the prison mental health and health-care teams. Safety checks were conducted in line with NICE guidelines.13
With regard to the research procedures of the trial, there was little risk to participant safety. Participants were aware that they could withdraw from the trial if they wished to do so. Participants who become upset or distressed were offered support by the research team and/or the prison mental health team.
The health-care team followed national guidelines on safety,13 which are predominantly related to the monitoring of cardiovascular function. More specifically, the clinical care followed the following procedures:
- checks on heart rate and blood pressure before commencing treatments, and review of health-care records
- potential cardiovascular abnormalities evaluated for risk
- heart rate and blood pressure checked once a week for the first 5 weeks and at the end of the 8-week trial.
Other safety checks included the monitoring of AEs during assessments. In addition, participants were monitored daily by prison staff and any potential AEs were reported to the prison health-care team.
Procedures for recording and reporting adverse events
Safety remained the responsibility of the prison mental health care team. AEs of any medical or non-medical intervention were identified or recorded by the research team at each site, and were verified by the clinician who was part of the research team, an assigned medical colleague at specialist registrar grade or above who was a member of the prison health-care team, or the clinical lead for the project. The decision to stop treatment following an AE remained the responsibility of the clinical team. Minor AEs that did not fall under official reporting procedures were reported to the clinical team, for example sleep disturbance, minor levels of anxiety or dysthymia, small increases in heart rate and blood pressure, reduced appetite and other minor physical symptoms that did not endanger patients or cause more than minor distress. All other AEs from medication were recorded and reported in line with The Medicines for Human Use (Clinical Trials) Regulations 200483 and Amended Regulations 2006.84
The research team, acting on behalf of King’s College London as sponsors, delegated the delivery of the sponsor’s responsibility for pharmacovigilance [as defined in Regulation 5 of the Medicines for Human Use (Clinical Trials) Regulations 200483] to the KHP CTO. Reporting of serious adverse events (SAEs) continued until the last participant’s final dose had been completed. For each participant, AEs were recorded from randomisation and reporting of AEs was from the time of first dose of the trial medication to the end of their involvement in the trial (last dose at the end of 8 weeks). All SAEs, serious adverse reactions (SARs), suspected unexpected serious adverse reactions (SUSARs) and important medical events (IMEs) (except those specified in the protocol as not requiring reporting) were reported immediately by the chief investigator or designated site investigators to the KHP CTO, in accordance with the current pharmacovigilance policy. Janssen-Cilag Ltd were notified at the same time.
The KHP CTO reported SUSARs and other SARs to the MHRA, who are competent authorities of other European Economic Area states in which the trial is taking place. The chief investigator reported to the relevant ethics committees. In this trial, there were no SUSARs or SARs, and only one SAE, identified as an IME.
Reporting timelines were as follows:
- SUSARs that are fatal or life-threatening must be reported no later than 7 days after the sponsor is first aware of the reaction. Any additional relevant information must be reported within a further 8 days.
- SUSARs that are not fatal or life-threatening must be reported within 15 days of the sponsor first becoming aware of the reaction.
The chief investigator and KHP CTO (on behalf of the co-sponsors) submitted a Development Safety Update Report relating to this trial IMP to the MHRA and REC annually.
Trial Steering Committee
A TSC was convened to provide overall supervision of the trial and ensure that the trial was conducted to the rigorous standards set out in the Medical Research Council guidelines for GCP.85 The TSC monitored progress, adherence and safety. The TSC chairperson was Professor Jenny Shaw (Consultant Forensic Psychiatrist, University of Manchester); the other TSC members were Dr Ylva Ginsberg (Consultant Psychiatrist specialising in ADHD in prisoners, Stockholm, Sweden), Dr Peter Mason (Forensic Psychiatrist and specialist in ADHD, Cheshire and Wirral), Mr Anthony Davis (R&D Manager, Oxleas NHS Foundation Trust, London), Dr Ulrich Müller-Sedgwick (Barnet, Enfield and Haringey Mental Health NHS Trust, London) and Mrs Beverley Nolker (Prison Officers’ Association Learning, London, and user representative). Other user representatives were included, but had to be replaced during the trial. Non-independent members were lead applicants in London and Edinburgh (Philip Asherson and Lindsay Thomson). Other members of the TMG attended TSC meetings as observers and to report to the TSC. During the trial, four TSC meetings were conducted, all face to face, and all members of the TMG were able to attend these meetings.
Data Monitoring Committee
A DMC was convened to monitor the safety of participants, the ethical conduct of the trial and the quality of the data. The committee consisted of three members experienced in clinical trials, including an independent statistician. The DMC chairperson was Professor Seena Fazel, University of Oxford (an experienced Forensic Psychiatrist). Other members were Professor Chris Hollis, University of Nottingham (an expert on the Clinical Management of ADHD); and Adrian Cook (a trial statistician). DMC meetings were timed to occur prior to TSC meetings so that they could report to the TSC. There were four DMC meetings during the trial. The first two were face-to-face meetings, which the majority of members attended, and the last two were by telephone conference.
Trial management
The project was led by Professor Philip Asherson in London, supported by a trial manager. The trial manager liaised with the trial monitors and ethics board when required and supported the completion and ongoing management of the project. The research psychiatrist at HMP YOI Isis co-ordinated all daily activities on site. Principal investigator Lindsay Thomson led the project in Edinburgh and was supported by the local research assistant, who conducted similar day-to-day co-ordinating tasks in HM YOI Polmont. Weekly telephone conference calls were held throughout the trial with Philip Asherson, Lindsay Thomson, the trial manager and the on-site clinical research teams (psychiatrists and research assistants) to allow for discussions regarding recruitment, diagnosis and other procedures being followed by the clinical teams.
A TMG was established to review AEs and reporting, to review the progress of the trial, to review and report on file notes and potential protocol violations, and to discuss changes to the protocol. This group, led by the chief investigator, Philip Asherson, consisted of Lindsay Thomson, the statisticians, the trial manager and the clinical teams. Other co-applicants were invited to these meetings and occasionally attended. During the initial stages of the trial, the TMG met at the same time as the TSC. After the first two TMG meetings, the frequency of these meetings was increased to monthly (from December 2017, when 94 participants had been randomised to the trial) to facilitate a greater level of ongoing monitoring of the trial. Prior to these regular monthly TMG meetings, informal meetings between the trial manager (LJ) and, separately, with the trial statistician (RH) and with the clinical research associate (CRA) monitoring the trial on behalf of the sponsor were held, file notes were reviewed, and protocol deviations were discussed and documented as required.
Ethics issues specific to this project
The trial was conducted in compliance with the principles of the Declaration of Helsinki (1996),86 the principles of GCP87 and in accordance with all applicable regulatory requirements, including but not limited to the Research Governance Framework and the Medicines for Human Use (Clinical Trials) Regulations 2004,83 as amended in 200684 and any subsequent amendments. The protocol and related documents were submitted for review to the East of England–Essex REC and to the MHRA for clinical trial authorisation. Annual progress and safety reports were completed, and a final report was submitted to the KHP CTO (on behalf of the sponsor), the REC and the MHRA within the timelines defined in the regulations.
Osmotic release oral system-MPH is licensed for first time use in young people with ADHD and severe impairment under aged 18 years only, although NICE also recommends OROS-MPH as a first-line treatment for ADHD in adults. The 8-week trial included a placebo group, so we were denying this group a recommended treatment for ADHD during this period. However, currently, prisoners with ADHD are rarely treated because of uncertainty regarding the validity of the ADHD diagnosis, efficacy of treatment, and concerns about potential drug abuse and diversion in prison populations. To address the issue of unequal access to treatment, we offered treatment to all participants once the trial was completed. Care was taken to ensure that no coercion was involved in recruiting prisoners to the trial. It was made clear that taking part in the trial would not have an effect in either a negative or positive way on their time in the prison or the prison regime.
Some participants might benefit from treatment that was offered as part of the clinical trial.
Participants were informed that their anonymised research data from this trial will be stored securely and may be shared with other scientists or research groups if this helps understanding of the trial findings. The data may also be used in combination with data from other similar studies. All personal information was stored in a secure place (a locked cabinet in a locked office in the prisons) and not allowed to leave the prison. Outside of the immediate health-care team involved in this research, no one will be able to match personal information (name and prison number) with the information gathered for research. Personal details were not linked to clinical or prison records in the research database. After completion of the trial, personal identifiable data relating to the trial will remain in the prison and be destroyed in accordance with GCP guidelines, and after authorisation from KHP CTO, on behalf of the sponsor. To achieve anonymity, each participant had an allocated identifiable number accessible to the local research team only.
Quality assurance
Monitoring of this trial to ensure compliance with GCP and scientific integrity was managed and oversight was retained by the KHP CTO Quality Team.
Data handling
The chief investigator, Philip Asherson, acts as custodian for the trial data. Data were stored on a database set up by KCTU. Patient data were pseudo-anonymised. All personal identification information and research identifier codes were stored on a password-protected computer by the clinical research teams. All data were stored, handled, processed and archived in line with the Data Protection Act 199888 and the Medicines for Human Use (Clinical Trials) Amended Regulations (2006).84
Data management
Data were stored on a MACRO system trials database (Elsevier UK, London, UK), set up and managed by the KCTU. This allowed for full audit information and for checks on data entry that were used to ensure the integrity of the data collection and monitoring of the trial progress. At the end of the trial, all research records were transferred by secure courier service to the relevant NHS trust R&D department for archiving. The Trial Master File will be archived on behalf of the sponsor using Iron Mountain Inc. (Boston, MA, USA).
Researchers had unique log-in access to the trial database for their site. The system was designed to allow users to register participants and then collect and track data and manage their progress throughout the trial. The data entry system had restricted answer options, including values for ‘not applicable’, ‘not done’ and ‘unknown’. Any data field was tested for valid dates. The data entry system had pre-set minimum or maximum values on all questions requiring numerical values; anything outside of that range was flagged to the user to ensure that the value entered was correct. Data were collected on hard copies from various sources, such as participants’ records and prison and medical records. At the end of the trial, for each participant, the researcher conducted a SDV checklist to ensure that all the data entered had been entered correctly by double-checking the paper copies and any other source data, such as electronic medical and prison records. This was then signed and dated by the researcher and sent to the trial manager.
The trial manager and CRA had read-only access, but were able to raise SDV or discrepancies (data clarification request). The trial statistician conducted high-level checks after data extractions and any data checks needed were sent to the trial manager. The CRA monitored the trial on behalf of KHP CTO and 10% of participants had all of their research data checked against source data. In addition, the primary end point (CAARS-O score) was verified at 100% for all participants.
Patient and public involvement
Valuable feedback was given by participants in the pilot study and by prison staff. We also had positive feedback from Her Majesty’s Inspectorate of Prisons regarding the pilot trial.89 Throughout the set-up and duration of the trial, we received advice from service users with ADHD or professionals working with prisoners. Special attention was paid to ensure that the information sheet and consent forms were written in appropriate language, as it was expected that there would be low levels of reading ability in this setting.
We lost contact with one service user member of the TSC during the trial, who was replaced by a service user with experience of both ADHD and prisons. Further insights were also given by Beverley Nolker, a service user who works for Prison Officers’ Association Learning. We had hoped to more directly include additional service users with both experience of ADHD and prisons; however, it was difficult to identify suitable individuals with the relevant experiences and to maintain their involvement and engagement.
We worked closely with Dr Kai Syng Tan, a King’s College London artist in residence with lived experience of ADHD. Kai visited HMP YOI Isis during the trial to understand the prison environment and what might motivate participants to engage in the trial. Following this, Kai generated images that emphasised choice, control, autonomy, self-care, self-respect, and, at the same time, was mindful of the stigma attached to ADHD. She designed an image that was used during the trial to remind participants of medication times for all their medications. The image was produced in the form of a flyer that prisoners could stick on the walls of their cells or use as a coaster.
At the end of the project dissemination took place at the Great Hall, Strand Campus, King’s College London. The event, attended by around 100 delegates, focused on the prevalence of ADHD in prisons, ADHD and comorbidities, what research has been done to further the management of ADHD in prisons and what ADHD services in prisons should look like. The event was free and open to the public, and generated a lot of interest from professionals in ADHD services and forensic settings and from members of the public.
Two consensus meetings, leading to publications, were organised and led by co-applicant Susan Young before and during the trial.21,49 These included multidisciplinary groups of health-care and criminal justice system professionals and service users across sectors relevant to the management of children, adolescent and adult offenders. The initial meeting in November 2009 provided an overview of recommendations for ADHD services in the criminal justice system, and was the foundation for setting up the ADHD service at HMP YOI Isis, and the subsequent pilot study for this trial. The second meeting, in November 2016, provided an updated consensus opinion and recommendations for the identification and treatment of offenders with ADHD in the prison population.