We consider there to be four main categories of CRIS stakeholder: (1) those whose data is contained in the system, namely SLaM patients, (2) those who care for SLaM patients, (3) those who use the system for research, audits, and service evaluations including academic researchers and NHS staff, and (4) the general public. In terms of prioritisation, SLaM patients and carers are considered the first priority for CRIS communications activities, this is to ensure that SLaM is meeting transparency requirements under the UK GDPR 2018.

SLaM has produced leaflets and posters targeting patients and the public to inform them of CRIS activity, including a CAMHS leaflet designed specifically for young people and their carers. These are displayed in SLaM waiting areas and team bases and represent the first point of contact that many patients will have with CRIS. This publicity material explains the purpose of processing deidentified clinical data for research and provides information on how people can find out more about CRIS, including how it has been used in data linkages. In 2023, the CRIS Team in collaboration with a mental health service user and the CRIS Oversight Committee undertook a project to update and redesign the leaflets and poster. The updated material includes new features such as a QR code to enable service users to more easily access further information on CRIS.

The CRIS leaflets and poster are one of several mandatory communication materials which SLaM services are required to display. Audits by SLaM’s Information Governance (IG) Team ensure that services are displaying the required materials appropriately. Leaflets and posters are translated or produced in large print as required.

To comply with Articles 13 and 14 of the UK GDPR 2018, SLaM publishes a privacy notice [25] which includes concise information on how personal data are used within SLaM. This privacy notice includes information on CRIS and how health data are used for research as well as contact details for patients who wish to opt-out of their data being used within CRIS.

The CRIS webpages on the NIHR Maudsley BRC website [26] aim to be the main source of information for SLaM patients, academics, SLaM staff, and the general public who would like to find out more about CRIS. The webpages contain information about the history of CRIS, the CRIS Team, outputs from CRIS research such as publications, CRIS data linkages, and NLP algorithms as a resource. It is also a place to find out how to contact the CRIS Team and get involved with CRIS work, as well as information on opting out for SLaM service-users who do not want their information included in CRIS. The CRIS webpages are also the main source of information available to academic users prior to meeting with a member of the CRIS Team and is a large component of the wider CRIS communications plan which aims to ensure that all potential CRIS stakeholders can access relevant information about CRIS in several different formats.

In 2023, SLaM was awarded a grant from Health Data Research UK (HDR UK) to undertake a website review project to update the website in order to better meet the HDR UK Transparency Standards [27]. The Transparency Standards were co-developed by the Pan-UK Data Governance Steering Group and HDR UK’s Public Advisory Board, with the aim of improving the transparency of processes for accessing health data for research. These standards were designed to guide best practice for transparency around how data are accessed and the information available to researchers and members of the public. This includes recommendations around transparency of application processes and criteria, website navigation, appropriate use of language based on target audience, and improving how organisations communicate the impact of health data research to improve and save lives.

The project involved both a lay review of the website conducted with five adult mental health service users, both patients and carers, as well as an academic review by two early career researchers. The review suggested several improvements to the website, such as adding more visual content, improving website navigation, and making the text and design more accessible. Feedback was reviewed and compared against the Transparency Standards. Changes were then made to webpages, including updating images, adding additional information, and making accessibility changes. Patient consultees were sent a further online questionnaire to provide feedback on the updated pages which was positive [28].

The CRIS Team holds regular drop-in sessions designed for anyone who would like to find out more about CRIS, this includes both academic users and SLaM patients. These sessions are widely publicised via the CRIS website as well as through SLaM, NIHR Maudsley BRC, and Institute of Psychiatry, Psychology and Neuroscience (IOPPN), King’s College London (KCL) communications channels. Including, for example, the weekly SLaM staff e-newsletter, NIHR Maudsley BRC social media channels, the CRIS newsletter, and physical posters.

Since initial development, CRIS has been presented at many different events, aimed at all CRIS stakeholders. Examples include the annual SLaM staff conference, department-specific patient and carer groups, and department-specific clinical teams as requested. We continue to seek opportunities to present CRIS and discuss our work through various SLaM and NIHR Maudsley BRC events and meetings.

Guided by the FAIR principles [15] to support the findability of CRIS as a resource, CRIS is included in the Catalogue of Mental Health Measures [29]. Designed to provide easy access to information about the mental health measures included in British cohort and longitudinal studies, the Catalogue of Mental Health Measures aims to facilitate mental health research by maximising the use of existing data. Further to this, the Catalogue also interfaces with the HDR UK Innovation Gateway [30]. Established in 2020, the Gateway acts as a common entry point for researchers to discover and request access to UK health-related datasets.

Due to the strict governance surrounding the use of CRIS [17], data cannot be shared via open access repositories. Therefore, to support the accessibility of the data, the CRIS Team have produced guidance to support researchers with applying to use and accessing CRIS data. To ensure that the process for obtaining access is transparent, the link to the online application form and guidance notes, as well as a detailed ‘Welcome to CRIS Guide’ are made available via the public website on a dedicated Information for Researchers page [31]. The CRIS Security Model is also made available, this details the process by which CRIS applications are reviewed and approved by the patient-led CRIS Oversight Committee [24].

All projects approved by the patient-led CRIS Oversight Committee are publicly available to view via the CRIS website [32]. The CRIS projects database includes the name of the lead applicant, the project title, lay summary, and approval date. Users can search on any of these fields to explore what research is being conducted using CRIS.

The use of health data by commercial agencies is considered particularly controversial, in part due to the lower levels of trust that the public has in the private sector in comparison with public section organisations [11]. The potential use of CRIS in commercial partnerships is included in the CRIS information leaflet and all existing use of CRIS in commercial partnerships is transparent and documented in detail on the CRIS website [33]. For industry partnerships, the preferred protocol has been for data to be accessed by university or SLaM academic staff and the findings generated according to the specifications of the commercial partner, with no direct access to data by employees of the commercial partners themselves. An example of the type of project conducted using CRIS in a commercial partnership includes the real-world evaluation of the effectiveness of existing practice [34]. The above model whereby data is accessed and analysed by academic staff helps to maintain the independence of the results as well as ensuring that projects meet the conditions of the CRIS Security Model [17] and there have been no concerns raised or complaints to date.

PPI groups have been set up to support the ongoing use of CRIS and to provide advice and recommendations for projects using CRIS data. The Data Linkage Service User and Carer Advisory Group (DL-SUCAG) provides advice and feedback to researchers conducting projects involving data linkage [35]. Launched in January 2024, the Clinical Informatics Service (CIS) Advisory Group aims to identify the priorities of SLaM patients and carers for how clinical health record data should be used for research and informing patient care as well as advising on NLP development and application. The CRIS for Dementia Service User & Carer Advisory Group (C4D-SUCAG) ensures that dementia research conducted using CRIS is led by the concerns identified by people with dementia and their families. The CRIS PPI groups facilitate meaningful collaboration between SLaM patients and the researchers using CRIS data and an evaluation of the DL-SUCAG in 2019 found that both researchers and patients valued the experience of attending the group.

Targeted PPI relating to specific data linkages with CRIS is also undertaken. A recent example includes a podcast series highlighting the data linkage between CRIS and the 2011 UK Census data from the Office for National Statistics (ONS) [22]. The four-part series focusing on the data linkage and initial findings from the study as well as discussions around tackling systemic mental health inequalities and the future possibilities of mental health data, is widely available online [36] as well as on all major podcasting applications, including Apple Podcasts and Spotify.

Another example of transparency activities related to data linkage includes the project-specific leaflets and poster developed for the linkage between CRIS and the HIV and AIDS Reporting System (HARS) [37]. In collaboration with clinical colleagues, these leaflets were distributed to specific teams within SLaM who specifically support patients with HIV or who are more likely to come in to contact with patients with HIV. The distribution of the leaflets led to greater awareness of the research project which in turn led to further transparency opportunities including a presentation to both patients and clinical staff at an event to celebrate National HIV Testing Week.

A list of all publications resulting from CRIS projects is publicly available on the CRIS website [38]. Researchers publishing data from CRIS are encouraged to do so in Open Access journals or to save the final version in an Open Access repository so that they are accessible to a wider audience.

An academic seminar is held monthly where CRIS researchers are invited to present their work. Sessions are open to all and are advertised to both CRIS academic users [31] and SLaM patients and the public [39] via the CRIS website. These sessions are held in a hybrid format to enable attendance both virtually and in person.

A recent communications project saw researchers discussing their CRIS research as part of a video campaign. The six CRIS Research Insights videos [40] showcase some of the work being carried out by academics using CRIS. The launch of the videos in May 2024, supported by the NIHR Maudsley BRC Communications Team, included a comprehensive social media campaign across Facebook, X (formally Twitter), and LinkedIn. The campaign led to 11,300 impressions and 305 engagements on X, and 7,000 impressions and 2,600 video views on LinkedIn. In total, the social media campaign resulted in an additional 1,600 visits to the CRIS webpages.

Launched in March 2024, the CRIS newsletter is published every quarter and sent to all CRIS academic users as well as individuals who have chosen to sign up through the link on the public CRIS website. It contains information on recent CRIS updates as well as highlighting important CRIS research findings and news. The newsletter is currently delivered to over 550 email addresses and the second edition, published in July 2024, was opened by 206 (37%) recipients and links within the newsletter were clicked by 28 (7%) recipients.

To support the interoperability and reusability of CRIS data, all data extractions conducted by the CRIS Team are done according to project-specific data extraction specifications. These specifications define the cohort of interest and the output variables required for each project. Although one of the largest mental health Trusts in Europe, SLaM’s catchment area only covers southeast London and limited national services, it is therefore important that findings from CRIS studies can be replicated at other sites to ensure reliability and generalisability of the results. Despite different mental health Trusts using different electronic patient record systems, meaning that code to extract data cannot always practically be applied to other sites, the data extraction specifications can be shared to enable equivalent datasets to be extracted across different systems. An example of this in practice was a project conducted during the COVID-19 pandemic where equivalent data were extracted across ten different UK mental health service providers which found that mental health services underwent profound changes during the first national lockdown beginning in March 2020 [4]. Other examples of multi-site collaborative research include investigations of crisis interventions to prevent readmission to acute mental healthcare [41], ethnic group differences in characteristics at dementia diagnosis [42], clozapine side effects [43], crisis presentations associated with social media exposures [44], antidepressant pharmacotherapy and dementia progression [45], and a recent estimate of demand for future Alzheimer’s disease therapies [46].

To further support reusability, a Cloud service has been developed to enable the sharing of NLP applications developed using CRIS with other mental health Trusts. The Mental Health Text Analytics Cloud (MH-TAC) is a prototype cloud-based platform set up within the SLaM infrastructure that allows other mental health Trusts to upload text files into secure domains where they can be processed using previously developed and validated NLP algorithms. The NLP-derived meta-data are then returned to users.

There are four categories of CRIS stakeholder which CRIS transparency activities aim to target: SLaM patients, carers, users of the resource, and the general public. One of the main challenges of ensuring comprehensive transparency is ensuring that communications activities are varied enough to reach these diverse groups of stakeholders.

With regards to patients and carers specifically, it can be challenging targeting harder-to-reach groups; this is particularly important as a major advantage of health records databases is that they provide better representation for these populations than traditional recruitment-based (and, hence, inevitably selected) research cohorts. SLaM services support patients throughout the lifespan, including patients with varying levels of capacity. CRIS communication activities need to ensure that they reach all these different groups. Current communication activities are varied to try and reach as many people as possible and include both physical leaflets and posters as well as online resources such as the CRIS webpages and video campaign. Regardless, we acknowledge that despite our best efforts our communications activities may not reach all patient groups.

Through the recent review of the CRIS webpages [28], we found that different people like to take in information in different formats, for example text, pictures, or videos. Throughout the project, there were challenges involved in designing a website that was informative to both lay public readers and potential research users. We learnt that providing dedicated pages for patients and academic users can help to tailor relevant information to different stakeholders.

Data science is challenging, and reports suggest that the general public have limited knowledge about data and the regulations surrounding it [47]. It can therefore be difficult to obtain meaningful engagement from patients and the public without requiring further training. Although, it can be challenging to find a balance between increasing awareness without coercing viewpoints. However, in our experience with our patient and carer advisory groups, providing a basic level of training enabled members to engage effectively and was well received by group members [35]. We have also engaged with patients and carers in the development of our leaflets, posters, and website content to ensure that the content is pitched at a level that is understandable by the general public. Where possible we have also tailored content, for example by having a CAMHS specific CRIS leaflet.

Working with data from electronic medical records can reduce researchers’ opportunities to engage with participants compared to what would be built into projects involving traditional recruitment and data collection. The transparency activities across CRIS, such as the PPI groups and engagement with the patient-led CRIS Oversight Committee, give researchers the opportunity to engage with patients and receive feedback on projects which can have a valuable impact on their research whilst also supporting transparency in the use of the CRIS data resource.

This paper describes the activities undertaken at the NIHR Maudsley BRC CRIS mental health case register, to incorporate transparency along the path from data collection to data use and through to dissemination of research findings. We believe the material presented in this paper will be of use to other data providers looking to foster transparency in their data resources.