Clinical Record Interactive Search (CRIS) provides a means of analysing de-identified data from the Oxford Health NHS Foundation Trust (OHFT) electronic patient records. CRIS unlocks and transforms clinical data held in Trust systems to provide a rich and pseudonymised resource allowing researchers and clinicians to investigate hypotheses and identify patient cohorts. CRIS also provides an invaluable tool for service evaluation and clinical audit.
Oxford Health NHS Foundation Trust are also part of the CRIS Network. This is a group of highly innovative NHS Mental Health Trusts who are working together to accelerate research work in dementia and mental health. For further information on this network and its members please see the CRIS Network website.
Access to clinical information is clearly a sensitive issue and a security model was developed which has been considered and approved by the Oxford Health NHS Foundation Trust Caldicott Guardian and the Trust Executive.
Trust Electronic Patient Record (EPR) systems hold a wealth of useful patient data. CRIS transforms this information into a pseudonymised database appropriate for research, service evaluation and clinical audit use.
The transformation of data enables CRIS users to access the wealth of information recorded during patient care by healthcare professionals. This includes data recorded in coded and structured form, for example dates and numbers, plus data held in unstructured free text form, for example, within written assessments, progress notes and correspondence. After transformation, free text data can constitute a large part of the content of information held within CRIS, although this will depend on the amount of unstructured data held in the source EPR system.
The transformation process uses the patient’s EPR system identifier to derive a unique ID for each patient in the database. This ID does not allow CRIS users to identify patients. However, where patients have given appropriate consent, the ID can be used by authorised personnel to contact patients who have been identified as potential recruits to an ethics approved research project.
Access to CRIS requires either an NHS contract or a Research Passport.
If you have an NHS contract that is not with Oxford Health please contact the HR department within your own Trust and ask them to prepare an NHS to NHS Proforma.
Once your right to access CRIS has been approved you can register your CRIS project.
To register your CRIS research project please download the research registration form and email your completed application to firstname.lastname@example.org. A member of the CRIS team will then contact you to establish the feasibility of your CRIS question. To assist you in determining your approach to interrogating the data we have produced a sample data plan which can be found here.
For a service evaluation or clinical audit, please visit the Trust Clinical Audit pages to download the relevant application form. To determine whether your project is Research, Service Evaluation or Clinical Audit please click here.
Additional information regarding the suggested content for the ‘Summary, Research and Scientific Justification’ fields, please see here for further information.
Once your project has passed feasibility your project will be formally registered on the CRIS system by a member of the CRIS team. Your project will then be submitted to the CRIS Oversight Group to seek their approval.
Access to the CRIS facility is via a secure private network on a research platform with UKSeRP. Once your right to access CRIS has been approved the CRIS Admin team will set up your UKSeRP user account. This will enable you to apply for a CRIS user account, access to the CRIS facility along with statistical analysis software and your project files.
In addition to a UKSeRP user account, see above, a CRIS user account will also be required by all project lead CRIS researchers. This will enable the formal registration of your CRIS project to the CRIS system.
To obtain a CRIS user account, log on to UKSeRP and access CRIS via the UK-CRIS desktop icon. From the CRIS home screen select ‘Register’ to submit a user application, which will be approved by a member of the CRIS Admin team. The CRIS Admin team will register your CRIS project once it has passed feasibility and you will automatically be added as a project user.
A training package which provides information regarding the navigation of CRIS and how to conduct a basic and advanced search is available via the CRIS Network website which includes forums, useful documentation and information relating to the CRIS Network. It may be accessed from any computer. All CRIS Users should register for a CRIS Network account which will then allow them to view the training documentation.
Once registered you can log in and access the CRIS user training here.
CRIS can only be accessed from the OHFT network. Data from CRIS must be kept within the OHFT firewall and can only be saved on the CRIS shared drive on OHFT computers. CRIS data CANNOT be saved on personal or encrypted USB sticks. CRIS data must not be emailed from OHFT machines to your personal email or a University email. Please be aware that additional permission is required for derived data to be analysed outside of the OHFT firewall, from the CRIS Oversight Group. CRIS researchers are only permitted to access their own CRIS Project folders and files located either within the CRIS Remote Desktop or the G:\Oxfordshire\CRIS Projects. Access is not permitted to other projects folders and files. Please also note that currently SPSS and R are the only statistical programmes that are available within the Research and Development Team.
The security model includes regular audits of searches carried out using CRIS (all searches by all users are recorded and can be audited). For this to be possible, we keep a record of all projects carried out involving CRIS analysis along with general specification of the type of searches which will be required.
Ethics and research governance approval assume anonymity of the data analysed. As with any dataset, there is the potential within this database to compromise anonymity by generating unique variable combinations or rare categories. CRIS users are asked to consider whether this issue may occur and strategies to avoid compromising anonymity. Alternatively they may wish to obtain specific ethics approval for an analysis where this risk is likely to be significant. Searching under clinicians’ names are a sensitive issue; this type of information in research would have to be justified.
The CRIS Oversight Group, led by the Trust Caldicott Guardian will review all requests to use CRIS as a de-identified database. It is important for the Trust to demonstrate that OHFT clinical data are used responsibly and for projects with demonstrable research and clinical importance.
The future of CRIS, as with other aspects of OHFT research, depends on successful bids for future funding. This in turn requires evidence of use of the database, hence the need to keep a record of individual projects.
The CRIS Oversight Group has a role in facilitating CRIS analyses and to advise on how best to extract robust data. The database is potentially complex and users will be encouraged to collaborate and share expertise and hands-on experience. The information submitted in the project application form will be used to provide a database for this purpose to assist future researchers.
All work, which requires the use of CRIS must be reviewed and approved by the CRIS Oversight Group. This should be taken in to account when completing any applications or considering a collaboration, which require the use of CRIS data. Failure to inform the CRIS Oversight Group of a grant, research application or collaboration may result in the project being declined.
If you have any questions please email CRIS.email@example.com
Below are links to the Top 10 priorities for future research in mental health, agreed by the Priority Setting Partnerships (PSP). Why not consider conducting a CRIS search to answer one of these questions.
James Lind brings patients, carers and clinicians together in Priority Setting Partnerships (PSPs) to identify and prioritise the Top 10 uncertainties, or unanswered questions, about the effects of treatments.
The aim of this is to make sure that health research funders are aware of the issues that matter most to patients and clinicians.
How are patient details protected?
CRIS information is held securely with strict arrangements about who can access the information. This will include Trust staff, clinicians and approved researchers. The information will only be used for the purpose of health and care research, service evaluation, clinical audit or to contact you about future opportunities to participate in research.
Where there is a risk that you can be identified your data will only be used in research that has been independently reviewed by an ethics committee. For further information on ‘patient information and health and care research’ see the Health Research Authority (HRA) or for CRIS specific information see our CRIS Privacy Notice
Below find a list of publications related to CRIS:
Breilmann J, Girlanda F, Guaiana G, Barbui C, Cipriani A, Castellazzi M, et al. Benzodiazepines versus placebo for panic disorder in adults. The Cochrane database of systematic reviews [Internet]. 2019;3(3):CD010677.
Gligic, L., Kormilitzin, A., Goldberg, P. and Nevado-Holgado, A. (2020). Named entity recognition in electronic health records using transfer learning bootstrapped Neural Networks. Neural Networks, [online] 121, pp.132–139.
Goodday SM, Kormilitzin A, Vaci N, Liu Q, Cipriani A, Smith T, et al. Maximizing the use of social and behavioural information from secondary care mental health electronic health records. Journal of Biomedical Informatics. 2020 Jul;107:103429.
Hofer, M., Kormilitzin, A., Goldberg, P. and Nevado-Holgado, A. (n.d.). Few-shot Learning for Named Entity Recognition in Medical Text.
Iqbal E, Govind R, Romero A, Dzahini O, Broadbent M, Stewart R, et al. The side effect profile of Clozapine in real world data of three large mental hospitals. arXiv:200109698 [cs, stat] [Internet]. 2020 Jan 27
McDonald K, Smith T, Broadbent M, Patel R, Geddes JR, Saunders KEA. Prevalence and incidence of clinical outcomes in patients presenting to secondary mental health care with mood instability and sleep disturbance. European Psychiatry. 2020;63(1).
Senior M, Burghart M, Yu R, Kormilitzin A, Liu Q, Vaci N, et al. Identifying Predictors of Suicide in Severe Mental Illness: A Feasibility Study of a Clinical Prediction Rule (Oxford Mental Illness and Suicide Tool or OxMIS). Frontiers in Psychiatry. 2020 Apr 15;11.
Vaci N, Liu Q, Kormilitzin A, Crescenzo FD, Kurtulmus A, Harvey J, et al. Natural language processing for structuring clinical text data on depression using UK-CRIS. Evidence-Based Mental Health [Internet]. 2020 Feb 1;23(1):21–26.
Walker, S., Potts, J., Martos, L., Barrera, A., Hancock, M., Bell, S., Geddes, J., Cipriani, A. and Henshall, C. (2019). Consent to discuss participation in research: a pilot study. Evidence Based Mental Health, [online] p.ebmental-2019-300116.
This is a prescribed app. It should only be used alongside a face to face intervention provided by a mental health worker. Check with your local service to see if they subscribe to the app.