BRISSKit: Connecting researchers with clinical data

7 November, 2012

The UK government is starting to recognise the value of data to the economy and is developing policy that aims to realise that worth. One of the largest and, potentially, most useful repositories of information in the country is the NHS, housing vast quantities of patient data with myriad applications. Government policy is beginning to increase the amount of clinical data that is exposed to biomedical researchers; however, the difficulty of accessing information whilst respecting the boundaries of patient privacy is hampering vital progress.

These are the issues being addressed by the University of Leicester’s Biomedical Research Infrastructure Software Service (BRISSKit) project team. The BRISSKit is an integrated, modular, open-source platform providing bench-to-bedside support for clinicians and researchers working with biomedical data. The components enable connection of electronic patient records with clinical data and tissue sample handling data while the database architecture allows anonymised versions of sensitive data to be hosted for university researchers. Instances of the platform can be run locally or from the cloud, providing the system with flexibility and scalability.

The BRISSKit has been developed in collaboration between Leicester University and the Glenfield Hospital Cardiovascular Biomedical Research Unit (BRU) and is now well enough established to roll out pilot studies at UCL’s Faculty of Health Sciences, the Birmingham Cancer Studies Group, Leicester Respiratory and Diabetes BRUs and the University of Loughborough Lifestyle BRU.

Currently, the platform hosts implementations of four open-source tools:

  • CiviCRM – Constituency relationship management software used to handle participant and study management. Participant records include a permission to contact field facilitating cohort management.
  • Onyx/RedCAP – Delivers patient questionnaires and handles associated data capture.
  • CaTissue – Handles biospecimen tracking and inventory for patient tissue samples.
  • I2B2 – Data warehousing software that connects to all other components, drawing down data and enabling complex queries to be applied to patient datasets eg: How many people over 50 that drink alcohol and have a low red blood cell count do we have a saliva sample for?

Community Event and Hackday

Last month I attended the BRISSKit community event held at the University of Leicester’s Gilbert Murray Hall. This was an opportunity for the members of the project’s user base to come together, catch up on progress and plug in to the two-day hackathon which followed. There was a good range of expertise represented at the event with health researchers, data managers and app developers putting the software through its paces and seeing what it could do.

We saw an impressive live demo of the platform with a use-case example illustrating the ease with which the elements interact, enabling data to flow through the system. Nick Holden and Olly Butters demonstrated the registration of a new patient and then showed how the system facilitated the collection of data and samples and, ultimately, drew information into the central database structure. They finished by querying the database to illustrate how data harvested from the component systems was now fully integrated.

Presentations from members of the pilot groups working with the platform demonstrated the practical benefits that introducing the tools into existing workflows had already achieved; one unit experienced an eight-fold increase in data logging speed when processing patient blood samples. These presentations, and others, are now available on the BRISSKit project website.

Hackday projects tackled issues such as the ingest of real data into the I2B2 system, integration of RedCAP data capture software with CiviCRM to enhance consent reporting; the development of a single login for all four of the BRISSKit components and coding a mobile phone-based barcode scanner to connect to CaTissue. Again, presentations from these projects are available on the BRISSKit website.

Broadening Participation

At present, the project is open to a few selected pilot groups but will be connecting with the wider community from the start of next year. The modular nature of the platform and the ease with which it can integrate with existing, proprietary systems should ensure that many biomedical researchers will find something to interest them in the project. Providing easy access to clinical data for researchers and supporting open source applications puts the BRISSKit project in an excellent position to meet the future challenges facing the biomedical sector.