EHI Live 2012

Over the last couple of days we (Jonathan, Robert, Saj and I) have been exhibiting BRISSKit at the E-Health Insider EHI Live 2012 event at the N.E.C. in Birmingham. It's the second year in a row we've had a stand in the skunkworks section. The skunkworks section was organised by Malcolm Newbury, and was set up to showcase open source developments in the health industry. Needless to say there was a lot less fancy suits in the skunkworks section compared to the more corporate stands trying to flog propriety systems :)

Our stand was well equiped this year, we had a couple of posters (user poster and technical poster) explaining what BRISSKit is, a banner, a glossy 12 page booklet and a slide-show running all day showing screen-shots. The picture on the left shows the stand in all its glory. We also had our laptops ready to show interested folk a live version of the stack working! Big thanks to Becca Wilson of the Space IDEAS Hub for sorting out all our promotional material.

As well as the stand we also had a couple of slots to give a talk about the project. Jonathan did this on the first day, with a live demonstration from me, on the second day Robert and Saj did the same thing.

It was definitely a useful event - we had lots of interest in the stand, and we had some good conversations with people on other stands that I think will lead to new and exciting opportunities...

See you next year EHILive!

Kirsty Pitkin's picture

Event Report: BRISSKit Community Day and Hack Event

BRISSkit Hack Event

The BRISSKit Community Day brought together project partners, interested biomedical research groups, developers and domain experts to learn about the BRISSKit toolkit: a national shared service designed to host, implement and deploy biomedical research database applications that support the management and integration of tissue samples with clinical data and electronic patient records.

In opening the event, Jonathan Tedds emphasised that BRISSKit team wanted to use this opportunity to work closely together with the biomedical research community to shape what they do, and are keen to work with all the partners involved to arrive a something sustainable that enables people to innovate and feed that innovation back into the biomedical community.

The community day took the form of a series of presentations and a round table discussion to collect feedback about the issues raised. This fed into a two day hack event, with developers working to resolve a series of practical challenges identified by members of the community to improve and extend aspects of BRISSKit.

This report summarises the community day and provides links to further resources, including videos from many of the presentations.

BRISSKit project update and sustainability

Dr Jonathan Tedds

Jonathan Tedds opened the event by providing an update on the progress of the BRISSKit project and and outlining their plans to ensure the sustainability of the platform.
He described the unique features of BRISSkit, including their use of mature open source applications, which have been fully customised for the UK. They key aim of BRISSKit is to to allow seamless integration between these tools over a web browser, so research groups have easy access to core infrastructure, without the need to build customised servers.

The project has completed the first phase of customisation and integration, is currently moving into its second phase, which will involve working with pilot groups and developing BRISSkit further to create a modular, scalable platform.

Tedds outlined the community feedback they have received so far, which has included fears surrounding the security of cloud computing, and described their responses to this feedback, such as exploring UHL N3 compliant hosting.

He concluded by emphasising the estimated returns of the project, demonstrating that 40 researchers using BRISSkit over five years will justify the funding from HEFCE. The team plan to roll BRISSKit out over a wider area than just Leicester, and are keen to engage with new partners within the community over next phase of the project.

You can watch Jonathan Tedds' presentation in full on YouTube.

BRISSKit demo: Leicester Cardiovascular BRU

Nick Holden / Dr Olly Butters

Nick Holden narrated a live demonstration of the BRISSKit suite by Olly Butters, illustrating the direction the platform is taking and how the services will integrate.

The demonstration focused on their use of CiviCRM to manage the recruitment of participants in a study at the NIHR Leicester Cardiovascular Biomedical Research Unit. They showed how to add a patient to a study and give permission for the data to be automatically passed to other elements of the BRISSKit toolkit, such as caTissue. They described their single data entry system API client for caTissue, which has sped up the process of entering data eightfold, and walked through how information is passed between other elements of the toolkit, including Onyx and i2b2 to improve the researcher workflow.

Holden and Butters went on to demonstrate how data from different sources within the toolkit can be used in the i2b2 interface to build searches based on data from both caTissue and Onyx, and to select samples to study. Holden completed the circle by showing how exporting information to CiviCRM can help CiviCRM to identify suitable cohorts to approach for future studies, where appropriate consents exist.

Holden emphasised that the elements of the BRISSKit toolkit have been selected because they are open source and have open APIs, which means that you can swap any element of the toolkit for another tool, provided the tool provides an API in a similar format. As a result, the team have started to think of BRISSKit as a platform, rather than a kit, that allows people to bring their own tools to the party and plug them in.

You can watch this demonstration in full on YouTube.

BRISSKit Contexts

NHS research context

Peter Knight

Peter Knight provided an overview of the broader picture beyond BRISSKit, describing a number of national projects to expose clinical data, such as prescribing and primary care data. He outlined key legislation changes in the field, including legal requirements in the Heath and Social Care Act requiring CCGs to make research as a key tenant. The Health and Social Care Information Centre has a new legal status to requisition data, specified to a standard, which, together with the Act, means that all the necessary legislative mechanisms are now in place to allow data to flow. He stressed that the issue is how we operate and manage that information.

Knight lamented that NHS trusts often have different ideas about what constitutes information governance, which often results in barriers to using patient data. However, he noted that the information commissioner has made it very clear that use of patient data for research purposes will be the default, as there is an exemption to the data protection act to allow use for research.

Looking forward, Knight described proposed changes to EU data protection law and the ramifications of these changes for biomedical research, such as the impact of a new right to be forgotten and a right to data portability. These issues will affect the way we manage and use health data.

Knight concluded by discussing standards and how these feed into a sustainable business environment, stressing that we spend £16bn on research in this country and £8.2bn specifically on medical research, so the challenge to make this work has real value to the economy.

You can watch an interview with Peter Knight, discussing the significance of BRISSKit on YouTube.

University research context

Prof. Anthony Brookes

Brookes described the broad vision of biomedical research at the University of Leicester and the forthcoming launch of their a new data-to-knowledge practice facility to demonstrate the strength of activities in Leicester.

He noted that whilst there is lots of wonderful infrastructure work going on, much of this is benefitting researchers, not healthcare directly. He noted that researchers and healthcare practitioners are two different ends of the spectrum and there is a gap between these to worlds. Brookes argued that we need to develop a bridge using knowledge engineering, and described efforts in Leicester to create a community of 200 participants to discuss what is needed to bridge this gap, called I4Heath.

In exploring this issue further, Brookes emphasised that we need to use research data to identify knowledge that is clinically useful, and defined the difference between what he termed knowledge engineering 1.0 - which favours researchers and allows them to play with the data - and knowledge engineering 2.0, which looks at the data to identify the bits that will be useful to healthcare professional, and presents this data in a useful way.

Brookes concluded by describing an ideal knowledge portal, where tools can be brought together and integrated to distill information and the role BRISSKit can play in such a portal.

You can watch Professor Brookes' presentation in full on YouTube.

Bio-Ontologies, BRISSKit and research

Dr Tim Beck

Tim Beck provided an example of how BRISSKit has contributed to the wider research community by allowing researchers to bind questionnaire terms to a bio ontology as during the question development stage.

Beck explained how this issue had been highlighted when they attempted to map Leicester cardiovascular BRU patient questionnaire data with SNOMED CT. Before BRISSKit, researchers had to load the bio ontologies into i2b2 directly using commercial tools, which often didn't work. The BRISSkit team have been able to develop a bridge between the questionnaire tool Onyx and i2b2 to over come this.

He described a pilot project within BRISSKit to successfully flow data between Onyx and i2b2, which allowed users to manually bind terms to a bio ontology as they create the questions, thus creating data which can be ported into i2b2. Whilst there is currently a problem with a lack of sub ontologies, and there are efforts underway to build a tool that will help to produce a customised DAG input for i2b2.

You can watch Tim Beck's presentation in full on YouTube.

BRISSKit Partners and Community Presentations/Discussion

Information Governance

Dr Andrew Burnham

In this presentation, Burnham explored the information governance issues that are specific to BRISSkit, including those which are likely to arise from the move from research project to a sustainable, assured service.

His work on the project involves examining how information governance tool kits should apply to BRISSKit, and what sort of assurance an Acute Trust Information Governance (ITG) toolkit and the Secondary Use Team ITG would need to work with BRISSKit.

Burnham described their efforts so far, which include creating a data management plan using the DCC DMPOnline tool, and surveying other NHS standards, such as care records guarantee and NHS Records Management. The BRISSkit team have also considered how to include security management assurances, such as ISO27001/ ISO27002 accreditation.

Burnham appealed for guidance to take this forward, as the team needs to understand what users need to make sure BRISSKit meets their requirements.

You can watch Andrew Burnham's presentation in full on YouTube.

UCL Institute of Child Health

Dr Tito Castillo

Castillo provided an overview of his team's efforts to achieve ISO27001 accreditation for UCL's Institute of Child Health, including a step by step guide to the process.

He explained the preliminary steps involved in developing an Information Security Management System (ISMS). This involved obtaining management support, defining the assets, carrying out a risk assessment and developing a risk treatment plan. From this treatment plan they developed their ISMS, which is an intensive document management process. Once the ISMS was completed, they were required to go through two phases of auditing to become accredited, and ongoing auditing to monitor any corrective action.

The compliant system they developed was based on their requirements for a secure data enclave or "safe haven" with security controls, secure end points, a scalable architecture using standard technology and open information standards. The resulting system involved a virtual desktop environment supported by a secure data centre provider that had already achieved ISO27001 compliance.

Castillo concluded by exploring the scope to work together with BRISSKit to create a federated service to combine their secure end point structure with the range of BRISSkit services.

You can watch Tito Castillo's presentation in full on YouTube.

Leicester Respiratory BRU

Dr Robert Free

Robert Free described Leicester Respiratory BRU - an NHS and University research partnership to promote the development of new effective therapies for the treatment of asthma and chronic constructive pulmonary disease - and outlined their aims to create an integrated research system with multiple components, including BRISSkit, and to use open source tools where possible.

Free gave an overview of their informatics plan, which linking data from their own bespoke clinical research tools with BRISSkit components. He demonstrated a prototype of the processes involved and described the rationale behind their approach in more detail.

Crucially, Free believes that using BRISSKit components is important, as it will save time and open opportunities for collaboration. He concluded some ideas for such collaborative efforts, including development of a hospital patient lookup, refining study enrolment and integrating data from i2b2 into their clinical research tool.

You can watch Robert Free's presentation in full on YouTube.

Birmingham School of Cancer Studies

Paul Mason

Mason provided a clinical trial perspective on the issues raised by the BRISSKit project, based on his experience at the Cancer Research UK Clinical Trials Unit, where his department supports informatics across a wide range of trials. Their approach involves creating their own bespoke systems to address regulation issues and to deal with the volume of historical data generated over long term clinical studies.

Mason was able to trace the history of work in the area prior to the emergence of BRISSKit, focussing particularly on the efforts of the Cancer GRID proposal, which was an e-science project funded between 2003-2006. Some of the ideas explored by Cancer GRID project were are similar to those now pursued by BRISSkit, but Cancer GRID did not achieve as much as was hoped for a variety of reasons, including modest funding and wide-reaching aims. Mason described many of the practical difficulties the project encountered in detail, observing that the clinical trial domain is too complicated. Whilst the Cancer GRID project is still ongoing, Mason felt little progress had been made to create a connected infrastructure prior to the emergence of the BRISSKit project.

Mason explained what he finds exciting about the BRISSkit project at its potential to change biomedical research practices. He stressed that the basic idea is good, the team are taking a pragmatic approach, the scope is not overly ambitious - dealing with manageable bite sized chunks, and emphasised the value of good communication from the team. He outlined the CRCTU's next steps to engage more fully with the BRISSkit project, which elicited discussion about the boundaries of the BRISSkit project. Mason advised against including clinical trials at this stage, but believes there will be a natural evolution over time.

You can watch Paul Mason's presentation in full on YouTube.

Leicester Lifestyle BRU

Tim Skelton

Skelton described the evolution and aims of the Leicester Loughborough Lifestyle BRU, which is designed to support translational clinical research in priority areas of high disease burden, specifically focussing on lifestyle health, by bridging the gap between research and the bedside.

The Lifestyle BRU has two research themes focused on physical activity and metabolic health. Much of their data comes from GP data, so they are interested in how that can be accessed and used. Skelton outlined some of the questions they are looking to explore and some of the different sources of data for this work.

Their overarching strategy is to collaborate, and see partnering with BRISSkit as a way to achieve this.

You can watch Tim Skelton's presentation in full on YouTube.

Breakout Groups: Feedback on components

Facilitated by Neil Beagrie

Beagrie asked the participants to split into small discussion groups to provide feedback about the issues raised throughout the day to help shape the future development of the BRISSKit project. Each group was asked to reflect on the following questions:

1. Are you willing to adopt open source tools like i2b2, caTissue, CiviCRM etc.?
2. Would you be willing to pay for the ongoing support of the tools that you subscribe to?
3. Which component(s) of BRISSKit demonstrated today will be of greatest interest to you?
4. What potential benefits could you see in BRISSkit as a service for you?
5. Are there any potential negative factors which could have a major influence on our level of interest?

Malcolm Teague from JANET gave us his perspective on some of these issues and why he is excited by the BRISSKit project in this short interview...

Hack Days

Whilst the BRISSKit Community Day offered an opportunity for talking, the two-day hack event that followed provided an opportunity for doing.

Participants in the community day provided the inspiration by outlining some of the practical developments they would like to see within the BRISSkit toolkit. These ideas were debated further to establish which could be realistically tackled within the two days allowed, before the developers got to work to produce either a proof-of-concept or a fully fledged solution for each of the chosen ideas.

In this short interview, event co-ordinator Malcolm Newbury explains the purpose of the hack event and how it has benefited the BRISSkit project already...


A range of ideas were proposed for the hack event, including:

  • REDCap/CiviCRM/i2b2 integration
  • DAG Builder/CiviCRM and caTissue integration
  • A CaTissue web client
  • i2b2/HES data upload
  • i2b2/cancer data upload
  • A barcode scanning through iPhone to caTissue
  • Federated access to the BRISSKit tools
  • Information security/operating procedures
  • LDAP access/single sign on

A spokesperson for each idea pitched to the rest of the hack day participants to attract people with similar interests or the skills necessary to work on their idea throughout the event.

Watch the quick fire pitches to get a full picture of the range of ideas identified:

Developer Perspectives

Throughout the event, we collected a series of interviews with participants to capture their perspectives on BRISSkit, the hack event and their individual projects.

In this interview, Paul Mason from the Cancer Research UK Clinical Trials Unit describes his involvement with several hack day projects looking at aspects of i2b2 integration, and discusses some of the different issues that arise when using clinical trials data...

In this interview, Jens Schwanke and Christian Bauer from the University of Gottingen discuss the different cultural approaches to privacy between the UK and Germany, and how BRISSKit could be adapted for use in Germany...

In this interview, Nick Holden from the University of Leicester discusses how his hack day project changed throughout the course of the event and how this process has altered his thinking about similar problems...


At the end of the hack event, participants were asked to report back on their progress in a series of final presentations. This showed the range of useful outcomes and contributions to the BRISSkit project that the teams were able to produce in a short, but intense period of rapid development.

The final presentations included:

The Trinity

This was a combination of three separate projects, each looking at different aspects of i2b2 integration using demonstration data from HES and cancer research clinical trials data. The presentation included a demonstration of this integration working in practice, presented by Anthony Thomas, Christiana McMahon and Christian Bauer.

Single Sign In

Malcolm Newbury from Guildfoss described his group’s efforts to use LDAP to address the problem of creating a single sign in across the four separate tools within the BRISSkit platform to help reduce the account management overheads.

caTissue Barcode Scanner

Adam Roberts demonstrated an smartphone application developed over the two days to scan very tiny barcodes from the end of sample vials and import the information from those barcodes into caTissue.

caTissue API Client

Nick Holden explained why his group’s project to develop a generic caTissue API client did not result in any code, but has resulted in a series of feature requests to improve the original caTissue programme for all users.

Linking CiviCRM to REDCap

Robert Free described his group’s successful efforts to integrate CiviCRM and REDCap, and to create a simple CiviCRM study creator as a Drupal plugin.

Prize Winners

All of the outcomes from this event are extremely valuable to the BRISSKit project and the community using these tools. However, thanks to the generous support of Aine McGuire at ScraperWiki, there were some prizes available for projects that particularly impressed our team of judges.

Chair of judges Jonathan Tedds announced the winners and drew the event to a close with his reflections on the progress made during the hack event.

If you want to see more, a selection of action shots from the event are available in the BRISSKit Hack Event Flickr set.

Community meeting and hack event

Well the dust has finally settled after last weeks community meeting and hack event. I think everyone agrees that it was a resounding success! Well done us :) We will be doing a write up of all the projects in the next week or two, and we will be posting videos from the event too! Watch this space...

Hack event preparations and closing the loop

If you've been living in a cave, on Mars, with no internet connection, then you might not know that we are having a community event and hack event in a couple of weeks time. If you haven't heard then I'd recommend you head over to the events page and have a look, it should be good fun. In preparation for this I'm having to build several instances of our software stack for people to play with, this has pushed the size of our virtual infrastructure up a lot (about 100 VMs now), and so far it is all managing ok!

In other news; Saj has been working on getting data from i2b2 back into CiviCRM, and it's almost done! This is a key part of the BRISSkit infrastructure as it means we will be able to do cohort queries in i2b2 then push the results back into a group in CiviCRM. The reason this is an essential part of the software is that it connects the pseudo-anonomised data back up to the real identifiers, so you could contact the cohorts you have generated (assuming you have the right privileges etc to do so of course). Awesome.

CiviCon 2012

Yesterday some of us went to CiviCon 2012 down in London. This was the first CiviCon any of us had been to, so we didn't know what to expect. Needless to say the CiviCRM guys put on an excellent conference and we all learnt a lot!

Between us we went to a lot of sessions, I really enjoyed the API and extending CiviCRM sessions, I definitely think we will be using the stuff we learnt there in the future. The talks on templating and drupal web forms were really informative too.

Nick gave a lightning talk on BRISSKit, it really was lightning quick - 6 minutes and 42 seconds quick! It was very well received by the room, which was nice.

All in all it was a good day out. Thanks CiviCRM!


Drupal migration 2

If you can read this then it all worked!

I've just finished migrating our drupal install (i.e. this) to drupal 7. This not only gives us a more modern install to work with, but it means we can use CiviCRM behind it! You'll note that now our events page is driven by CiviCRM :)

If you happen across any issues then let me know via the contact page.

i2b2 (AUG) Second Annual Conference 2012

Saj Issa, Malcolm Newbury and Andy Carruthers attended i2b2's AUG conference in Boston, MA on 24th July 2012.

A fantastic event which allowed the team to showcase the Brisskit stack of i2b2, catissue, onyx and civicrm to the i2b2 community. Demo was well received and got a lot of attention. Most notably from Shawn Murphy of i2b2 and Uli Sax & the team from Universities of Erlangen & Goettingen, Germany who expressed their excitement of how Brisskit provides an end to end solution.

Much networking was done and valuable contacts were gained, most notably from EU where there are potential i2b2 partners in Germany, France and Italy.

Key points of conference :
* A new version of i2b2 v1.7 will be released in Feb 13, it will use Jboss 7.1, Bamboo, JIRA, Axis 2 v1.61
* A new feature for temporal queries will be added.
* The identity management cell has been improved.
* A new feature to manage clinical trials will be added to i2b2 v1.7.

Overall the community has done some amazing work.

So.. personally what excites me.. Smart i2b2 plugins seem to be very powerful and interesting !!

Malcolm Newbury's picture

Brisskit's CiviCRM used to link up busy clinicians at #nhshackday

Had a great weekend using CiviCRM, MULE ESB, and Iphone to provide a directory service to busy doctors who need to get hold of local department phone numbers or colleague's bleeper numbers. deeper summary of problem here:
Proves once again - the maturity of Brisskit's components and wisdom underlying their selection.


Bio-ontology: building bridges between Onyx and i2b2

We have been investigating solutions for structuring the patient questionnaire data collected in Onyx so they can be queried using the i2b2 ontology interface. It would also be advantageous to allow data to be interoperable across installations. The “semantic glue” we have found are bio-ontologies.

The SNOMED CT terminology has been selected and approved for use in the NHS, so we started with binding questionnaire data to SNOMED CT. But, wouldn’t it be a huge advantage to allow more flexible ontology use? How about using LOINC (Logical Observation Identifiers Names and Codes) or ICD-10 (International Classification of Diseases) instead? Or, for that matter, any of the bio-ontologies in BioPortal?

To address this, Rob, Jeff and I have developed a solution to allow questionnaire data to be bound to bio-ontology terms contained within BioPortal and then be imported into i2b2 while maintaining the ontology term parent-to-child inheritance. More details soon…


Drupal migration

We are migrating the Drupal CMS and wiki today. Moving them from a VirtualBox machine hosted by RCS to a VMWare machine hosted by Eduserv.

We've already migrated a test version a few days ago, and so today's move should just be a case of updating the copy of the databases to get the latest content and altering the DNS records to point to the new server.

If this blog entry makes it to the new server, then we know the database updates went through ok.