Last week, I had just returned from San Francisco after a face-to-face meeting about the CDISC work to build a repository of shared semantics. After a long trip home, I indulged in my interest in cycling and caught up with the Tour of Italy, the Giro d’Italia. The Giro, along with the Tour of Spain – the Vuelta – and the Tour de France are the three premier events on the annual cycling calendar. This year was the 100th running of the Giro and, after being out of professional cycling for three years, Lance Armstrong was competing in the event for the first time. Lance was there to raise awareness of his Livestrong charity and its work in fighting cancer. With Lance at the event the media interest exploded.
It is the impact that Lance Armstrong has as part of his work with Cancer that made me think. Lance was posting daily videos from the tour about how it was all going, interviewing other riders etc. But as part of each video he would also mention a victim of cancer and detail the individuals fight against the disease. The one that caught my attention was a young boy in the UK dealing with cancer. He was being supported by a charity, Cyclists Fighting Cancer, that helps youngsters with specially equipped bikes to help them with muscle rehabilitation. The 30 second mention touched all the right buttons for me, it was UK-based, it was cancer – a disease that has affected my own immediate family – it was cycling, it was Lance Armstrong. It had an impact; I donated to the charity.
It also made me think about the influence CDISC has had and continues to have. It is to have an impact ensuring that those working to beat cancer and all the other diseases are able to share and use data with ease.
My thoughts went back to the meeting in San Francisco and the repository work. The CDISC repository of shared semantics (CRSS) work could, to my mind, have as great an impact as anything else CDISC have achieved to date, if we get it right. And that is the important qualifier, if we get it right. For CDISC the project is a significant effort and therefore it is only right that we are sure of our needs, our solution and our plans before proceeding.
As an aside, I am going to refer to the project as CRSS. There project may well be renamed in the near future to give it a more meaningful and non-IT name, but for the purposes of this post, CRSS will do.
At the CDISC Board meeting in April 2009, held in Budapest, the board decided that we should proceed with the CRSS project. The aim is to make a decision at the next CDISC Board face-to-face meeting – to be held in September 2009 – as to whether or not CDISC should build a production repository. To allow the board to make that decision we need to answer some simple questions. Unfortunately the answers may not be quite so straightforward. The questions are of the fundamental type: What is it we want to build, why do we want to build it, is there a business case for building it, what are the risks?
To allow for such questions to be answered we will use a methodology borrowed from the software industry called the Unified Process. The first phase of the process is the inception phase and it is this phase that we will execute between now and September. Emerging from the inception phase will be the required information that the board needs to make it decision. If the decision is yes then the next phase (the elaboration phase) can proceed but that is not yet in scope.
I was going to write something about the inception phase but a document supplied by Charlie Mead contains words that are better than anything I could conjure up:
In the Inception Phase, the project’s overarching ‘business case’ is critically examined in terms of vision and scope, critical risks that must be managed, budget and resource requirements, etc. The Inception Phase ends with a ‘Go/No Go’ decision on the project, thereby ensuring that projects with unmanageable risks, insufficient resources, and/or unclear scope or vision are terminated early rather than late.
As part of the inception phase we will execute a small pilot “sub-project” to address two risks.
One of the biggest issues we as CDISC face is the scale of the work. SDTM has given us a solid base to build on and built the common “safety” domains. Now we need to build the therapeutic area domains. We will attempt to achieve this by bringing together the varied definitions used within research organizations (biopharma, CROs, academic institutions etc) to date and attempt to harmonize them. But this is a large undertaking and we need to assess the scale of the work, the necessary processes and costs (metrics). This will allow us to determine if the project’s business case is sound.
The second risk is around ensuring the quality of the definitions placed into the repository. We want to ensure these are of high quality and protect against such things as overlapping or duplicate definitions. To assist in this, we want to evaluate some of the work being performed in semantic wikis and ontologies to assess whether or not these can help us populate the repository with high-quality definitions. Such technology may also help in improving the process of harmonizing the content for the repository.
Given the scale of the work and the importance placed upon it by the CDISC board we have decided to allocate significant resources to the project. Rhonda Facile will be leading the project with Julie Evans and Bron Kisler assisting. Julie is working on some of the key documents and with modelling support, while Bron will assist in the areas of content governance and business requirements. This may result in existing work being impacted but we are taking steps to add extra resources to minimise this.
So the inception phase is there to ensure we get it right, to ensure that CRSS has the impact we think it can have. It will allow CDISC to improve our existing standards such as SDTM and CDASH by improving the quality of definitions contained within those standards, it will provide a platform by which CDISC can expand those standards into therapeutic areas and it will provide the means by which those standards can be delivered in an electronic form to our members and consumed by applications to help streamline the clinical research process. Not a bad impact.
Dave Iberson-Hurst
Dave
Are you aware of the National Center for Biomedical Ontology? (http://www.bioontology.org/) A collaborative partner, perhaps, for this kind of exercise?
Also, there’s an interesting conference on biomedical ontologies later this month (July 24-26, 2009) “ICBO: International Conference on Biomedical Ontology” – http://icbo.buffalo.edu/
Barrie