A member of the Clinical Programming Team writes about their experience at the CDISC European interchange in Stockholm held on 18th-19th April 2012.
As part of the CDISC European Interchange meeting, I contributed a presentation on how Quanticate have used the CDASH standards to facilitate data capture and transformation to the SDTM standard.
Although the title of the Interchange was ‘Standards Accepted for Progress – CDISC the Reliable Partner’ most of the emphasis was on evolving standards and the use of metadata from protocol through to ADaM dataset and CSR output. The opening session on the ‘CDISC Technical Roadmap’ generated a lot of interest. This session focused on CDISC, the updating of standards more regularly, and giving CDISC adopters a ‘roadmap’ of the different standards released. CDISC will be making more use of the BRIDG model and beginning the transition to Share. In particular, the creation of SDTM Therapeutic Area (TA) domains is the current focus. They have quite an aggressive project timeline for the creation of these SDTM domains, which is good news in terms of getting some TA standards out there (the SDTM standard has been lacking in this) but bad news in that there will be a lot of new domains to review and possibly incorporate into company-specific CDISC standards. Also of note: the CDISC website has been remodelled so that it gives each standard in context. For example, SDTM and ADaM are in the Foundation standards category, with SDTM providing the data for ADaM. See CDISC Foundation Standards for more detail.
There was also an interesting presentation on the adoption of CDISC standards at a large pharmaceutical company. The biggest challenges facing the adoption of new standards were discussed: training and culture shift. To gain acceptance of the standards, a company CDISC Network was set up to:
• Cultivate Subject Matter Experts
• Share Information
• Develop a CDISC Roadmap
• Encourage CDISC Team Participation
• Create Pilot Projects
• Review and Implement New Standards
The company roadmap had a set of targets for each standard implementation but they found standards adoption ‘impossible’ without the relevant tools to support it. Getting investment in these tools and adoption of the implementation and road map also requires management buy-in. The reasons given to management to adopt the standards were:
• CDISC is steadily evolving (see Technical Roadmap above)
• FDA requires SDTM, define.xml and ADaM is not far off
• Peer organisation support: more and more vendors can supply CDISC compliant tools
• CDISC communications: get involved in webinars, ESUG Face2Face events, Interchanges, training
• Externalisation: what are competitors doing? Will we need to share data? Who has adopted these standards and how have they done it?
This was followed by an excellent presentation on their adoption of CDISC ODM to capture data electronically, where the initial goal of the company was to move to an EDC supplier that would facilitate the creation of SDTM data from their clinical trials. They went through a number of iterative steps in the consultation phase in evaluating EDC vendors and found that the most important functional aspect of these tools was to be able to specify the eCRF and database using ODM which can then be imported into an EDC system. They found 2 vendors who would do this.
The company have chosen to adopt one of these tools and use a range of EDC vendors depending on the protocol. The most important aspect of this project is now to be able to specify the metadata in ODM: SDTM compliance is merely an outcome of getting this right.
These talks, along with another on using ODM to migrate clinical data, all pointed to the growing importance of metadata and the integral part it has on enabling standards compliance and traceability of data points from protocol to CSR.
I also attended the stream dedicated to Electronic Health Records (EHR) and the flow of data between hospitals and clinical research organizations. CDISC are doing a lot of work around this as part of the IMI (Innovative Medicines Initiative), using the BRIDG model to link HL7 and other healthcare standards to the CDISC ODM and other CDISC standards. CDISC gave an interesting talk on this initiative (called EHR4CR) highlighting: the use of datamarts to capture data from the hospital workflow system so that data integrity would not be compromised and workflow interrupted, the introduction of a platform of tools and services that healthcare and clinical research can use, and the formal standardisation that is required so that data from different sources and systems can be interrogated. For more information on this project, please see here: EHR4CR
The final stream was dedicated to the ADaM standard. The FDA talked through the use of ADaM datasets in the review of clinical trials: some reviewers will use both SDTM and ADaM datasets, but some are ‘suspicious’ of the derivations in ADaM and find that the sponsor’s metadata have ‘holes’. To this end, a Data Users Guide is being created by Work Group 4 from the FDA/PhUSE Computational Sciences Symposium which was held at the beginning of April. When I questioned the presenter on why define.xml was not being used, he stressed that it was, but the FDA felt it was not currently good enough to trace data from CRF to ADaM dataset. This deficiency in define.xml is something that needs to be addressed..
Overall the Interchange was a fantastic event to attend to learn more about the evolving standards and to meet people with a passion for these standards and for driving clinical trials data to a standard that can be used, re-used and interrogated quickly and consistently.
Related Blog Posts: