Summer 2007
New Challenges, New Solutions: Acquiring, Accessing and Utilising Clinical Trial Data
A joint meeting between the PSI and ACDM was held for statisticians and data managers to share experiences and challenges being faced in the current climate of the requirement to collect data faster, more efficiently and to make the most of the information obtained.
Seven very interesting and thought provoking presentations were given throughout the day covering topics such as communication, study design, data standards and system validation. There was one strong common theme that came from all the presentations: the importance of cross-functional collaboration during the whole life of the trial.
The first presentation concentrated on the importance of communication within the study team. It is shown that words account for only 7% of the message people take away from an encounter and that body language and tone account for the remaining 93%. In the current culture of large, global study teams and the involvement of CROs it is not always possible or practical to have face to face meetings therefore the majority of communication within teams takes the form of teleconferences or emails.
The presenters highlighted that communication within teams should be structured, clear and regular. Within Roche they have created a document called the ‘Data Review Plan’ which defines the roles and responsibilities of the different team members and lists all tasks that need to be performed with their associated timelines. This document is created at study inception and reviewed/updated regularly during the course of the study. As all members of the team help to create, agree the contents and sign this study plan it acts as a ‘Contract’ for the conduct of the trial.
The second topic to be presented was ‘How statisticians can help bring added value to clinical trials’. Great benefits can be achieved by continued involvement of the study statistician and ongoing review of the data rather than the traditional approach of the study statistician only being involved in the protocol design and end analysis. Benefits include:
Analysis for futility or efficacy at a formal Interim Analysis (as defined in the protocol).
Reduce timelines at the end of the study.
Monitor recruitment and screening failure rates. This helps the study teams track the trial and assess if it is on target and if not put corrective actions in place at an early stage e.g. re-train sites on eligibility criteria, add new sites to the study or CAP recruitment at the country/site level to avoid bias.
The last point was supported by a later talk ‘How statisticians and modelling can help predict recruitment rates and resource requirements’. One company uses a Patient Recruitment Simulation Model to predict patient recruitment which takes into consideration recruitment on a weekly basis and at a site level thus allowing for regulatory approvals and site start up timelines. The model is updated during the course of the study with actual rates seen to provide a more accurate picture. This information allows CRF flows to be forecast which is used in resource management.
This was followed by a talk on Adaptive Designs. The use of adaptive design trials can lead to cheaper and faster drug development programs. By conducting these smaller trials first and determining doses that are well tolerated and efficacious, it leads to well planned, efficient larger trials where there is no need to recycle or re-do analyses to gain the desired results.
Many considerations should be taken into account when deciding on whether an adaptive design is practical. A few main points highlighted were:
Recruitment needs to be controlled; fast recruitment does not allow sufficient time to react to the data being collected to adjust the treatment given to the next subject.
Speed of data capture and review is critical, there is minimal time to make and implement decisions. As a result Electronic Data Capture is found to be the technology of choice which also allows for easier changes to the CRF which may also need to be implemented.
Availability of drug supplies. Adaptive design trials have led to just-in-time drug management systems (these can save Pharmaceutical companies millions of dollars in drug wastage across all types of trials).
An update on where CDISC is with different initiatives was provided. CDISC founded in 2000 now has over 170 corporate members and all submissions to the FDA are now being made using the CDISC standards. One of the initiatives to be discussed was the BRIDG Model where CDISC, HL7 (healthcare), FDA and ICH are working together towards producing interfacing standards that will allow semantic interoperability. The goal is that subject data will only need to be entered once thus negating the requirement for source document verification. Further information on the BRIDG model and CDISC can be found at and

The final presentation of the day was on ‘Implementing AGILE systems development and Validation’. The traditional method of system validation occurs at the end of the development. It does not allow for changes in specifications to be incorporated (all user requirements must be defined up front) and involves a lot of testing and documentation.

The AGILE method is designed to perform validation and user acceptance testing in smaller, more frequent chunks. The idea is that the process is more adaptive; it allows for modifications due to the close interactions with the end users, eliminates waste (only 60% of features regularly used) and therefore results in a faster delivery of the system. This method is used in many industrial sectors and offers potential benefits to the pharmaceutical industry.
During the day many techniques employed to manage clinical trials by various companies were explored e.g. adaptive design trials, CDISC. However the main message that I took away was the importance of communication and collaboration between the whole study team from the start to the end of the trial, ensuring the expertise of each group is fully exploited. That processes, timelines and agreements are clearly documented and available so the team are fully aware of their role within the trial. Although this is something I think we are all aware of, it is something that can be overlooked with the pressures of meeting timelines etc and it was great to hear how companies are successfully putting this in place.
Cath Jones