Clinical Studies and Data Collection and Reuse
Clinical studies drive medical research, but they're complex and costly affairs. In order to be scientifically useful, they must record large amounts of data on the health and treatment regimens of a carefully selected group of test subjects. The IT infrastructure to collect this valuable dataset is often created from scratch for each study, a time-consuming and expensive process that leads to non-standard systems that compile data in non-standard ways.
Could standardized software increase the effectiveness and reduce the cost of clinical studies, and thereby accelerate medical research? Together with colleagues from the CancerGrid project in the United Kingdom, we decided to find out.
Building on the work of CancerGrid, which set out to improve information management for large-scale Phase III clinical trials, we employed a semantics-driven approach and developed models, standards, and software for collecting medical research data. We broadened the scope of the work to include early-phase (Phase I and II) studies, in which many more observations—including detailed molecular and imaging data—are made, but in a smaller number of subjects. Despite the importance of early-phase, experimental medicine in the development of new therapies, data from these studies is rarely reused or combined. We aimed to show how appropriate information systems support can be rapidly provided, facilitating study management and data reuse, at little or no cost to the researchers.
Our project focused on early-phase experimental studies in cancer, but the technology has proven to be widely applicable and is now deployed in institutes around the world. For example, it is being used to support a clinical study on the effectiveness of different ways of administering pneumonia vaccines to children in Nepal. Aside from allowing the trial to be run more cost-effectively, our system is collecting the data in a standardized way that will allow it to be easily reused in future studies.
The storage of clinical trials data required the creation of a robust and highly flexible IT infrastructure that would permit rapid reconfiguration to support different studies while maintaining the consistent data definitions needed to ensure that the data can be reused. Microsoft SharePoint and Microsoft InfoPath provided the needed robustness and configurability, demonstrating the value that these standard business technologies can bring to scientific research. The academic researchers shared their experiences with Microsoft product development teams, contributing to the design process for the next generation of Microsoft Office tools.
Learn more about this research:
- Towards programming languages for genetic engineering of living cells
- A programming language for composable DNA circuits
- Abstractions for DNA circuit design