In my first blog of this series, titled Pharma Lifecycle Plagued with Integration and Data Management Challenges (Part I – Overview), I outlined six phases of the pharma lifecycle and set the stage to provide deeper insights into the integration and data management challenges each phase faces. The first stage I’d like to tackle in depth is Research and Development (R&D), which broadly covers drug research and discovery.
Besides the analysis of the compounds which make up drugs, access to patient data for research is paramount. Patient data can come from a variety of sources, typically data acquisition applications such as Electronic Data Capture (EDC) or Electronic Health Record (EHR) systems. Once the “business relationship” has been determined to gain access to the data, the technical issues arise since we are far from being at a point where we can rely on standards to give us quality data in a usable format. Therefore, much energy is spent by data scientists acting as “data janitors” to clean up the data before it can be used.
The following white paper, Pharmaceutical R&D Company Leverages Cloud Integration Platform to Bring Its Collaborative Environment to Life, details how real world evidence patient data is acquired from hospital institutions and harmonized using cloud-based services before being deposited intotranSMART where it is available to scientists in a timely manner for research—no janitorial work required. This is a collaborative environment, although without knocking down the barriers related to integration and management of the data, the platform is useless.
As you will see, key challenges include:
- Integration of low-dimensional clinical data with high-dimensional omics data. The availability of high-throughput genomic analyses has increased the need for adequate tools to manage and explore these multiple scale, incongruent, incomplete and complex big healthcare datasets so that different questions can be postulated.
- Substantial costs and the intellectual demands of managing computational complexity and data protection. Open source analytic tools such as tranSMART require complex infrastructures (e.g. web and Java servlet servers, databases, solr, and rserve) that are most likely out of reach for the average translational clinician or researcher.
I’d love to hear more about the integration and data management challenges you’re having around R&D and how you’ve solved them. Or, if you haven’t yet, let’s brainstorm how we can, collaboratively!
Until next time,
Gary Palgon, VP Healthcare and Life Sciences Solutions