RWE and Risk: A Primer
Balancing the Demand for Deep Data with Patient Privacy
As pharma companies have become more sophisticated in their use of real-world evidence (RWE), they have moved beyond standard datasets from vendors to access a deeper and wider variety of datasets, often working in partnership with local health systems. Indeed, leading pharma companies have built comprehensive networks and data platforms can that provide a shared understanding to teams across the organization about the reality of what is happening in healthcare. The increasing number and variety of datasets analyzed — including novel sources from social media through to medical imaging — are delivering ground-breaking insights.
However, accessing a growing range of data sources will necessitate new capabilities, including the critical need to protect patient privacy. Many pharma companies cite protecting privacy as one of their primary imperatives in building RWE into their capabilities, but also a key barrier to making progress. Real-world data (RWD) is patient-level data drawn from a variety of sources that all contain varying amounts of protected health information (PHI). Removal of PHI is a critical first step to using this data in RWE analysis. The challenge is how to effectively anonymize the data without diminishing data quality in an exponentially increasing number of contexts.
Fortunately, new software enabled capabilities now exist to address this urgent challenge. Best practice approaches and guidelines have emerged advocating for a risk-based approach to de-identification in order to balance the competing goals of anonymity and quality. Levering an automated de-identification process that uses a risk-based methodology ensures a continuous — and legally compliant — flow of data for RWE analysis.
In this primer, we describe the techniques, software platforms and highlight example use cases of how pharma companies can take both sustainable and secure approaches to access new datasets and build RWE data networks.