So what? In an August readout from Fitbit, researchers reported diagnostic accuracy of AUC 0.77 in detecting COVID across 1,100+ patients using the same digital biomarker triad with 4 days of prior symptom history. This updated readout not only validates the initial findings among a much larger study group, but shows slight improvement in overall diagnostic performance by layering patient-reported symptom data into the “triad” as a fourth digital biomarker. While still in its infancy, the digital biomarker space is showing early promise as a complementary (and in the distant future, perhaps substitutive) approach to traditional biomarkers for detecting disease. The overwhelming trend toward multi-omic signatures in the “traditional” biomarker space is already coming to bear in digital biomarkers, which we expect will explode in complexity, diversity of biometrics and behavioral markers, and patient-reported data in the years to come. In addition to the Fitbit DETECT study, other major COVID-focused digital biomarker development programs underway include Huami’s ~1.3M-user study in China and the Robert Koch Institute’s study in Germany with 500,000+ enrolled.

TLDR: After investing in Seqster in February, Takeda has now announced an extended partnership with the RWD start-up dubbed the “mint.com” of health data by its CEO, Ardy Arianpour. Under the partnership, Takeda will gain access to the Seqster Research Portal (SRP) which matches patient-level clinical data from 3,000 hospitals / health systems and 150,000 doctor’s offices with patients’ molecular and wearables-generated data. Takeda plans to leverage the portal to launch 12 distinct use cases across the organization in the next few weeks.       

So what? Takeda’s announcement comes on the heels of its recent partnership with Amazon Web Services and Accenture aimed at cloud-driven organizational infrastructure transformation. Other global pharma giants have inked similar partnerships (e.g., Merck-AWS, Google-Pfizer, Google-GSK, Google-Bayer) establishing the infrastructure to support larger, more real-time use of RWD. The Takeda-Seqster partnership also follows a trend of pharma companies partnering with RWD aggregators (e.g., Roche / Genentech-PicnicHealth) to tap into larger, more multi-omic datasets with potential for massive scale, automated data linkage (e.g., Datavant), and near real-time reporting. Stay tuned for our interview next week with Seqster CEO, Ardy Arianpour, and Takeda Head of Emerging Technologies & Digital Partnerships, Emir Roach!

So what? Each player in this partnership arguably has the resources and runway to roll-out directly competitive offerings on real-world data for new therapies. However, perhaps the most differentiating aspect of their joint offering is one that these players could not build alone. This is the opportunity for seamless, longitudinal visibility on the  performance of newly approved therapies, which would appeal to pharma and payers alike. Longitudinal data capture is one of the key pain points observed with disparate RWD offerings available today, because patient data is so often siloed across disparate institutions and IT systems (e.g., EMR, LIS, EDC, etc.). Datavant’s market-leading “patient key” tokenization technology is particularly crucial to building longitudinal, patient-level datasets spanning data sources, as evidenced by its growing number of data partnerships (e.g., Parexel-TriNetX, Takeda-Seqster, Blue Health Intelligence, Medable, Precision Digital Health, Prognos, Symphony Health, Trialbee, Castor). Given Medidata’s entrenchment in clinical trials, Datavant’s open data ecosystem and TriNetX global reach across ~170 HCOs, these companies are incredibly well-positioned to facilitate longitudinal data capture that could offset the need for post-market studies requiring active patient follow-up.