Solving a Critical Challenge: A Discussion on Data Gaps in Life Sciences 

Solving a Critical Challenge: A Discussion on Data Gaps in Life Sciences 

On a recent LinkedIn Live, I sat down with Arnaub Chatterjee of Datavant to discuss a topic critical to our work: contextual datasets, in particular, mortality data. We talked about how our mission at Veritas Data Research is to make this data not just accessible, but truly useful for life sciences. Our partnership with Datavant, a leader in secure data sharing, is a significant advancement in this effort. 

As Arnaub noted, we’re working together “so that more organizations can better discover, or evaluate, or link to mortality data in a privacy-compliant way,” ultimately empowering the industry to “answer critical questions with much more confidence.” 

I’ve always believed that mortality should be a key endpoint for healthcare analytics, but historically, it’s been difficult to access. Traditional sources’ utitlity has been challenged with significant gaps and restrictions, creating a substantial blind spot for researchers. As I noted in our conversation, “The Social Security Administration’s Death Master File, once comprehensive, now covers only about 15% of deaths due to state opt-outs.” Similarly, deaths found in EHRs or claims data represent a fraction, maybe only 15% to 20%, of deaths nationally. Alternatively, the CDC’s National Death Index (NDI) is a research-grade data set with great completeness, but it is not commercially-available for industry use. 

At Veritas, we built our data collation workflow from the ground up to address these issues. Through a robust process of collecting data from over 40,000 different sites every week—including funeral homes and crematoriums—we’ve been able to build a data set with over 90% coverage that is accessible and usable for commercial applications. Plus, we can provide this data with remarkable timeliness, with 90% of our data available within two weeks of the date of death. 

Beyond simply reporting the fact of death, we’ve focused on ensuring our data is usable and trustworthy for high-stakes applications. This focus means building transparency around how we consolidate multiple records from different sources, and ensuring thatour data is also fully traceable, which is vital for use as regulatory-grade evidence. As I noted to Arnaub, we’re “big believers in the FAIR principles and FDA’s emerging guidance on traceability and provenance of data,” which is why every one of our records can be traced back to its originating source, whether that’s an obituary, cemetery, or a government disclosure. 

The availability of complete, timely, and traceable mortality data is unlocking new applications that were previously out of reach. While this data’s role is well-understood for use cases like survival analysis in Health Economics and Outcomes Research (HEOR), its role in clinical trials and operational efficiencies is just beginning to be explored. 

As I explained, “the ability to append mortality data [to clinical trial data] to do survival analyses quickly and efficiently is something I think we’re going to see really start to happen over the next couple years.” This shift is now also being driven by regulators. The FDA recently issued draft guidance encouraging the measurement of overall survival in cancer studies, even if it’s not the primary endpoint. This presents a new burden on the industry, but as I noted, our data can help them “measure that survival endpoint in a much more efficient way.” In that same vein, we’ve seen from studies such as one from UCLA in late 2023¹, that a surprising percentage of patients in health systems’ EHRs are actually deceased, resulting in a lot of wasted time and resources. By having access to timely mortality data, we can help health systems and others prevent insensitive outreach that’s probably not appreciated by the family of the deceased. 

For a deeper dive into these and other insights, you can watch the entire conversation here.