Interview, Fireside Chat, Podcast
a16z Podcast | Dark Data in Healthcare
- The healthcare system is expected to undergo structural transformation driven by the aging baby boomer demographic and a cultural shift where future generations demand immediate, mobile-accessible data similar to consumer payment experiences, moving away from reliance on oral histories or episodic encounters.
- Longitudinal data is projected to become increasingly critical for managing chronic conditions and patients living longer (from an average of 48 to 83 years), necessitating solutions that address data portability and the expiration of records in current patient portals which fail to maintain 80-year histories.
- A "data refinery" is anticipated to convert unstructured documents like PDFs and XML into machine-readable formats, enabling algorithms to automate clinical trial matching and reduce the current failure rate where upwards of 90% of trials go unfilled.
- Real-world evidence is predicted to gain significant utility for payers to validate drug efficacy beyond FDA approval and for identifying trends in crises such as the opioid epidemic, potentially becoming more valuable than traditional clinical trials for determining outcomes.
- Friction in data retrieval is expected to decrease from 30 minutes to a third of a second, allowing doctors to utilize patient-provided information and facilitating an "industrial strength" version of patient-created data that could rescue therapies currently stalled due to data issues.
- The environment is forecast to enable "permissionless innovation" where users build on data platforms with unpredictable use cases, potentially forming coalitions to crowdsource specific clinical trial requests and establishing peer-to-peer care ecosystems for chronic condition management.
- Significant risks include the fragmentation of digital data outstripping EHR vendor consolidation, the lack of institutional responsibility for maintaining long-term patient records, and the uncertainty regarding the specific outcomes and engagement types when previously siloed data is fully freed.
- Patients are expected to become the legally, ethically, and morally incentivized entities responsible for aggregating their fragmented health information, as the system moves from acute to chronic care and institutions fail to keep records current.