Federated epidemic surveillance

Epidemic surveillance is a challenging task, especially when crucial data is fragmented across institutions and data custodians are unable or unwilling to share it. This study aims to explore the feasibility of a simple federated surveillance approach. We conduct hypothesis tests on count data behind each custodian’s firewall and then combine p-values from these tests using techniques from meta-analysis. We propose a hypothesis testing framework to identify surges in epidemic-related data streams and conduct experiments on real and semi-synthetic data to assess the power of different p-value combination methods to detect surges without needing to combine or share the underlying counts. Our findings show that relatively simple combination methods achieve a high degree of fidelity and suggest that infectious disease outbreaks can be detected without needing to share or even aggregate data across institutions.Author summary: Timely and trustworthy epidemic surveillance requires data which is often fragmented across institutions that cannot or will not share it for reasons of privacy, regulatory restrictions, competition, etc. We show that infectious disease outbreaks can be detected without sharing any raw data at all. To accomplish this, we introduce federated surveillance: a method for pushing the computation behind these custodians’ firewalls, identifying sufficient statistics that can be shared to test for the presence of an outbreak without requiring access to any sensitive information. Across a variety of settings, federated surveillance provides nearly the same ability to detect outbreaks as fully centralized data, and significantly more than could be achieved via any single data source in isolation. Our results show this more readily implementable form of data sharing can provide substantial value for future pandemic preparedness.