Abstract
Researchers and policymakers have proposed systems to detect novel pathogens earlier than existing surveillance systems by monitoring samples from hospital patients, wastewater, and air travel, in order to mitigate future pandemics. How much benefit would such systems offer? We developed, empirically validated, and mathematically characterized a quantitative model that simulates disease spread and detection time for any given disease and detection system. We find that hospital monitoring could have detected COVID-19 in Wuhan 0.4 weeks earlier than it was actually discovered, at 2,300 cases (standard error: 76 cases) compared to 3,400 (standard error: 161 cases). Wastewater monitoring would not have accelerated COVID-19 detection in Wuhan, but provides benefit in smaller catchments and for asymptomatic or long-incubation diseases like polio or HIV/AIDS. Monitoring of air travel provides little benefit in most scenarios we evaluated. In sum, early detection systems can substantially mitigate some future pandemics, but would not have changed the course of COVID-19.
Introduction
In the wake of COVID-19, global health organizations and governments are proposing multi-billion dollar investments in "pathogen-agnostic" surveillance. The objective is to monitor hospitals, sewage, and airports to catch the next "Disease X" before it explodes into a global crisis. However, there has been a notable lack of quantitative evidence regarding the actual benefit of these systems. How much earlier would they trigger an alarm? Which system is best for which type of virus? This research addresses the need for a predictive framework to guide resource allocation, moving beyond qualitative reasoning to provide a rigorous assessment of pandemic early-warning systems.
The Modeling Framework
To compare these systems, we developed a branching process simulation model that tracks individual infections and the probability of those individuals being "captured" by a specific surveillance system. Each infected person has a probability ($p_{test}$) of being detected. The simulation stops when a specific threshold ($d$) of detected cases is met, accounting for a system-specific detection delay ($g$).
Beyond simulations, we derived a compact formula to estimate the expected number of cases at detection ($E[C]$): $$E[C] \approx \frac{d \times R_0^g}{p_{test}}$$ This formula highlights that detection time is most sensitive to the delay ($g$) and the reproduction number ($R_0$), as they scale exponentially. We validated the model by retrospectively predicting the detection times of the first COVID-19 cases across 50 US states in 2020, achieving a mean absolute error of less than one week.
Pathogen-Surveillance Matching
Our analysis demonstrates that there is no universal "best" system. Instead, the optimal strategy depends heavily on the pathogen's epidemiological profile:
- Hospital Monitoring: Excels for severe diseases with high hospitalization rates, such as Ebola.
- Wastewater Monitoring: Superior for pathogens with high asymptomatic spread or long incubation periods, such as Polio. For diseases like HIV/AIDS, advanced surveillance could detect an outbreak up to 110 weeks earlier than the status quo.
- Air Travel Inefficiency: Monitoring international flights provided almost no benefit in the scenarios evaluated, primarily due to the low probability of a traveler being both infected and shedding enough virus to be detected during a specific flight window.
A critical factor for wastewater is the "dilution effect." In massive cities like Wuhan, the sheer volume of sewage dilutes the viral signal, meaning thousands of people must be infected before the virus is detectable. Wastewater is most effective in smaller communities (under 480,000 people); above this threshold, the clinical "noise" in hospitals often provides a faster signal.
Case Study: COVID-19 in Wuhan
Applying this model to the initial outbreak in Wuhan reveals the "Lead Time Illusion." For a fast-spreading respiratory virus like COVID-19, the lead time provided by advanced systems is surprisingly small. Because cases grow exponentially, the difference between "status quo" detection and "advanced" detection is often just a few days.
Hospital monitoring would have detected the Wuhan outbreak at approximately 2,300 cases—only 0.4 weeks earlier than the actual detection at ~3,400 cases. Wastewater monitoring would have actually lagged behind the actual detection date in a city of that size.
Discussion and Limitations
While early detection is a powerful tool for certain pathogens, its impact on the course of a COVID-like pandemic would have been marginal. Furthermore, early detection only helps if it is followed by a coordinated, politically feasible public health response.
Technical challenges also remain; the model assumes that metagenomic sequencing or multiplex PCR can reliably identify a novel pathogen at low concentrations in complex samples. Future work must also account for the cost-effectiveness of these systems to determine the "dollars per week of lead time" for these technologies. In sum, while early detection systems can substantially mitigate some future pandemics, they are not a silver bullet for all viral threats.
Summary updated