Aligning SARS-CoV-2 Indicators via an Epidemic Model: Application to Hospital Admissions and RNA Detection in Sewage Sludge
Preprint
Abstract
Ascertaining the state of coronavirus outbreaks is crucial for public health decision-making. Absent repeated representative viral test samples in the population, public health officials and researchers alike have relied on lagging indicators of infection to make inferences about the direction of the outbreak and attendant policy decisions. Recently researchers have shown that SARS-CoV-2 RNA can be detected in municipal sewage sludge with measured RNA concentrations rising and falling suggestively in the shape of an epidemic curve while providing an earlier signal of infection than hospital admissions data. The present paper presents a SARS-CoV-2 epidemic model to serve as a basis for estimating the incidence of infection, and shows mathematically how modeled transmission dynamics translate into infection indicators by incorporating probability distributions for indicator-specific time lags from infection. Hospital admissions and SARS-CoV-2 RNA in municipal sewage sludge are simultaneously modeled via maximum likelihood scaling to the underlying transmission model. The results demonstrate that both data series plausibly follow from the transmission model specified and provide a 95% confidence interval estimate of the reproductive number R 0 ≈ 2.4 ±0.2. Sensitivity analysis accounting for alternative lag distributions from infection until hospitalization and sludge RNA concentration respectively suggests that the detection of viral RNA in sewage sludge leads hospital admissions by 3 to 5 days on average. The analysis suggests that stay-at-home restrictions plausibly removed 89% of the population from the risk of infection with the remaining 11% exposed to an unmitigated outbreak that infected 9.3% of the total population. Highlights: A maximum likelihood method for aligning observed lagged epidemic indicators via an underlying transmission model is derived and illustrated using observed COVID-19 hospital admissions and SARS-CoV-2 RNA concentrations measured in sewage sludge to model a local SARS-CoV-2 outbreak The method enables direct estimation of the reproductive number R 0 from the observed indicators along with the initial prevalence of SARS-CoV-2 infection in the population at risk The analysis suggests tracking SARS-CoV-2 RNA concentration in sewage sludge provides a 3 to 5 day lead time over tracking hospital admissions, consistent with purely statistical time series analysis previously reported The model enables estimation of the fraction of the population compliant with government-mandated stay-at-home restrictions, the size of the exposed population, and the fraction of the population infected with SARS-CoV-2 over the outbreakKeywords
All Related Versions
- Published version: Health Care Management Science, 24 (2), 320.
This publication has 13 references indexed in Scilit:
- Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States, March 23-May 12, 2020JAMA Internal Medicine, 2020
- Reconciling early-outbreak estimates of the basic reproductive number and its uncertainty: framework and applications to the novel coronavirus (SARS-CoV-2) outbreakJournal of The Royal Society Interface, 2020
- OM Forum—COVID-19 Scratch Models to Support Local DecisionsManufacturing & Service Operations Management, 2020
- Logistics of Aggressive Community Screening for Coronavirus 2019JAMA Health Forum, 2020
- Incidence, clinical outcomes, and transmission dynamics of hospitalized 2019 coronavirus disease among 9,596,321 individuals residing in California and Washington, United States: a prospective cohort studymedRxiv, 2020
- Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus–Infected PneumoniaNew England Journal of Medicine, 2020
- Estimation in emerging epidemics: biases and remediesJournal of The Royal Society Interface, 2019
- Intrinsic and realized generation intervals in infectious-disease transmissionProceedings Of The Royal Society B-Biological Sciences, 2015
- The concept ofRoin epidemic theoryStatistica Neerlandica, 1996
- Theoretical Statistics1974