Estimating the number of infections using wastewater surveillance data

Introduction Severe acute respiratory syndrome coronavirus (SARS-CoV-2) has resulted in one of the worst pandemics in history. As of September 24th, 2021, there were more than 231 million cases of coronavirus 2019 disease (COVID-19) worldwide, with over 4.7 million deaths. A primary challenge associated with this novel virus is that a significant proportion of COVID-19 goes undetected by the healthcare surveillance system due to the propensity for mild symptoms or asymptomatic conditions, resulting in infected individuals not being tested. SARS-CoV-2 primarily infects the respiratory system but is also commonly detected in the gastrointestinal tract. As a result, infected individuals can excrete SARS-CoV-2 in their stool. The fraction of individual who shed the virus in stool has varied across studies between 45% to 67%. Viral shedding in stool results in genetic material from the virus being present in wastewater and can be quantified as a community-scale indicator of infection dynamics. Surveilling SARS-CoV-2 in wastewater is a cost-effective means to track the spread of COVID-19. Results suggests that SARS-CoV-2 viral prevalence in wastewater can be a leading indicator of clinical COVID-19 cases. Most of the previous studies indicated that wastewater surveillance data has a 2-7 days lead time compared to clinical data. The next step in wastewater-based epidemiology is to use wastewater surveillance data to predict the number of infections in a community. There are different models and equations that have been developed to estimate number of infections using wastewater SARS-CoV-2 concentration data. The mass balance equation is one of the most common models to predict the number of infected individuals for a sewer shed (Eq. 1): Number of infections = (Viral load in wastewater — Wastewater flow rate)/(Production of stool per capita — Viral load in stool — % Individual who shed the virus in their stool) (Eq.1) Viral load in wastewater (viral copies/L), and wastewater flow rate (L/day) data can be calculated easily during wastewater surveillance projects. The data for daily production of stool per capita (g stool/capita-day), viral load in stool (virus copies/g stool), and fraction of SARS CoV-2 infected individual who shed the virus in their stool are referenced in the literature. Among these factors, a major source of uncertainty is the viral load in stool (shedding rate of SARS-CoV-2), which varies over the course of disease and varies between individuals who react differently to viral infection. In the present study, we aim to investigate the viral load factor and better understand this uncertainty. We will use Monte Carlo Simulation to take all these uncertainties into account to retroactively predict infection rates. To validate the results of this study, we will implement the mass balance model in a few wastewater surveillance projects throughout the US to estimate the number of infections in both small and large communities. Mass balance model critical factors Daily production of stool: Rose et al. have done a comprehensive review on the characteristics of feces in different countries. Table 1 shows their findings for feces in high-income countries. Baes on this information (high Skewness) the distribution of daily stool production in high-income countries most closely resembles a log-normal distribution. To confirm this assumption, we will use actual data from Rose et al. study and fit several distributions to determine the best fit. We will then use the fitted distribution in the mass balance model to predict the number of infections different communities using Monte Carlo simulation. SARS-CoV-2 load in stool: The major point of uncertainty in all developed methods to estimate the number of infections using wastewater surveillance data is the rate in which individuals shed the virus in their stool. There are few studies that have quantified the concentration of SARS-CoV-2 in stool of patients over time (see Figure 1). To use this data in our mass equation model and Monte Carlo Simulation, we will first determine the data distribution using statistical software (e.g., Minitab). After finding the best distribution model for daily production of stool, and shedding rate of SARS-Cov-2 in feces, we will execute the mass balance model (Eq. 1) using Monte Carlo Simulation with 10,000 iterations to predict infection rate in different communities. We will then compare our results to the clinical data in the local communities to estimate the accuracy of our model. As mentioned previously, the mass balance equation has a few limitations including: not considering the degradation rate of SARS-CoV-2 in sewer shed, and considering the non-domestic wastewater flow like stormwater flow, or industrial flow (in case of using the wastewater treatment plant data for wastewater flow). To find the expense of these limitations, we will compare our method (mass balance model) with more sophisticated models that have been developed for predicting the number of infections.