Arthur R. Schmidt’s research while affiliated with University of Illinois Urbana-Champaign and other places

Current modeling practices for environmental and sociological modulated infectious diseases remain inadequate to forecast the risk of outbreak(s) in human populations, partly due to a lack of integration of disciplinary knowledge, limited availability of disease surveillance datasets, and overreliance on compartmental epidemiological modeling methods. Harvesting data knowledge from virus transmission (aerosols) and detection (wastewater) of SARS-CoV-2, a heuristic score-based environmental predictive intelligence system was developed that calculates the risk of COVID-19 in the human population. Seasonal validation of the algorithm was uniquely associated with wastewater surveillance of the virus, providing a lead time of 7–14 days before a county-level outbreak. Using county-scale disease prevalence data from the United States, the algorithm could predict COVID-19 risk with an overall accuracy ranging between 81% and 98%. Similarly, using wastewater surveillance data from Illinois and Maryland, the SARS-CoV-2 detection rate was greater than 80% for 75% of the locations during the same time the risk was predicted to be high. Results suggest the importance of a holistic approach across disciplinary boundaries that can potentially allow anticipatory decision-making policies of saving lives and maximizing the use of available capacity and resources.

Nucleic acid-based assays, such as polymerase chain reaction (PCR), that amplify and detect organism-specific genome sequences are a standard method for infectious disease surveillance. However, challenges arise for virus surveillance because of their genetic diversity. Here, we calculated the variability of nucleotides within the genomes of 10 human viral species in silico and found that endemic viruses exhibit a high percentage of variable nucleotides (e.g., 51.4% for norovirus genogroup II). This genetic diversity led to the variable probability of detection of PCR assays (the proportion of viral sequences that contain the assay’s target sequences divided by the total number of viral sequences). We then experimentally confirmed that the probability of the target sequence detection is indicative of the number of mismatches between PCR assays and norovirus genomes. Next, we developed a degenerate PCR assay that detects 97% of known norovirus genogroup II genome sequences and recognized norovirus in eight clinical samples. By contrast, previously developed assays with 31% and 16% probability of detection had 1.1 and 2.5 mismatches on average, respectively, which negatively impacted RNA quantification. In addition, the two PCR assays with a lower probability of detection also resulted in false negatives for wastewater-based epidemiology. Our findings suggest that the probability of detection serves as a simple metric for evaluating nucleic acid-based assays for genetically diverse virus surveillance. IMPORTANCE Nucleic acid-based assays, such as polymerase chain reaction (PCR), that amplify and detect organism-specific genome sequences are employed widely as a standard method for infectious disease surveillance. However, challenges arise for virus surveillance because of the rapid evolution and genetic variation of viruses. The study analyzed clinical and wastewater samples using multiple PCR assays and found significant performance variation among the PCR assays for genetically diverse norovirus surveillance. This finding suggests that some PCR assays may miss detecting certain virus strains, leading to a compromise in detection sensitivity. To address this issue, we propose a metric called the probability of detection, which can be simply calculated in silico using a code developed in this study, to evaluate nucleic acid-based assays for genetically diverse virus surveillance. This new approach can help improve the sensitivity and accuracy of virus detection, which is crucial for effective infectious disease surveillance and control.

Nucleic acid-based assays, such as polymerase chain reaction (PCR), that amplify and detect organism-specific genome sequences are a standard method for infectious disease surveillance. However, challenges arise for virus surveillance because of their genetic diversity. Here, we calculated the variability of nucleotides within the genomes of ten human viral species in silico and found that endemic viruses exhibit a high percentage of variable nucleotides (e.g., 51.4% for norovirus GII). This genetic diversity led to variable probability of detection of PCR assays (the proportion of viral sequences that contain the assay’s target sequences divided by the total number of viral sequences). We then experimentally confirmed that the probability of the target sequence detection is indicative of the number of mismatches between PCR assays and norovirus genomes. Next, we developed a degenerate PCR assay that detects 97% of known norovirus GII genome sequences and recognized norovirus in eight clinical samples. In contrast, previously developed assays with 31% and 16% probability of detection had 1.1 and 2.5 mismatches on average, respectively, which negatively impacted RNA quantification. Additionally, the two PCR assays with lower probability of detection also resulted in false negatives for wastewater-based epidemiology. Our findings suggest that the probability of detection serves as a simple metric for evaluating nucleic acid-based assays for genetically diverse virus surveillance. Importance Nucleic acid-based assays, such as polymerase chain reaction (PCR), that amplify and detect organism-specific genome sequences are a standard method for infectious disease surveillance. However, challenges arise for virus surveillance because of the rapid evolution and genetic variation of viruses. The study analyzed clinical and wastewater samples using multiple PCR assays and found significant performance variation among the PCR assays for genetically diverse norovirus surveillance. This finding suggests that some PCR assays may miss detecting certain virus strains, leading to a compromise in detection sensitivity. To address this issue, we propose a metric called the probability of detection, which can be simply calculated in silico using a code developed in this study, to evaluate nucleic acid-based assays for genetically diverse virus surveillance. This new approach can help improve the sensitivity and accuracy of virus detection, which is crucial for effective infectious disease surveillance and control.

Rainfall-derived infiltration and inflow (RDII) is extraneous water in a sanitary sewer system that originates from surface runoff. Most RDII enters sanitary sewer systems through illegal connections or mechanical faults, especially in aged sewer systems. In this study, the physical process of three primary RDII sources: roof downspout, sump pump, and leaky lateral, are investigated using physics-based models. These three sources represent three different flow paths: direct connection of impervious catchments, mixed flow through coarse porous media followed by a direct connection, and percolated flow through compacted soil. Due to the differences in medium and the lengths of flow paths, flow responses of these three RDII sources differ in time and magnitude. In turn, they can be distinctly identified from each other. The typical flow response of each RDII source is represented as an impulse response function (IRF), a flow response to a pre-specified representative rainfall computed using physics-based models. The total RDII flow hydrograph is presented as a combination of these three IRFs. The weighting factors of each IRF are calculated using a genetic algorithm technique in a test sewer basin in a suburb of Chicago, IL. The model results suggest leaky lateral might be the biggest RDII contributor to the system. The model performance was compared with one of the more widely used RDII estimation methods, the Storm Water Management Model RTK method. While the RTK method shows better performance overall, the IRF method provides a unique solution with robust performance. The suggested physics-based approach may shed light on identifying local RDII issues with more detail, facilitating more effective management of a sewer system.

Wastewater-based epidemiology (WBE), an emerging approach for community-wide COVID-19 surveillance, was primarily characterized at large sewersheds such as wastewater treatment plants serving a large population. Although informed public health measures can be better implemented for a small population, WBE for neighborhood-scale sewersheds is less studied and not fully understood. This study applied WBE to seven neighborhood-scale sewersheds (average population of 1471) from January to November 2021. Community testing data showed an average of 0.004 % incidence rate in these sewersheds (97 % of monitoring periods reported two or fewer daily infections). In 92 % of sewage samples, SARS-CoV-2 N gene fragments were below the limit of quantification. We statistically determined 10–2.6 as the threshold of the SARS-CoV-2 N gene concentration normalized to pepper mild mottle virus (N/PMMOV) to alert high COVID-19 incidence rate in the studied sewershed. This threshold of N/PMMOV identified neighborhood-scale outbreaks (COVID-19 incidence rate higher than 0.2 %) with 82 % sensitivity and 51 % specificity. Importantly, neighborhood-scale WBE can discern local outbreaks that would not otherwise be identified by city-scale WBE. Our findings suggest that neighborhood-scale WBE is an effective community-wide disease surveillance tool when COVID-19 incidence is maintained at a low level.

Wastewater-based epidemiology (WBE), an emerging approach for community-wide COVID-19 surveillance, was primarily characterized at large sewersheds such as wastewater treatment plants serving a large population. Although informed public health measures can be better implemented for a small population, WBE for neighborhood-scale sewersheds is less studied and not fully understood. This study applied WBE to seven neighborhood-scale sewersheds (average population of 1,471) from January to November, 2021. Community testing data showed an average of 0.004% incidence rate in these sewersheds (97% of monitoring periods reported two or fewer daily infections). In 92% of sewage samples, SARS-CoV-2 N gene fragments were below the limit of quantification. We statistically determined 10 -2.6 as the threshold of the SARS-CoV-2 N gene concentration normalized to pepper mild mottle virus (N/PMMOV) to alert high COVID-19 incidence rate in the studied sewershed. This threshold of N/PMMOV identified neighborhood-scale outbreaks (COVID-19 incidence rate higher than 0.2%) with 82% sensitivity and 51% specificity. Importantly, neighborhood-scale WBE can discern local outbreaks that would not otherwise be identified by city-scale WBE. Our findings suggest that neighborhood-scale WBE is an effective community-wide disease surveillance tool when COVID-19 incidence is maintained at a low level. Graphical abstract