In a recent study posted to medRxiv* preprint server, researchers characterised trends of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA levels in wastewater.

Study: Identifying trends in SARS-CoV-2 RNA in wastewater to infer changing COVID-19 incidence: Effect of sampling frequency. Image Credit: DedMityay/Shutterstock

Wastewater-based epidemiology (WBE) relies on infectious disease marker levels within the wastewater to evaluate disease incidence in the neighborhood. The coronavirus disease 2019 (COVID-19) pandemic has increased attention to WBE. Specifically, SARS-CoV-2 RNA concentrations in wastewater-settled solids correlate well with incident clinical COVID-19 cases (lab-confirmed cases) in the identical sewer shed.

Moreover, trends in wastewater SARS-CoV-2 RNA concentrations precede those in incident COVID-19 cases in communities. Nevertheless, there’s uncertainty in interpreting and using WBE data to assist decision-making. Further, there have been limited efforts regarding monitor trends actively. Usual trend measures, similar to rates of change, easy moving averages, et cetera., provide less attention to trend stability and statistical significance and might often be confusing/misleading.

In regards to the study

In the current study, researchers used every day measurements of SARS-CoV-2 RNA in wastewater to check three trend evaluation methods and evaluate their performance. Day by day sampling from the San José-Santa Clara regional wastewater facility commenced on November 15, 2020, until September 15, 2022. The authors used data on SARS-CoV-2 nucleocapsid (N) gene concentrations in gene copies/g of dry weight normalized by pepper mild mottle virus (PMMoV) concentrations.

Three analytic metrics – percent change (PC), Mann-Kendall (MK) trend test, and relative strength index (RSI), were used to discover trends in N or PMMoV over time. RSI is beneficial to tell trend stability, whereas the opposite two inform about statistical significance. The RSI was calculated for the seven-day right-aligned moving average of N or PMMoV using a 14-day look-back period.

For PC, the authors used the formula that the US Centers for Disease Control and Prevention (CDC) uses for computing trends within the wastewater SARS-CoV-2 RNA concentrations. The MK trend test, a non-parametric test, was used to guage monotonic trends in a time-series dataset. The MK trend test was applied to check raw N or PMMoV data using a 14-day look-back period.

Heatmaps were created for your entire evaluation period and individually for 3 COVID-19 waves attributable to 1) SARS-CoV-2 Delta, 2) Omicron BA.1, and three) Omicron BA.2 and BA.5 variants. Further, the authors down-sampled the dataset to repeat PC and MK trend test methods. The down-sampled dataset was generated for all combos covering frequencies between two samples/week and 6 samples/week.

Wastewater SARS-CoV-2 RNA trends were stratified as increasing, decreasing, or stable using the three metrics. Furthermore, the MK trend test and PC calculated the upward and downward sensitivity and specificity for every down-sampled dataset. Upward/downward sensitivity was defined as appropriately identifying an upward/downward trend, whereas specificity was the power to discover no trend.

Findings

All methods identified increasing, decreasing, and stable trends of SARS-CoV-2 RNA concentrations in wastewater. The MK trend test and PC identified upward trends ahead of RSI at the start of Delta and BA.2/BA.5 waves. Specifically, the MK trend test identified the increasing trend 17 and 12 days before RSI initially of Delta and BA.2/BA.5 waves, respectively.

Likewise, PC identified the increasing trend 16 and 26 days before RSI at the start of the Delta and BA.2/BA.5 waves, respectively. Moreover, the MK trend test and PC identified the downward trends at the tip of the COVID-19 waves before RSI. Within the down-sampling analyses, each the MK trend test and PC had poor (upward/downward) sensitivity for low sampling frequencies but improved with the rise in frequency.

The MK trend test attained acceptable upward/downward sensitivity with a minimum of 5 samples/week. Alternatively, PC reached acceptable upward/downward sensitivity with at the very least 4 samples/week. Notably, specificity remained similar and really high for each methods in any respect sampling frequencies.

Conclusions

The MK trend test and PC provided more early warnings of upward and downward trends using every day data than the RSI. Down-sampling analyses suggested that a sampling frequency of at the very least 4 and five samples per week was mandatory to discover trends using PC and the MK trend test, respectively. Since PC required fewer samples/week than the MK trend test, WBE programs with budget constraints may prefer PC.

Overall, the MK trend test and PC are inference-based methods and will be used to categorise trends in a regular manner. These trend evaluation approaches could also be adopted by WBE programs to tell public health departments on how COVID-19 case trends change, particularly because the rate of clinical SARS-CoV-2 testing declines.

*Essential notice

medRxiv publishes preliminary scientific reports that usually are not peer-reviewed and, due to this fact, mustn’t be considered conclusive, guide clinical practice/health-related behavior, or treated as established information.