Monitoring emerging pathogens using negative nucleic acid test results from endemic pathogens in pig populations: Application to porcine enteric coronaviruses

Abstract

This study evaluated the use of endemic enteric coronaviruses polymerase chain reaction (PCR)-negative testing results as an alternative approach to detect the emergence of animal health threats with similar clinical diseases presentation. This retrospective study, conducted in the United States, used PCR-negative testing results from porcine samples tested at six veterinary diagnostic laboratories. As a proof of concept, the database was first searched for transmissible gastroenteritis virus (TGEV) negative submissions between January 1st, 2010, through April 29th, 2013, when the first porcine epidemic diarrhea virus (PEDV) case was diagnosed. Secondly, TGEV- and PEDV-negative submissions were used to detect the porcine delta coronavirus (PDCoV) emergence in 2014. Lastly, encountered best detection algorithms were implemented to prospectively monitor the 2023 enteric coronavirus-negative submissions. Time series (weekly TGEV-negative counts) and Seasonal Autoregressive-Integrated Moving-Average (SARIMA) were used to control for outliers, trends, and seasonality. The SARIMA's fitted and residuals were then subjected to anomaly detection algorithms (EARS, EWMA, CUSUM, Farrington) to identify alarms, defined as weeks of higher TGEV-negativity than what was predicted by models preceding the PEDV emergence. The best-performing detection algorithms had the lowest false alarms (number of alarms detected during the baseline) and highest time to detect (number of weeks between the first alarm and PEDV emergence). The best-performing detection algorithms were CUSUM, EWMA, and Farrington flexible using SARIMA fitted values, having a lower false alarm rate and identified alarms 4 to 17 weeks before PEDV and PDCoV emergences. No alarms were identified in the 2023 enteric negative testing results. The negative-based monitoring system functioned in the case of PEDV propagating epidemic and in the presence of a concurrent propagating epidemic with the PDCoV emergence. It demonstrated its applicability as an additional tool for diagnostic data monitoring of emergent pathogens having similar clinical disease as the monitored endemic pathogens.

Fig 1. Time-series of transmissible gastroenteritis virus (TGEV) polymerase chain reaction (PCR) results from 2010 through 2014.

The black line represents the number of submissions with TGEV PCR-negative results while the red line represents the number of submissions with a TGEV PCR-positive result. The blue rectangle indicates the first peak on TGEV-negative submissions on week of April 8^th, 2013.

Fig 2. Time-series of transmissible gastroenteritis virus (TGEV) polymerase chain reaction (PCR)-negative results in relation to porcine epidemic diarrhea virus (PEDV) PCR-positive results from 2012 through 2014.

The black line represents the number of submissions with TGEV PCR-negative results while the red line represents the number of submissions with at least one PEDV PCR-positive result. The blue square represents the weekly time points (April 8^th, 2013). in which a potential signal of increased TGEV-negative submissions occurred signalizing the PEDV emergence in the United State in 2013 based on PCR results (April 29^th, 2013).

Fig 3. Anomaly detection algorithms (alarms represented by the colored dots) using the transmissible gastroenteritis virus (TGEV) polymerase chain reaction (PCR)-negative submissions (black line) in relation to porcine epidemic diarrhea virus (PEDV) PCR-positive submissions (red line).

Seasonal Autoregressive-Integrated Moving-Average (SARIMA) TGEV-negative fitted values (green line), SARIMA TGEV-negative residuals values (pink line).

Fig 4. Time-series of transmissible gastroenteritis virus (TGEV) and porcine epidemic diarrhea virus (PEDV) polymerase chain reaction (PCR)-negative in relation to delta coronavirus (PDCoV) PCR-positive results from 2013 through 2014.

The black line represents the number of submissions that included TGEV and PEDV PCR-negative results, the blue line represents the number of submissions that included at least one PDCoV PCR-positive result.

Fig 5. Anomaly detection algorithms (alarms represented by the colored dots) using the transmissible gastroenteritis virus (TGEV) and porcine epidemic diarrhea virus (PEDV) polymerase chain reaction (PCR)-negative submissions (black line) in relation to porcine delta coronavirus (PDCoV) PCR-positive submissions (blue line).

Seasonal Autoregressive-Integrated Moving-Average (SARIMA) fitted values (green line), and SARIMA residuals values (pink line).

Fig 6. Distribution of enteric polymerase chain reaction (PCR)-negative submissions (black line) in 2023, Seasonal Autoregressive-Integrated Moving-Average (SARIMA) enteric-negative fitted (green line), and SARIMA enteric-negative residuals (pink line).