Skip to main content

Pig farmers’ willingness to share health information has made it possible to develop accurate forecasts of porcine epidemic diarrhea virus (PEDV) spread within a decision-making timeframe. New research funded by the Swine Health Information Center (SHIC) at North Carolina State University indicates predictability of this forecasting depends on the stage of the spread within the study region.

Being able to forecast disease outbreaks before they occur allows specific control strategies to be tailored to farms in the short-term and the long-term, offering an opportunity to prevent infection. In this study, researchers developed a way to forecast PEDV outbreaks by generating weekly high-resolution maps to track spread and identify current and future PEDV high-risk areas.

“After PEDV broke in 2013, it opened up a whole new avenue of information sharing,” says Paul Sundberg, SHIC executive director. “There was less concern about privacy issues and confidentiality. The industry realized if we share information, we’ll all be stronger. If we take advantage of that willingness to share information, then the question becomes, what do you do with it?”

This project is one of the outcomes of that question, Sundberg says.

A Look at the Study

Three epidemiological transmission models were fully developed, calibrated to true infections and compared: a) a novel epidemiological framework called PigSpread developed specifically to model disease movement in swine populations, b) SimInf, a previously developed framework that has been used more broadly to model disease transmission a national level in EU and Brazil, and c) PoPS (Pest or Pathogen Spread), a framework for modeling the spread of pests or pathogens across a landscape.

The models were calibrated on true weekly PEDV outbreaks from three spatially related swine companies within the study region. Model outputs had general agreement with observed outbreaks throughout the study period with some variability between models, the research shows.

The analysis estimates of the combined strategies of herd closure, feedback, and reinforcement of on-farm biosecurity reduced the incidence of outbreaks in sow farms by 14% and in gilt development units (GDU) by 20% when deployed weekly in sow and GDU farms located in risk areas.

“One of the things that Dr. Machado found in this project, is that he was able to predict with some certainty, outbreaks of PEDV within a region,” Sundberg says. “He took the characteristics of the farm and put them together such that he could correlate some of those characteristics with an outbreak of PEDV historically. And if you do that historically, then you can use that same information to look forward and say, if these things are going on, you’re likely to have an outbreak.”

There are multiple programs that can make these predictions. One of the advantages of this project was that it compared different programs and their ability to be able to actually predict an outbreak, Sundberg says.

“We’ve learned we need to switch between models to make predictions every week,” Machado says. “They have different performances as the disease evolves over time and different assumptions.”

Disease Outbreak Forecasts

Within the outbreak models, the researchers then tested a combination of strategies that might reduce between-farm transmission, knowing this is key to maintain control of outbreaks while minimizing production disruptions.

“We know the most important route of transmission every week, which allows us to focus on trying to break transmission based on the most predominant route of transmission,” says Gustavo Machado, assistant professor in the Department of Population Health and Pathobiology in the College of Veterinary Medicine at North Carolina State University.

Producers can anticipate what’s going to be needed for the next epidemic, Machado adds.

“We were able to build a [computer] model where we can actually employ intervention,” he says. “Now we want to deploy the intervention, such as blind feedback to the farms, which is exposing the pigs to the circulating virus and see how that reduces their transmission. We can manipulate the biosecurity – we can enhance or decrease biosecurity and see how that facilitates the containment and spread.”

For example, in the figure below, Machado says combining the current control strategies used in the field can have a great impact in reducing the transmission of PEDV.

Figure 6. The impact of herd closure, feedback and enhancement of on-farm biosecurity on the average number of PEDV outbreaks for each farm sow (a), GDU (b), nursery (c), and finisher farms (d). The y-axis represents the number of PEDV outbreaks and the x-axis the effectiveness of feedback ranges from no applied (0%) to a range of 80%-95% efficacy. Horizontal dashed lines show the proportion of PEDV outbreak reductions based on the average number of outbreaks from the baseline scenario with no control. Galvis, J. et al, 2020, submitted.

What’s Ahead?

Sundberg says the industry is trying to learn how to walk before it runs when it comes to this area of research.

“These types of analyses teach us how to walk and how to use the information to the betterment of the producer,” Sundberg says. “As we learn better how to use that, we’ll be able to better apply it to more and more things and maybe even to diseases like PRRS, which are really costly for producers.”

Machado says these study results open up a whole new dimension where you can test your interventions and treatments in the computer before you actually do clinical research or be out in the field directly to see the impact.

“We can test those things in the modeling that we have, and then select the things that we really want to try in the field, such as a vaccine. A new vaccine comes into the market, we can test them in the computer first instead of going to the field and doing challenges,” Machado says.

SHIC is funding several projects focusing on different angles related to this study. He says they want to investigate the ability of these programs being able to apply this kind of information analysis more broadly, to include other diseases and do an even better job of prediction.

“Stay tuned – this is just the start of this effort. We’re just on the cusp of the ability to be able to predict disease. If we can do this with PEDV, can it help them on the farm and within the timeframe that they could actually make a difference and prevent it? One of the exciting things is if we can do this with PEDV, we may be able to extend it to other things as well,” Sundberg says.

Farm Journal’s Pork | Jennifer Shike | October 14,

Image Source –