Search tips
Search criteria 


Logo of canvetjReference to the Publisher site.Journal Web siteJournal Web siteHow to Submit
Can Vet J. 2016 January; 57(1): 84–89.
PMCID: PMC4677617

Language: English | French

Outbreak investigation of porcine epidemic diarrhea in swine in Ontario


Porcine epidemic diarrhea virus was first diagnosed in Ontario in January of 2014. An outbreak investigation was conducted and it was hypothesized that feed containing spray-dried porcine plasma contaminated with the virus was a risk factor in the introduction and spread of the disease in Ontario.


Enquête sur l’éclosion de la diarrhée porcine épidémique chez les porcs de l’Ontario. Le virus de la diarrhée épidémique porcine a été diagnostiqué pour la première fois en Ontario en janvier 2014. Une enquête a été réalisée sur l’éclosion et on a émis l’hypothèse que les aliments contenant du plasma porcin séché par atomisation contaminé par le virus représentait un facteur de risque pour l’introduction et la propagation de la maladie en Ontario.

(Traduit par Isabelle Vallières)

In January of 2014, porcine epidemic diarrhea virus (PEDV) was diagnosed in a swine herd in southwestern Ontario (1,2). Porcine epidemic diarrhea had not been previously diagnosed in Canada but was reported in the United States in May of 2013. Following the initial discovery, another 62 cases of PEDV were diagnosed between January 22, 2014 and July 21, 2014.

An outbreak investigation was conducted by veterinarians at the Ontario Ministry of Agriculture, Food and Rural Affairs to describe the outbreak of PEDV in Ontario and to generate hypotheses regarding the origin and spread of the outbreak. The investigation was conducted between January 23 and July 21, 2014, and consisted of 3 components: an outbreak investigation survey of cases with an in-depth investigation into the first 6 cases; environmental surveillance from transport trailers, processing plants, and an assembly yard; and feed testing based on results of the survey of the initial cases.

A case was defined as a farm in Ontario diagnosed with PEDV by reverse transcriptase-polymerase chain reaction (RT-PCR) on samples originating from animals. Porcine delta coronavirus was also diagnosed concurrently with the outbreak of PEDV; however, this report is limited to cases of PEDV.

Outbreak investigation survey

The outbreak investigation survey (available upon request) was designed based on a similar survey conducted in the United States (3). The survey collected information pertaining to the premises, characteristics of the facility and the herd, feed sources, and clinical presentation of the disease. Records of movements were compiled for animals, humans, feed, equipment, and other potential fomites moving on and off the farm during the 2-week period prior to the onset of clinical signs of disease or a positive PEDV RT-PCR test. Veterinarians, either government, private or both, interviewed producers using the survey.

Information was compiled on 25 of the first 28 cases that occurred from January 22 to February 28, 2014 (Table 1). In 3 of the 28 cases there was no response to requests for an interview. Due to limited human resources, information collected on the 3 non-respondents and the next 35 cases was limited to date of diagnosis, county, and production type. The next 35 cases occurred from March 4 to July 21, 2014. A summary of the 63 cases by production type is presented in Table 2.

Table 1
Proportion of case farms with and without exposure to the factor, PEDV in Ontario, January 22 to February 28, 2014
Table 2
Production type of case farms, PEDV in Ontario, January 22 to July 21, 2014

Spatiotemporal characteristics

The investigation revealed that the outbreak affected farms across a widespread area and within many counties in Ontario (Figure 1). The largest number of cases was reported in Bruce, Huron, Middlesex, Oxford, and Perth counties.

Figure 1
Distribution of the number of cases by county, PEDV in Ontario, January 22 to July 21, 2014.

Cases were significantly clustered in relation to controls to the 1st and 4th nearest neighbor. Case and control data were loaded into cluster detection and analysis software (ClusterSeer 2.5.1; Biomedware, Ann Arbor, Michigan, USA). A Cuzick and Edwards’ statistic (4) was calculated and the Statistical Distance Test was significant as well as the test for the 1st and 4th nearest neighbor (Table 3), indicating that cases were more likely to neighbor cases than controls. This clustering could reflect spread between neighboring farms (1st nearest neighbor) and spread within a production system as pigs move from farrowing barns to nearby finishing farms (4th nearest neighbor). A similar analysis of the spread of PEDV in Iowa and North Carolina found that cases neighbored cases to the 4th and 20th nearest neighbor, respectively (5). The outbreak had a temporal pattern of spread consistent with a point source with secondary transmission. An epidemic curve based on the date of diagnosis by epidemiological week and production type (finisher versus other) was constructed to visualize the outbreak in time (Figure 2). The epidemic curve had an initial peak and a second peak of cases, suggesting an outbreak with an initial source of exposure followed by another vehicle of transmission resulting in secondary cases (6).

Figure 2
Epidemic curve by epidemiological week, PEDV in Ontario, January 22 to July 21, 2014.
Table 3
Summary of spatial and temporal analyses, PEDV in Ontario, January 22 to July 21, 2014

In addition, the outbreak did not cluster in time. Case data were loaded into cluster detection and analysis software (ClusterSeer 2.5.1). Clustering in time was assessed using a modified cusum test (7) using a relative risk threshold of 1.1, a significance level of 0.05, and a series of 999 Monte Carlo simulations to determine the statistical significance. Three temporal clusters were detected but these were non-significant (Table 3).

Clinical presentation

Sows and piglets in the farrowing rooms were most severely affected by the disease, although clinical signs were variable across farms. Average morbidity and mortality rates were calculated based on data reported by producers who were interviewed (Table 1). Sows presented with clinical signs of anorexia, diarrhea, occasional vomiting and negligible mortality, while nursing piglets had clinical signs of diarrhea, vomiting, and an average mortality of 53%. Older animals, including starters, growers, and finishers, had mild clinical signs of disease including anorexia and diarrhea with an average morbidity of 29% to 40% and average mortality of 1% to 2%. Onset of the disease was variable; most farms reported a sudden onset of clinical signs although a few reported gradual onset. On 1 finisher farm, animals co-infected with other diseases experienced an increased mortality of 5%.

On average, herds recovered from the illness approximately 3 wk after onset and took longer to recover when ground intestinal material was used as a method of feedback to help develop herd immunity. A follow-up survey was conducted approximately 1 mo after the onset of clinical signs to determine the clinical progression of and recovery from disease (available upon request). Fifteen producers were questioned regarding clinical signs in their affected herd and the date on which they thought their herd had recovered from the disease. The average duration of illness was 19 d, although 3 producers reported ongoing effects from the outbreak (Table 4). Producers were also questioned on whether they used fecal material or ground intestine as a method of feedback. Herds that used ground intestinal material took almost twice as long to recover from the illness (19.6 d) compared with herds that used fecal material only (10.6 d). This difference was almost statistically significant (P = 0.054), although the sample size was small (n = 10).

Table 4
Results of follow-up survey, PEDV in Ontario, January 22 to February 28, 2014

As the outbreak progressed, more finisher than farrowing farms became affected. Most of the first 28 cases were farrow-to-finish farms, while the last 35 cases were predominantly finisher farms (Table 2). This could be attributed to a different type of exposure, previously infected animals moving through the production system, breaches in on-farm biosecurity, or vehicles contaminated with PEDV entering premises.

Field investigation

An extensive field investigation of the initial 6 cases did not find any common risk factors associated with farm biosecurity, feed transporters, service providers, a rendering company, or livestock haulers. The private veterinarians involved with these 6 cases reported that the farms had excellent biosecurity. Fifteen feed delivery trucks from 3 feed companies were swabbed in 5 places, including the cab step, steering wheel and cab interior, floor mats and pedals, wheel wells, and blower pipe. All of these samples tested negative for PEDV. Two service providers were on the index farm approximately 7 d prior to the initial diagnosis of PEDV. The equipment taken onto the premises was swabbed and tested negative for PEDV. No service providers were reported to have entered the premises of the other 5 cases in the 2 wk prior to the onset of disease. A vehicle belonging to a rendering company was in contact with 10 farms following contact with the index case; this vehicle was swabbed and all samples were negative on testing for PEDV. All farms in contact with this vehicle were notified and asked to monitor for signs of PEDV; no clinical signs consistent with PEDV developed in these herds. Livestock trailers in contact with the first 4 cases were sampled and were negative on tests for PEDV.

Environmental surveillance

Surveillance testing of hog transport and processing networks demonstrated that the virus was not highly prevalent across Ontario. To determine the extent and spread of the virus throughout Ontario, environmental samples were taken between January 24 and March 17, 2014 from transport trailers and hog processing plants handling approximately 85% of the hogs marketed in the province. Trailers were sampled using a protocol modified from one validated for detection of porcine reproductive and respiratory virus (8). After the hogs were unloaded, trailers were swabbed using a cloth (Swiffer Sweeper Dry Unscented Sweeping Cloth, Procter and Gamble Company, Toronto, Ontario) soaked in a 90% saline/10% propylene glycol solution. The solution was extracted from the cloth and tested for PEDV using the RT-PCR test. Only 7.5% of 1555 trailers tested had samples positive for PEDV. Producers who shipped hogs on contaminated trailers were contacted and on follow-up testing 3 farms were diagnosed with PEDV.

Environmental sampling, as described previously, was conducted on January 24, 2014 at an assembly yard that had contact with 8 out of 25 cases within 2 wk prior to the outbreak and PEDV was detected in all 10 samples collected. It is not possible to determine the contribution of this contact, if any, to the initial infection with PEDV. This is attributed to the structure of the transport industry in which animals are often congregated at a site prior to movement to a larger yard. This structure provides ample opportunity for mixing of animals from various sources and transmission of pathogens. Further tracing activities were not performed at the assembly yard. In the United States, it was suggested that the transport process (including collection points and vehicles) can be a source of transmission of PEDV if adequate hygiene measures are not implemented (9).

Feed investigation

It was hypothesized that spray-dried porcine plasma (SDPP) contaminated with PEDV was a risk factor for the introduction and spread of PEDV in Ontario. The investigation found that 21 out of the 25 cases investigated were associated with shipments of creep or nursery feeds from a single feed company. Samples from these feeds and a batch of imported SDPP used to manufacture the feed were sampled and were positive for PEDV on RT-PCR testing. The feed also tested positive on RT-PCR testing for transmissible gastroenteritis virus, porcine respiratory coronavirus and the North American strain of porcine reproductive and respiratory virus. The Canadian Food Inspection Agency (CFIA) was notified and subsequently sampled the SDPP and feed, which were sent to the National Centre for Foreign Animal Disease in Winnipeg for testing. Both the SDPP and the feed contained PEDV genetic material, as determined by RT-PCR (10). The CFIA also conducted swine bioassays and found that the SDPP blood plasma contaminated with PEDV was capable of infecting inoculated piglets; however, they could not infect piglets with PEDV using the complete feed containing the PEDV-contaminated SDPP (10). Nevertheless, the CFIA’s epidemiological assessment provided support for a link between pelleted swine feed containing a contaminated lot of imported SDPP, and the outbreak of PEDV that started in Ontario in January 2014 (Pascale Aubry, manuscript in preparation). However, the role, if any, of spray-dried porcine plasma or pelleted swine feed in general in the epidemiology of PEDV requires further investigation. A recent study demonstrated that complete feed contaminated with PEDV was infective to naïve pigs using natural feeding behavior; however, this feed did not contain any animal by-products and may represent post-processing contamination (11). The feed containing the contaminated SDPP was voluntarily withdrawn from the Canadian market on February 9th, 2014, represented by the dashed line on the epidemic curve (Figure 2). The epidemic curve could suggest an initial peak of cases in non-finisher farms exposed to the contaminated SDPP followed by secondary transmission related to spread between farms. This could also explain the observed shift in the type of hog production system affected (Table 2).

In conclusion, PEDV affected many farms over a widespread area in Ontario without any significant temporal clustering. The disease caused high morbidity and mortality in nursing piglets and had less impact on older animals, although clinical signs were variable across farms. Extensive surveillance of hog transportation and processing networks showed that the virus was not highly prevalent across Ontario at the time of the investigation. An in-depth outbreak investigation of the cases suggested that feed containing PEDV-contaminated spray-dried porcine plasma could be a source of the virus, although a swine bioassay was unable to confirm this link in complete feed.


The authors thank Dr. Sue Burlatschenko of Goshen Ridge Veterinary Services for assisting the outbreak investigation. The authors acknowledge numerous other individuals involved in this investigation, including staff of the Ontario Ministry of Agriculture, Food and Rural Affairs; and the Animal Health Laboratory, University of Guelph. The authors also thank the many producers and industry partners who participated in and provided information for the investigation. CVJ


Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (gro.vmca-amvc@nothguorbh) for additional copies or permission to use this material elsewhere.


1. Kochar HS. Porcine epidemic diarrhea in Canada: An emerging case study. Can Vet J. 2014;55:1048–1049. [PMC free article] [PubMed]
2. Ojkic D, Hazlett M, Fairles J, et al. The first case of porcine epidemic diarrhea in Canada. Can Vet J. 2015;56:149–152. [PMC free article] [PubMed]
3. American Association of Swine Veterinarians. PEDV Investigation 2013 [monograph on the Internet] Perry, Iowa: American Association of Swine Veterinarians; c2013. [Last accessed March 6, 2014]. Available from:
4. Cuzick J, Edwards R. Spatial clustering for inhomogeneous populations. J Royal Stat Soc Series B Method. 1990;52:73–104.
5. Perez A. Spatial epidemiology of PED. Proceedings of the Allen D. Leman Swine Conference 2014; [Last accessed November 6, 2015]. Available from:
6. Fontaine RE, Goodman RA. Describing the findings. In: Gregg MB, editor. Field Epidemiology. 2nd ed. New York, New York: Oxford University Press; 2002. pp. 90–91.
7. Levin B, Kline J. The cusum test of homogeneity with an application in spontaneous abortion epidemiology. Stat Med. 1985;4:469–488. [PubMed]
8. Kenney K, Polson D. Validation of Swiffer® cloth-origin neutralizing broth samples for detection of PRRS virus in the environment. Proceedings of the American Association of Swine Veterinarians; Phoenix, Arizona. 2011; pp. 95–108.
9. Lowe J, Gauger P, Harmon K, et al. Role of transportation in spread of porcine epidemic diarrhea virus infection, United States. Emerg Infect Dis. 2014;20:872–874. [PMC free article] [PubMed]
10. Pasick J, Berhane Y, Ojkic D, et al. Investigation into the role of potentially contaminated feed as a source of the first-detected outbreaks of porcine epidemic diarrhea in Canada. Transbound Emerg Dis. 2014;61:397–410. [PMC free article] [PubMed]
11. Dee S, Clement T, Schelkopf A, et al. An evaluation of contaminated complete feed as a vehicle for porcine epidemic diarrhea virus infection of naïve pigs following consumption via natural feeding behavior: Proof of concept. BMC Vet Res. 2014;10:176. [PMC free article] [PubMed]

Articles from The Canadian Veterinary Journal are provided here courtesy of Canadian Veterinary Medical Association