|Home | About | Journals | Submit | Contact Us | Français|
The purposes of this study were to describe the clinical signs observed in PRRS positive herds during a porcine reproductive and respiratory syndrome (PRRS) outbreak in Ontario and to determine associations between these clinical signs and herd demographics and PRRS control strategies. All PRRS polymerase chain reaction-(PCR)-positive submissions to a diagnostic laboratory between September 1, 2004 and August 31, 2007 were identified (n = 1864). After meeting eligibility requirements and agreeing to voluntary study participation, producers from 455 of these submissions were surveyed for information on clinical signs observed in their herds, herd demographics, and PRRS control strategies used in their herds at the time that the PCR-positive samples were taken. Larger herd size was associated with an increased risk of reporting abortion, weakborn piglets, off-feed sows, and sow mortality in sow herds, and with an increased risk of reporting mortality in finishing herds. When disease control strategies were examined, use of a commercial PRRS vaccine in sows and gilts was associated with a decreased risk of reporting weakborn pigs and high pre-weaning mortality, while the use of serum inoculation in breeding animals was associated with an increased risk of reporting off-feed sows and sow mortality. Providing biofeedback of stillborn/mummified piglets, placenta or feces to gilts was associated with an increased risk of reporting respiratory disease and mortality in finishing pigs while all-in/all-out flow in farrowing rooms was associated with an increased risk of reporting sow mortality and weakborn piglets.
L’objectif de la présente étude était de décrire les signes cliniques observés dans les troupeaux positifs pour le syndrome respiratoire et reproducteur porcin (PRRS) lors d’une poussée de cas de PRRS en Ontario et de déterminer les associations entre ces signes cliniques et les données démographiques de ces troupeaux ainsi que les stratégies de contrôle du PRRS. Entre le 1er septembre 2004 et le 31 août 2007, toutes les soumissions à un laboratoire de diagnostic positives par réaction d’amplification en chaîne par la polymérase (PCR) pour la présence de PRRS ont été identifiées (n = 1864). Après avoir rencontrés tous les critères d’éligibilité et avoir acceptés de participer volontairement à l’étude, les producteurs associés à 455 de ces soumissions ont été recensés pour des informations sur les signes cliniques observés dans les troupeaux, les données démographiques du troupeau, et les stratégies utilisées dans leurs troupeaux pour limiter le PRRS au moment où les échantillons positifs par PCR ont été prélevés. Une taille de troupeau plus grande était associée avec un risque augmenté de rapporter des avortements, des porcelets faibles à la naissance, des truies sans appétit, et de la mortalité de truie dans les troupeaux de truies, ainsi qu’un risque plus grand de rapporter de la mortalité dans les troupeaux de porcs en finition. Lorsque les stratégies de contrôle de la maladie étaient examinées, l’utilisation d’un vaccin commercial contre le PRRS chez les truies et les cochettes était associée avec un risque diminué de rapporter des porcelets faibles à la naissance et une mortalité pré-sevrage élevée, alors que l’utilisation d’inoculation de sérum chez les animaux reproducteurs était associée avec un risque augmenté de rapporter des truies sans appétit et de la mortalité chez les truies. Une bio-rétroaction de porcelets momifiés/mort-nés, de placenta ou de fèces à des cochettes était associée avec un risque augmenté de rapporter des problèmes respiratoires et de la mortalité chez les porcs en finition, alors que la pratique du tout-plein/tout-vide dans les chambres de natalité était associée avec un risque augmenté de rapporter de la mortalité chez des truies et des porcelets faibles à la naissance.
(Traduit par Docteur Serge Messier)
In the late 1980’s, outbreaks of a new disease causing pneumonia, poor growth, and increased mortality in growing pigs and reproductive losses in sow herds were reported in the United States and Europe (1–3). A viral etiologic agent was eventually identified and the disease it caused became known as porcine reproductive and respiratory syndrome (PRRS) (4,5). Today, this syndrome is one of the most economically important diseases affecting the swine industry worldwide. For example, in 2005, the projected cost of PRRS to the American swine industry was estimated to be about $560 million (US)/y (6).
In Ontario, the first reported outbreaks of a PRRS-like disease occurred in the late summer and early fall of 1987 (7). Subsequent analysis of stored sera collected from pigs between 1978 and 1982 showed PRRS virus antibody had been present in Ontario herds as early as 1979 (7). The virus spread rapidly throughout Ontario and it has since been estimated that between 40% to 80% of herds in central Canada are seropositive (7).
In the fall of 2004 and winter of 2005, swine veterinarians reported seeing an increase in the prevalence of PRRS in southwestern Ontario and an increase in the severity of clinical signs in these outbreaks. This epidemic was also detected at the Animal Health Laboratory (AHL) at the University of Guelph (Guelph, Ontario) when both the number of cases submitted for PRRS polymerase chain reaction (PCR) testing and the total number of cases testing PCR-positive for PRRS virus increased (8). The purpose of this paper is to describe the clinical signs observed in PRRS positive herds during the outbreak and to determine the association between these clinical signs and herd demographics and certain disease control strategies.
The study was designed to have both a retrospective and prospective component. For the retrospective aspect of the study, the AHL database [Vetstar Animal Disease Diagnostic System (VADDS); Advanced Technology Corporation, Ramsey, New Jersey, USA] was searched to find all retrospective cases with at least 1 positive result from PRRS reverse-transcriptase polymerase chain reaction (RT-PCR) testing on serum or tissue samples submitted to the laboratory between September 1, 2004 and January14, 2006. This test has been previously described (9). September 1, 2004 was chosen as the beginning date as this was when swine veterinarians had begun to see increases in the number of PRRS outbreaks in Ontario. January 14, 2006 marked the end of the retrospective portion of the study and the beginning of the prospective portion of the study.
In total, 860 cases with at least 1 PRRS PCR-positive sample were submitted to the AHL from September 1, 2004 to January 14, 2006. The reasons for submitting samples to the laboratory were not known but it was assumed that some submissions were prompted by clinical problems while others were for PRRS monitoring purposes. Four hundred positive cases were selected for further investigation. Included in the 400 cases were all samples that had been genetically sequenced at the request of the submitting veterinarian (n = 179) because sequence information was being used in another research project. An additional 221 PRRS PCR-positive cases were randomly selected using simple random sampling without replacement (SAS; proc survey) based on a sampling frame constructed from all remaining PRRS PCR-positive submissions during the aforementioned study period, for a total of 400.
The case reports for each of the 400 selected cases were individually examined. Those cases in which only “weak positive” PCR results were produced were excluded (n = 31) as the probability of there being enough PCR product to enable genetic sequencing for the other project was low. Positive cases that had been submitted by veterinarians who were not members of the Ontario Association of Swine Veterinarians (OASV) were also excluded (n = 6) as the project had been advertised through OASV meetings and publications and non-members did not know about the research project. In several instances, multiple PCR-positive cases had been submitted from the same site from September 1, 2004 to January 14, 2006 and more than one case arising from the same site was included in the list of 400. If < 30 d had passed between positive submissions from the same site, the 2nd submission was excluded (n = 34). This left a total of 329 retrospective PRRS PCR-positive cases out of the original 400 selected for follow-up.
Case reports from the selected retrospective submissions were examined for contact information for the owners of the sites from which the samples were submitted. In some cases, the veterinarians submitting samples that tested positive on PCR contacted the researchers to provide contact information for producers. For the remaining case reports for which adequate producer contact information was not available, the submitting veterinarians were contacted to gain the producer contact information. If this failed, the case was excluded.
The researchers were successful in following up with only 93/329 (28.9%) eligible retrospective cases that were originally selected. The reasons for failing to gather information on eligible retrospective cases included: the submitting veterinarian failed to respond to requests for client contact information (n = 76); the study ended before successful contact was made with the producer (n = 47); the submitting veterinarian declined to have the case participate in the project (n = 46); the producer declined participation (n = 33); the producer failed to respond to attempts to contact them (n =26); the veterinarian/producer didn’t know which site the sample(s) came from (n = 7); and the client phone number given was not in service (n = 1). Because of the low number of retrospective cases enrolled in the study, information from 28 additional retrospective cases was gathered, for a total of 121 retrospective cases. This additional information was collected from producers that were being interviewed for information on cases that had been submitted from their site(s) during the prospective timeframe of the study. These producers also happened to have retrospective PCR-positive cases that met the study selection criteria (not a “weak positive,” submitted by an OASV member veterinarian, > 30 d since a previous positive submission) but whose cases had not been randomly selected from the 860 PRRS PCR-positive cases submitted between September 1, 2004 and January 14, 2006.
For the prospective portion of the study, beginning on January 15, 2006, all new serum and tissue submissions to the AHL from Ontario with positive PRRS PCR test results were identified. As before, cases with weak positive results, those submitted by non-OASV members, and those cases in which < 30 d had passed since a previous positive submission from the same site were excluded. In late December 2006, serum and tissue gel-based PRRS RT-PCR testing at the AHL was replaced by combined multiplex real-time single-tube PRRS RT-PCR testing that included primers and probes for both European and North American strains of PRRSV. The primers and probes in this assay are proprietary to Tetracore (Rockville, Maryland, USA) and this test has been previously described (10). The results of the new test were reported as “positive,” “negative,” or “suspicious.” Cases with “suspicious” or “negative” results were excluded. The search for new PCR-positive PRRS cases ended on August 31, 2007.
In total, 1007 PRRS PCR-positive cases were submitted between January 15, 2006 and August 31, 2007. Of those, 898 met the study’s selection criteria. Due to time constraints and the on-going nature of the prospective portion of the study, detailed reasons for not following up on eligible prospective cases were not recorded. Successful contact was made with producers representing 348 separate cases. Fourteen producers declined to have their case participate in the study, so information on 334 prospective cases was gathered.
The producers who agreed to voluntarily participate were asked to complete a survey. For producers with multi-site operations, each site was treated individually and a separate survey was completed for each site from which PRRS positive submissions were received. For sites from which multiple positive samples had been submitted over time, a separate survey was completed for each submission. The survey collected information on herd demographics, the presence or absence of specific clinical signs, and herd management practices such as disease control strategies. The same survey was administered to both retrospective and prospective cases and producers were asked to provide information about what was occurring at the time that the PCR-positive samples were taken. Most surveys were completed by telephone interview with the first author or 1 of 2 trained research technicians [91.6% (417/455)]. Some producers [7.1% (32/455)] preferred to complete the survey themselves, in which case the survey was mailed or faxed to them. In other instances [1.3% (6/455)], herd veterinarians wished to administer the survey to their clients. All producer-and veterinarian-completed surveys were examined and, if there were any questionable responses, the researchers contacted the producers by telephone for clarification. All information was kept confidential and used for summary purposes only.
A hierarchical database (Microsoft Access; Microsoft Corporation, Redmond, Washington, USA) reflecting the questions on the survey was used to facilitate data entry. Data were checked for consistency and, if necessary, validated by telephone follow-up. The data were exported to SAS 9.1 (SAS Institute, Cary, North Carolina, USA) for statistical analysis. Descriptive statistics including means, standard deviations, minimum and maximum values for quantitative variables, and frequency counts and percentages for qualitative variables were calculated.
Associations between reported clinical signs suggestive of PRRS and selected demographic and management factors were tested using logistic regression models. For sites housing sows, the clinical signs included were abortion, stillbirths, weak-born piglets, high pre-weaning mortality, sow anorexia, and sow mortality. For sites housing nursery and/or finishing pigs, the clinical signs included were mortality and respiratory disease. Demographic factors examined included herd type, average sow inventory, and maximum nursery/finisher capacity (number of hog spaces in barn). Quadratic effects of quantitative variables were tested. In addition, lowess smoothing was done to further inspect linearity of quantitative variables with logit for each clinical sign. Management factors included in the analysis were disease control strategies (commercial vaccine use, serum inoculation, biofeedback) and pig flow [all-in/all-out (AIAO) or continuous] used concurrently to samples testing PCR positive. For the purposes of this study, serum inoculation referred to any deliberate injection of serum from PRRSV-infected pigs into other pigs to ensure their exposure to a farm-specific PRRS strain while biofeedback referred to the feeding of pig manure, dead/stillborn/mummified piglets or placenta to individual or groups of pigs. To determine pig flow for each site, producers were asked if they always, usually, occasionally, or never managed their sites, barns, rooms, and/or pens in an AIAO manner. For analysis purposes, farrowing rooms were considered AIAO if they were always managed in an AIAO manner, otherwise they were considered to be continuous flow. Nursery and finisher stages were classified as AIAO by site, barn, or room if they always or usually used AIAO by site, barn, or room, otherwise they were considered to be continuous flow.
For each clinical sign, univariable associations were tested. The following groups of explanatory variables were considered: herd size, herd type, pig flow, and disease control strategies. Only explanatory variables appropriate for each production class were used in the modelling process. Quadratic effects of quantitative variables were tested. Logistic regression models were estimated by generalized estimating equations with exchangeable correlation structure for herds repeatedly sampled within site over time. The Wald test was used for assessing statistical significance. Empirical standard errors were used.
The 455 completed surveys represented 338 separate sites. Most sites (75.7%) were represented only once. Of the remaining sites, 56 were represented twice, 18 were represented 3 times, 7 were represented 4 times, and 1 was represented 5 times. For sites that submitted multiple positive samples over time, the mean length of time between submissions was 230 d [standard deviation (s) = 157 d], with a minimum of 30 d and a maximum of 677 d. Most sites (33.4%) included in the study were farrow-to-finish sites while the fewest sites (3.3%) housed gilts only (Table I).
The average sow inventory was 577 sows with farrow-to-finish sites having the fewest and farrow-to-wean sites having the greatest number of sows (Table I). The average nursery pig capacity was 1563 pigs. Farrow-to-finish sites had the smallest nursery capacity while nursery-finisher sites had the largest. The average finishing pig capacity was 1443 pigs. Again, farrow-to-finish sites had the smallest finishing capacity while finishing-only sites had the largest (Table I).
The type of pig flow used by most sites housing nursery pigs was AIAO-by-room (69.8%). The remaining nursery sites used continuous flow (22.3%) or were AIAO-by-barn (7.9%). Over half of the sites housing finishing pigs (52.1%) used continuous flow in the finishing barn. The remaining sites housing finishing pigs were using AIAO-by-room (34.2%) or AIAO-by-barn (13.7%) flows. An almost equal proportion of sites housing sows reported using AIAO-by-room (50.3%) and continuous flow (49.3%) in farrowing rooms.
Vaccination was commonly reported with > 1/3 of producers using a PRRS vaccine in some manner (Table II). The most frequent uses of PRRS vaccine were in gilts at arrival (35.7%) and in gilts before breeding (30.3%). Some producers (4.5%) described using PRRS vaccine in some other manner, such as blitz vaccinating the entire herd. Porcine reproductive and respiratory syndrome vaccine was seldom used in growing animals with only 1.7% of sites housing nursing piglets and 2.5% of site housing nursery pigs reporting use of PRRS vaccine in these animals. Finisher pigs were never vaccinated against PRRS (Table II). All 3 types of modified-live PRRS vaccines available in Canada [Inglevac PRRS MLV (MLV), Inglevac PRRS ATP (ATP), and Reprocyc PRRS-PLE (PLE) (Boehringer Ingelheim Vetmedica, St. Joseph, Missouri, USA)] were being used. Across all sites, MLV was used most commonly (21.5%), followed by PLE (13.6%), and ATP (5.1%) (Table II).
Serum inoculation was not commonly used as a PRRS control method on participating sites. The most common use of serum inoculation was in gilts, but only 10.8% of sites that housed gilts used this technique (Table II). Biofeedback was more commonly used; with almost 1/4 of producers from sites housing gilts reporting using this practice (Table III). Serum inoculation and biofeedback were each used in nursery pigs on only one site and neither technique was used in nursing or finishing pigs (Table II).
When all PRRS control strategies were examined together in sites housing sows, the most commonly used combination of PRRS control strategies was providing biofeedback to both gilts and sows (5.8%). Other common combinations of PRRS control strategies included using PRRS vaccine in gilts at arrival and in sows at weaning (5.1%), using PRRS vaccine in gilts before breeding and in sows at weaning (4.1%), and using PRRS vaccine in gilts before breeding and in pregnant sows (4.1%). Slightly more than 1/4 (27.1%) of producers with sites housing sows reported using none of the PRRS control strategies considered.
Complete information on clinical signs was missing from surveys from 10 sites. Of the remaining 445 sites, 84.9% of producers reported seeing clinical signs associated with PRRS virus in their pigs (Table III). All of the clinical signs of interest (sow anorexia, abortions, stillborn piglets, weak-born piglets, pre-weaning mortality, sow mortality) were reportedly seen in at least 1 of the participating sites housing sows (n = 292) at the time that PRRS PCR-positive samples were taken (Table II). Producers from 15.8% of participating sites reported seeing all of the previously mentioned clinical signs in their sows while 5.8% reported seeing all clinical signs except sow mortality at the time the positive samples were taken. Almost 1/4 (24.7%) of sites reported seeing no clinical signs of PRRS in sows at the time the positive samples were taken. Producers from sites housing nursery pigs (n = 286) reported seeing respiratory disease (64.9%) and nursery pig mortality (69.2%) at the time that the samples that tested PRRS positive on PCR were taken. Slightly less than half of the producers with sites housing finishing pigs (n = 236) reported seeing respiratory disease (46.8%) and mortality (46.2%) at the time the PCR-positive samples were taken (Table III).
When examining associations between reported clinical signs in sows and herd type, more farrow-to-wean sites reported seeing abortions than both farrow-to-grower and farrow-to-finish sites (Table III). There were no differences in the proportions of sites reporting any other clinical signs among the 3 herd types where sows were housed. The proportion of sites reporting nursery pig respiratory disease or mortality did not differ among farrow-to-finish, farrow-to-grower, nursery-only, and nursery-finish sites (Table III). Producers from more finishing-only sites reported seeing finishing pig respiratory disease and mortality than producers from farrow-to-finish sites (Table III). When all clinical signs were considered together, a greater proportion of farrow-to-grower sites reported seeing clinical signs of PRRS than finishing-only sites. Similarly, more farrow-to-finish sites reported seeing clinical signs of PRRS than did farrow-to-wean sites (Table III).
When herd size was considered, there was an association between sow inventory and probability of observing clinical signs, with an increasing probability of observing clinical signs with increasing sow numbers, although not all associations were significant. As sow inventory increased by 100 animals the odds of reporting abortion [odds ratio (OR) = 1.07, P = 0.002, 95% confidence interval (CI):1.03–1.12], weakborn piglets (OR = 1.06, P = 0.01, 95% CI:1.02–1.11), sows off-feed (OR = 1.05, P = 0.04, 95% CI:1.00–1.10) and sow mortality (OR = 1.07, P = 0.003, 95% CI:1.02–1.11) increased. Using lowess smoothing, u-like association was visually apparent for 3 clinical signs (preweaning mortality, abortion, sow mortality) around a herd size of ~100 sows. Herds with < ~100 sows appeared to have higher logits of different magnitude than herds with ~100 sows, after which size, the smoothed logit started increasing linearly. However, quadratic effect during modelling was not identified as statistically significant for either of the above 3 clinical signs in sow herds. Clinical signs were not associated with nursery capacity but the odds of reporting finisher pig mortality was associated with increasing finisher pig capacity. For each 100 head increase in finishing capacity, the odds of reporting finisher mortality increased by 1.03 times (P = 0.03, 95% CI: 1.00–1.07).
Any use of a PRRS vaccine in sows reduced the odds of reporting weak-born pigs (OR = 0.56, P =0.027, 95% CI: 0.34–0.94) and preweaning mortality (OR = 0.46, P = 0.002, 95% CI: 0.28–0.74). Using a PRRS vaccine in sows at weaning reduced the odds of reporting weak-born pigs (OR = 0.45, P = 0.009, 95% CI: 0.24–0.82) and preweaning mortality (OR = 0.45, P = 0.005, 95% CI: 0.26–0.79). Using a PRRS vaccine in gilts at arrival also decreased the odds of reporting pre-weaning mortality (OR = 0.54, P = 0.029, 95% CI: 0.31–0.94) and mortality in nursery pigs (OR = 0.45, P = 0.026, 95% CI: 0.22–0.91). Conversely, the use of serum inoculation in gilts and in sows increased the odds of reporting sow mortality (OR = 2.86, P = 0.008, 95% CI: 1.31–6.24 and OR = 2.34, P = 0.05, 95% CI: 0.99–5.50, respectively). Using serum inoculation in sows also increased the odds of reporting sows off-feed (OR = 2.59, P = 0.037, 95% CI: 1.06–6.31). Providing biofeedback to gilts increased the odds of reporting both finisher pig respiratory disease (OR = 2.41, P = 0.04, 95% CI: 1.03–5.61) and finisher pig mortality (OR = 2.68, P = 0.018, 95% CI: 1.18–6.08).
Significant associations between pig flow and the presence of certain clinical signs were found. The odds of producers reporting weakborn piglets and sow mortality on sites where strict AIAO flow was used in the farrowing room were 2.02 (P = 0.006, 95% CI: 1.22–3.34) and 2.30 (P = 0.004, 95% CI: 1.30–4.08) times greater respectively, than on sites that were not using strict AIAO flow in the farrowing rooms. On sites where pigs were finished, the odds of producers reporting finisher respiratory disease and mortality were 3.33 (P = 0.01, 95% CI: 1.28–8.67) and 2.80 (P = 0.03, 95% CI: 1.09–7.23) times greater if they used AIAO-by-barn flow than if they used AIAO-by-room flow. Similarly, the odds of producers reporting finisher respiratory disease and mortality were 2.78 (P = 0.02, 95% CI: 1.16–6.66) and 3.31 (P = 0.007, 95% CI: 1.38–7.95) times greater if they used AIAO-by-barn flow than if they used continuous flow.
There was a wide range of clinical signs reported from PRRS PCR-positive sites and the proportion of sites reporting specific clinical signs differed from those reported by others. In a study of 52 PRRS positive herds in Illinois and Iowa, abortion, sow mortality, and preweaning mortality were seen in a smaller proportion of sow herds (38%, 20%, and 45%, respectively) and respiratory disease was seen in a greater proportion of nursery and finisher herds (74% and 72%, respectively) than in our study (11). These results corroborate earlier observations that the clinical syndrome of PRRS is highly variable (12). Some of this is likely due to differences in virulence between strains of PRRSv. It has been shown that the ability of PRRSv to induce clinical respiratory disease (13,14) and affect reproductive performance (15) is strain-dependent. Unfortunately, specific information on clinical signs of PRRS observed in Ontario herds prior to the 2004–2005 epidemic is unavailable. It is therefore difficult to ascertain whether the clinical signs observed in herds during and after the 2004–2005 outbreak differed from the clinical signs of PRRS that were seen in herds prior to the outbreak. Many other factors, such as management practices, the presence of other diseases, or whether the PRRS PCR-positive submission represented an outbreak in a previously naive site, a new outbreak in a previously positive site, or an endemically affected site, could influence the appearance of clinical signs on a PRRS positive site.
The proportion of sites in this study that used PRRS vaccine in their animals was lower than that reported in other studies of PRRS positive herds. While just over 1/3 of sites in the present study reported using a PRRS vaccine, 53% of 52 Illinois/Iowa herds (11) and 69% of 226 Quebec herds (16) reported use of a PRRS vaccine. Most striking is the difference in the proportion of herds using a PRRS vaccine in growing animals between the study herein and the Quebec study, a neighbouring province to Ontario. In the Quebec study, 42% of nursery-grow-finish units were using PRRS vaccine (16), while only 2.5% of sites in Ontario were using PRRS vaccine in growing animals, and only in nursery pigs. These differences in vaccine usage may reflect differences in herd type, management practices or PRRS prevalence between the different study areas. Also, reported outbreaks of clinical PRRS in PRRS-vaccinated pigs have led to doubts about the efficacy of currently available vaccines (17). Because of this, the use of commercial PRRS vaccines may have declined in favour of other PRRS control strategies.
Despite concerns over the efficacy of PRRS vaccines, results from our study indicated that the use of a PRRS vaccine in sows and gilts has positive effects in both the farrowing room and in the nursery. Others have also reported beneficial effects of PRRS vaccine use in PRRS infected pigs. One study showed that using a modified-live PRRS vaccine reduced clinical signs and improved average daily gain in both PRRS-naïve and previously PRRS-infected pigs experimentally infected with a heterologous PRRS strain (18). In other research, clinical signs were reduced in PRRS-vaccinated, heterologously challenged pigs compared with unvaccinated, challenged pigs (19). While some have found no reduction in clinical signs with the use of PRRS vaccine in PRRS positive herds (11), producers should not rule out using a PRRS vaccine as an aid to control clinical PRRS.
One newer PRRS control strategy is serum inoculation, which involves deliberate exposure of animals to a herd-specific strain of PRRSv. Some research into the effect of serum inoculation on herd performance and the expression of clinical signs of PRRS has been conducted. Some have reported the successful eradication of PRRSv from herds through the use of serum inoculation (20). Others found that serum inoculation reduced the severity of clinical signs of PRRS in both PRRS-naïve and previously PRRS-exposed pigs (21). Our study showed that the use of serum inoculation increased the risk of reporting sow anorexia and sow mortality. Immediately following serum inoculation, clinical signs of PRRS can be expected and have been reported by others. Batista et al (22) observed fever, anorexia, and depression in gilts for 48 to 72 h post-serum inoculation. If the PCR-positive samples from sites using serum inoculation in the present study were taken within several days of serum inoculation occurring, it may explain the association between the use of this PRRS control method and sow anorexia and mortality. Unfortunately, we do not know exactly when in relation to a PCR-positive sample being taken that serum inoculation occurred on participating sites, only that it happened sometime prior to the sample being taken. Another consideration is that many producers turn to serum inoculation after unsuccessfully trying other PRRS control strategies. It could be that the herds using serum inoculation were dealing with particularly virulent, difficult to control, strains of PRRS. These strains may have produced more severe clinical signs, hence the apparent association between serum inoculation and sow anorexia and mortality. In any case, producers considering this technique should be made aware of the association between serum inoculation and clinical signs of PRRS in sows.
A relatively commonly used PRRS control strategy, biofeedback to replacement gilts, was associated with finisher pig respiratory disease and mortality. Using biofeedback for PRRS virus exposure during the first week of gilt acclimatization has been recommended for effective PRRS control (23). However, to our knowledge, no peer-reviewed reports on the effects of using this PRRS control strategy on the expression of clinical signs of PRRS are available.
In all production phases, we expected that the use of AIAO would reduce the prevalence of clinical signs as this management practice minimizes disease transfer between pigs (24). Others have found that using AIAO flow has a protective effect for several clinical signs of PRRS in different stages of production (11). Therefore, the finding that AIAO flow in farrowing rooms and finishing barns was a risk factor for reporting weakborn pigs, sow mortality, and finishing pig respiratory disease and mortality was surprising and difficult to explain. It may be that that the use of continuous flow leads to a low PRRS prevalence with on-going morbidity and mortality problems while, with AIAO flow, a population may quickly go from being PRRS negative to PRRS positive with a rapid increase in pig morbidity and mortality. Producers using continuous flow in their barns may expect to see a low level of disease in their pigs and therefore may be less likely to report seeing clinical signs than producers using AIAO flow who may observe a sudden, dramatic increase in clinical signs in their pigs. Also, it is possible that recall bias may have affected this and other results of this study. Producers were asked to provide information about what was occurring in their herd at the time that PRRS PCR-positive samples had been taken. In some cases, particularly those from the retrospective portion of the study, producers had to recollect what was being done in their herd almost 2 y prior to the time the researchers contacted them. Unmeasured sources of variability among participating sites could also be contributing to our findings.
Larger herd size was associated with sow mortality, sow anorexia, abortion, weakborn pigs, and finishing pig mortality. Goldberg et al (11) also found associations between large herd size and sow mortality and nursery pig respiratory disease and speculated that the risk of an epidemic is greater in large herds due to the increased opportunity for virus transmission within these herds.
Caution is needed when interpreting the results of this study as the analysis was performed on PRRS virus positive sites only. Also, given that a site was PRRS PCR-positive, our question was whether or not selected demographic and management factors were associated with a particular clinical sign. Many other factors, not taken into account in this study, could influence the appearance of clinical signs in a PRRS PCR-positive herd. In this analysis, 4 groups of factors were explored, some with multiple levels. In total, each clinical sign was compared with between 20 (farrowing phase) and 23 (nursery phase) discrete levels of factors, which did not consider potential multiple comparisons within each factor. We did not adjust for multiple comparisons, partly in accordance with the exploratory nature of the analysis. Also, due to the difficulty the researchers faced in contacting eligible cases and the voluntary nature of producer participation in the study, selection bias may have influenced the results on multiple levels.
In conclusion, a wide range of clinical signs was reported from PRRS PCR-positive sites during and after an apparent PRRS outbreak in Ontario. Herd demographic factors, including herd size and herd type, were found to be associated with reporting clinical signs of PRRS. Producers using a PRRS vaccine in sows and gilts were less likely to report clinical signs so using commercially available vaccines should not be overlooked as a potential means to control clinical signs of PRRS. While using AIAO pig flow, biofeedback and serum inoculation were found to increase the risk of reporting clinical signs, use of these techniques should not be disregarded and PRRS control strategies should be evaluated on an individual herd basis.
This project was funded by Ontario Pork. The authors thank technicians Karen Richardson and Doug Wey for their assistance in collecting data, Mioara Antochi, Jane Coventry, Li Ge, Chris Mason, and Rebecca Marshall for PRRS RT-PCR testing at the AHL, and OASV members and participating producers for their cooperation in our research.