African swine fever: a New Zealand perspective on epidemiological risk factors for its occurrence.

This article reviews key epidemiological and clinical features of African swine fever (ASF). We identify particular aspects of New Zealand's pig populations (commercial, non-commercial, and wild) that may affect the risk of disease entry or spread. Review of published literature is supplemented by analysis of demographic and spatial aspects of the New Zealand commercial, non-commercial, and feral pig populations to provide context around risk factors for the disease that are most relevant to New Zealand. The current Eurasian outbreak of ASF, including recent spread into Oceania, has increased the risk of an incursion of the disease into New Zealand. Large volumes of fresh pork importation (including from countries affected by ASF), large non-commercial pig populations with substantial spatial overlap with the country's commercial industry, limited monitoring of compliance with waste food feeding regulations, and lack of mandatory premises identification for non-commercial pig holdings would likely contribute to the risk of spread of ASF in the event of an incursion. Awareness amongst veterinarians of these risk factors will contribute to national biosecurity and disease preparedness efforts in New Zealand.

Is transportation a risk factor for African swine fever transmission in Australia: a review.

African swine fever (ASF) is a viral disease of the pigs that was first described in Africa during the early part of the twentieth century. The disease has periodically occurred outside of Africa, including an ongoing epidemic in Europe and Asia that started in 2007; the disease has never occurred in Australia or New Zealand. Once introduced into a country, spread can occur through direct and indirect routes of transmission. Infected feral pig populations have the potential to act as a long-term reservoir for the virus, making eradication difficult. Just before and throughout the period of clinical signs, ASF virus is shed in oronasal fluids, urine, faeces and blood. This results in contamination of the pig's environment, including flooring, equipment and vehicles. Transportation-related risk factors therefore are likely to play an important role in ASF spread, though evidence thus far has been largely anecdotal. In addition to the existing AUSVETPLAN ASF plan, efforts should be made to improve transportation biosecurity, from the time a pig leaves the farm to its destination. Collection of data that could quantify the capabilities and capacity of Australia to clean and disinfect livestock trucks would help to determine if private and/or public sector investment should be made in this area of biosecurity. No peer-reviewed research was identified that described a specific process for cleaning and disinfecting a livestock truck known to be contaminated with ASF virus, though literature suggests that transportation is an important route of transmission for moving the virus between farms and countries.

The shedding of group A rotavirus antigen in a newly established closed specific pathogen-free swine herd.

A longitudinal study was undertaken in a newly established specific pathogen-free (SPF) swine herd to determine the dynamics of rotavirus antigen shedding in a closed swine facility. Pregnant SPF gilts which populated the herd, and their offspring, were monitored weekly for three consecutive lactations. Fecal samples were assayed for the presence of group-specific viral antigen by a solid phase immunoassay (ELISA). Results indicate that in the week prior to farrow, 35% of samples from gilts/sows contained rotavirus antigen. During nursing, 37% of the gilts'/sows' fecal samples also contained virus antigen. Over the course of three farrowings, every gilt/sow in the herd excreted virus antigen. Virus antigen was present in 25% of the samples tested from nursing pigs and in 70% of the samples tested from pigs in the postnursing period; 95% of the litters excreted virus antigen either while nursing or postweaning. Seasonal incidence in virus antigen excretion was noted with proportionally more suckling pigs virus antigen-positive in summer and proportionally more sows/gilts positive during winter. Diarrhea occurred only rarely in the sampled population. Although piglets shed rotavirus subclinically, ELISA positive feces from piglets of each lactation caused severe disease when fed to neonatal gnotobiotic pigs. Electropherotyping of these passaged viruses indicated minor variation in RNA banding patterns over time.

Evaluation of the sensitivity and specificity of two diagnostic tests for antibodies to pseudorabies virus glycoprotein X.

The diagnostic performance of 2 enzyme-linked immunosorbent assays (gX-T, gX-H) for antibodies to pseudorabies virus (PRV) glycoprotein X (gX) were evaluated using 311 serum samples from a nonvaccinated quarantined herd. When the standardized virus neutralization (VN) test, which uses the Shope strain (VN Shope), was used as the comparative diagnostic standard, the gX-T test had a 7% false-negative rate and a 52% false-positive rate, and the gX-H test had a 19% false-negative rate and a 19% false-positive rate. When the VN test with a Bartha recombinant strain (VN Bartha gIIIKa) was used as the diagnostic standard, the gX-T test had a 9% false-negative rate and a 26% false-positive rate, and the gX-H test had a 24% false-negative rate and a 11% false-positive rate. Thus, the gX-T test was more sensitive and the gX-H test was more specific. Additional diagnostic tests on 79 serum samples from a noninfected herd did not produce false positives for the gX-H test, but there was an 8% false-positive rate for the gX-T test. Previous studies from our laboratory have demonstrated that VN Bartha gIIIKa has higher sensitivity than VN Shope, without losing specificity, and thus is a better comparative diagnostic standard. When adding a suspect range to the gX-T test, using the same criteria as the suspect range for the gX-H test, the false-positive rate of the gX-T test was reduced to 5% when evaluated versus VN Bartha gIIIKa in the infected herd and to 1% for the PRV-negative herd.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of salinomycin on sow weight change during lactation and on sow reproductive performance.

Forty-five gravid cross-bred sows (mean parity 3.3 +/- .3) were randomly allotted to two dietary treatments: corn-soybean mean (CS) or CS plus 60 mg salinomycin per kilogram of diet (CSS). Sows were fed their respective diets through two successive parities with dietary treatment initiated at 100 d postcoitum and continued until weaning of the second successive litter. Therefore, sows fed CSS received salinomycin for 14 d before the first parturition and for approximately 153 d before the second parturition. Daily feed intake was restricted to 2 kg.hd-1.d-1 during gestation and to 3 kg.hd-1.d-1 from weaning to breeding. All sows. had ad libitum access to feed during lactation. Sows were weighed 7 d prior to parturition, at weaning and at breeding. Weaning-to-estrus interval and farrowing interval were recorded for all sows. Litters were weighed at birth and weaning. There were no differences (P greater than .05) between dietary treatments in sow weights before parturition, at weaning or at breeding for either first or second farrowing. The CSS-fed sows lost more weight from weaning to breeding after the first (P less than .03) and second (P less than .05) lactation periods than CS-fed sows. The CSS-fed sows tended to gain more (P = .06) weight during lactation than CS-fed sows. There were no differences (P greater than .05) between treatments in lactation feed intake, weaning-to-estrus interval, farrowing interval, litter size born or weaned, litter weights at birth or at weaning, or in sow culling rate.(ABSTRACT TRUNCATED AT 250 WORDS)

Purification of swine haptoglobin by affinity chromatography.

A globin-agarose affinity chromatography technique was used to purify swine haptoglobin. This technique provides a highly specific, single-step purification method without the contamination of extraneous serum proteins reported by previous studies. Complex formation between the haptoglobin isolate and swine hemoglobin confirmed that biological activity was maintained during the purification process. Immunoelectrophoretic and Ouchterlony immunodiffusion methods revealed that the swine haptoglobin isolate cross-reacted with polyvalent antisera against human haptoglobin.

Serum haptoglobin concentration as an indicator of weight gain in pigs.

Serum haptoglobin concentration was investigated as an indicator of weight gain in commercially-reared pigs. The serum haptoglobin concentrations and weights of 40 pigs were monitored on a weekly basis, from weaning to 13 weeks of age. All data were ranked based on the week 13 weights, and divided into high, middle and low weight gain groups. By the fourth week of the study, when the pigs were seven weeks old, serum haptoglobin concentrations could be used to differentiate pigs that would have a high weight gain at the end of the study from those which would have a low weight gain.

A model for the induction of Pasteurella multocida type-A pneumonia in pigs.

A model for the induction of pneumonia caused by Pasteurella multocida type-A was developed. Anesthetized pigs were dosed intratracheally with 10(10) colony forming units of P. multocida type-A suspended in a total dose of saline based on the pig's body weight (8 mL/kg). A severe bronchopneumonia was present when the treated pigs were euthanized seven days postinfection. The mean percentage of pneumonic lesions in the treated pigs, as determined by morphometric measurement, was 14.03 +/- 7.01 (X +/- SD). In contrast, in the control pigs, the mean percentage of pneumonic lesions was 0.59 +/- 0.52. For the treated pigs during the period following induction of pneumonia, weight gain and feed intake were reduced significantly (p less than 0.05) compared to the controls. It was shown that severe pneumonia could be induced without the use of other infectious agents, which could potentially confound the experimental model.

Efficacy of an in-feed preparation of ivermectin against endoparasites and scabies mites in swine.

In 2 trials, the efficacy of an in-feed preparation of ivermectin was evaluated in 40 pigs naturally infected with endoparasites and Sarcoptes scabiei var suis. Treated pigs (n = 10 in each trial) were fed a ration containing 2 ppm ivermectin for 7 days, followed by consumption of a nonmedicated ration for the remainder of the trial. Control pigs (n = 10 in each trial) were fed a complete, nonmedicated ration for the duration of the trial. Pigs in trial A were monitored for 14 days after treatment; those in trial B were monitored for 35 days after treatment. In trial A, treatment efficacy of ivermectin was 100% against Ascaris suum, Physocephalus sexalatus, Oesophagostomum dentatum, O brevicaudum, Metastrongylus spp; 99.8% against Ascarops strongylina; 90.9% against Trichuris suis; and 13.1% against Macracanthorhynchus hirudinaceus. At the terminus of the trial, statistically significant (P less than 0.05) differences were observed between numbers of treated and control pigs infected with A suum, Ascarops strongylina, and Oesophagostomum spp. On posttreatment day 14, S scabiei were not found in any scrapings taken from treated pigs, but were found in scrapings from 3 of 10 control pigs. The number of infested pigs in the treatment group was not statistically different from the number of infested pigs in the control group. In trial B, treatment efficacy was 100% for A suum and Metastrongylus spp; 96.9% for Ascarops strongylina; and 76.9% for M hirudinaceus. At the terminus of the trial, statistically significant (P less than 0.05) differences were evident between numbers of treated and control pigs infected with A suum, Ascarops strongylina, and Metastrongylus spp.(ABSTRACT TRUNCATED AT 250 WORDS)

Bioavailability, pharmacokinetics, and plasma concentration of tetracycline hydrochloride fed to swine.

A 2 X 2 crossover design trial was conducted in gilts to determine the bioavailability and pharmacokinetics of tetracycline hydrochloride. The bioavailability of tetracycline hydrochloride administered orally to fasted gilts was approximately 23%. After intravascular administration, the disposition kinetics of tetracycline in plasma were best described by a triexponential equation. The drug had a rapid distribution phase followed by a relatively slow elimination phase, with half-life of 16 hours. Its large volume of distribution (4.5 +/- 1.06 L/kg) suggested that tetracycline is distributed widely in swine tissues. Total body clearance was 0.185 +/- 0.24 L/kg/h. Other pharmacokinetic variables were estimated. In a second trial, 3 gilts were fed a ration containing 0.55 g of tetracycline hydrochloride/kg of feed. Resulting plasma concentration of tetracycline was determined at selected times during 96 hours after exposure to the medicated feed. Plasma drug concentration peaked (0.6 micrograms/ml) at 72 hours after access to the medicated feed.

Risk factors associated with transmissible gastroenteritis in swine.

Commercial production data base records from 2 Illinois farms, on which epizootic or enzootic transmissible gastroenteritis (TGE) was experienced, were accessed for an epidemiologic study. Risk factors investigated were sow parity, source of sows, location of farrowing crates, and breeding practices. At farm 1, an epizootic was experienced; at farm 2, an epizootic of TGE followed by enzootic TGE was experienced. Initially, crude risk ratios were calculated for these risk factors, and the crude risk ratios were subsequently adjusted for confounders and interactions, using multiple logistic regression techniques. After adjustment, parity-3 sows were 2.3 times more likely to have litters with TGE than were sows of all other parities on farm 1, and parity-1 sows were 2.6 times more likely to have litters that experienced TGE than were sows of all other parities on farm 2. A single boar on each farm was linked to increased likelihood of a sow's litter contracting epizootic TGE on each farm. Enzootic TGE was maintained by the periodic influx of outside-source gilts on farm 2; these gilts were 2.2 times more likely to have litters with TGE than were sows derived from farm 2. Sows housed in farrowing crates located under the cold air inlet of farm 2 were 1.7 times as likely as sows located in other rows to have litters with enzootic TGE.

Plasma concentrations of oxytetracycline in swine after administration of the drug intramuscularly and orally in feed.

Four pigs were used in a 2 X 2 crossover study to determine plasma oxytetracycline (OTC) concentration and OTC pharmacokinetic variables after IM administration of 2 OTC preparations--long acting OTC and a 100-mg of OTC/ml solution (OTC-LA and OTC-100, respectively)--at a dosage of 20 mg/kg of body weight. In a second study, 3 additional pigs were given ad libitum access to feed containing pure OTC (0.55 g/kg of feed). The mean (+/- SD) peak plasma OTC concentration after OTC-LA administration was 6.0 +/- 2.2 micrograms/ml at 30 minutes; the mean peak plasma OTC concentration after OTC-100 administration was 6.7 +/- 3.4 micrograms/ml at 90 minutes. Mean plasma OTC concentration after oral OTC administration in feed peaked at 0.4 micrograms/ml 48 hours after access to OTC-medicated feed and decreased to 0.25 micrograms/ml by the end of that study. Mean plasma OTC concentration was maintained at greater than 0.5 micrograms/ml for less than 48 hours after OTC-LA administration and for less than 36 hours after OTC-100 administration. Mean plasma OTC concentration decreased to less than 0.2 micrograms/ml by 72 hours after IM administration of either product. Calculation of area under the plasma OTC concentration-time curve (AUC) did not reveal significant difference between the 2 OTC formulations. There also was not significant difference (between OTC-LA and OTC-100) in the value of the disappearance rate constant after administration of either OTC formulation. The data did not indicate significant pharmacologic advantage of OTC-LA, compared with OTC-100, when either formulation was administered IM at a dosage of 20 mg/kg.(ABSTRACT TRUNCATED AT 250 WORDS)

Eradication of pseudorabies virus from three large swine herds achieved by management intervention and use of a vaccine with a deletion for glycoprotein I.

Three large farrow-to-finish swine herds in Illinois, quarantined because of infection with pseudorabies virus (PRV), were enrolled in an intensified PRV eradication program, with the goal being release from quarantine within 3 years. The intervention plan primarily relied on vaccination, using a vaccine with a deletion of the genes coding for glycoprotein I, in breeding and growing/finishing pigs and decreases of movement and mixing of growing/finishing pigs. The initial goal was to decrease viral spread in the growing/finishing pigs, thereby enabling production of seronegative replacement gilts. Off-site rearing of replacement gilts was implemented in 1 recently infected herd in which the seroprevalence in the growing/finishing group was high. Results of bimonthly serologic monitoring indicated that there was minimal spread of PRV in the growing/finishing pigs after 1 year. Increases in the number of sows culled combined with an increase in the number of seronegative replacement gilts entering the breeding group resulted in a reduction of sow seroprevalence, so that phased test and removal of seropositive breeding stock could commence in all 3 herds at about 18 months after initiation of the program. Persistent use of the test-and-removal procedure and repeated testing for release from quarantine were required for the most recently infected herd. All herds were released from quarantine within 3 years, indicating that a PRV eradication program based on vaccination and management changes designed to minimize the spread of PRV can be used in conjunction with test-and-removal procedures to effectively eliminate PRV from large farrow-to-finish swine herds.

Serum haptoglobin concentration in swine naturally or experimentally infected with Actinobacillus pleuropneumoniae.

Serum haptoglobin (Hp) concentrations were measured in swine that were naturally or experimentally infected with Actinobacillus pleuropneumoniae. In swine from a specific-pathogen-free herd, mean serum concentration of Hp (+/- SD) was 5.79 +/- 1.06 mg of cyanmethemoglobin-binding capacity (CHBC)/dl. Serum Hp concentrations in paired samples were measured at 7-day intervals in 40 swine randomly selected from a conventional herd that was experiencing an acute episode of pneumonia and deaths caused by A pleuropneumoniae serotype-5 infection. Day-0 and -7 serum Hp concentrations were 24.58 +/- 1.38 and 23.10 +/- 1.12 mg of CHBC/dl, respectively, with no significant difference between these measurements. In a second conventional herd with a history of chronic infection with A pleuropneumoniae serotype 5, serum concentrations of Hp measured in paired samples obtained 6 days apart were 12.36 +/- 0.81 and 18.63 +/- 0.76 mg of CHBC/dl, respectively, and were significantly (P < 0.05) different from each other. Twenty-nine 12-week-old conventional swine were challenged intranasally with A pleuropneumoniae serotype 1 (n = 19) and serotype 5 (n = 10). Serum Hp concentration increased from prechallenge concentrations of 7.49 +/- 1.38 and 15.10 +/- 1.22 mg of CHBC/dl, respectively, to 41.01 +/- 1.35 and 22.37 +/- 1.78 mg of CHBC/dl, respectively, 72 hours after challenge. For these 29 swine, serum Hp concentration was positively correlated with rectal temperature (r = 0.34; P < 0.001) during the immediate postchallenge period.

Prevalence of pseudorabies virus infection and associated infections in six large swine herds in Illinois.

Sera were collected from 6 large farrow-to-finish swine herds infected with pseudorabies virus (PRV) in Illinois. All herds were participating in the Large Herd Cleanup Study, a USDA-initiated project to evaluate the feasibility of eradicating pseudorabies from large farms (greater than 400 sows) by use of a combination of vaccination and management changes. Herd size ranged between 425 and 1,500 breeding females. Between April and July 1990, sera for measurement of PRV antibodies were obtained from 113 to 156 sows and 112 to 162 finishing pigs (body weight greater than 70 kg)/herd. Duplicate sera from 30 sows and 30 market-weight pigs/herd were obtained for measurement of serum antibodies to the following associated organisms: swine influenza virus, transmissible gastroenteritis virus, encephalomyocarditis virus, Actinobacillus pleuropneumoniae, Eperythrozoon suis, and 6 serovars of Leptospira interrogans. Prevalence of PRV antibodies attributable to field virus infection ranged between 53.8 and 100% for sows and between 0.7 and 97.3% for finishing pigs, as determined by the appropriate differential test for the vaccine being used on each farm. In only 1 herd, PRV seroprevalence was increased with higher sow parity. For associated infections, the risk of seropositivity attributable to PRV was not significant (for most infections) on all farms and varied among farms. Thus, pseudorabies did not appear, in general, to increase susceptibility to infection with other disease agents.

Progress after one year of a pseudorabies eradication program for large swine herds.

Six large farrow-to-finish swine herds quarantined for pseudorabies in Illinois participated in the USDA-initiated Large Herd Cleanup Study. These herds were monitored for antibodies to pseudorabies virus (PRV) for 1 year after the initiation of an intensive eradication program. Herd size ranged between 425 and 1,500 females of breeding age. Gene-deleted modified-live virus vaccines were used on all farms, with 3 of the 6 herds receiving a vaccine with a deletion of the gene for glycoprotein-I and the other 3 herds receiving a vaccine with a deletion of the gene for glycoprotein-X. The breeding herd and growing pigs were vaccinated on each farm. Each herd produced its own replacement gilts. In addition, management changes emphasizing all-in, all-out pig flow were initiated. One year after initiation of the vaccination program, sera for the measurement of PRV antibodies were obtained from sows and heavy finishing pigs (> 70 kg) from each of the farms. Prevalence of PRV antibodies attributable to wild-type virus infection ranged from 7 to 63% (median, 33%) for sows and from 0 to 42% (median, 4%) for finishers, as determined by the appropriate vaccine differential test. For each sow herd, there was a large decrease in the PRV seroprevalence rate after 1 year of the program (range, -21 to -68%; median, -42%). Examination of PRV prevalence rates by parity indicated decreased seroprevalences in the lower parities (< 2) in 3 of the herds, suggesting that vaccination reduced the spread of PRV.(ABSTRACT TRUNCATED AT 250 WORDS)

Serological survey of exposure to Erysipelothrix rhusiopathiae in poultry in New Zealand.

AIMS: To estimate the seroprevalence of antibodies to Erysipelothrix rhusiopathiae in chickens in New Zealand, and to estimate the effect of housing type, geographical location and age on seroprevalence. METHODS: A cross-sectional serological survey of a convenience sample of 545 broiler, breeder, and layer chickens in 55 flocks was conducted in 2010-2011. Birds were aged 5-83 weeks; housing types were free-range, shed, caged, and unknown; and flocks were located in the Auckland, Manawatu, North Canterbury, Otago, Taranaki, Waikato, and Wairarapa regions of New Zealand. An ELISA was used to measure antibodies to E. rhusiopathiae. Samples with an optical density reading ≥ 1.50 were considered to be positive. Logistic regression analysis was used to model the effect of housing type, geographical location and flock age on the prevalence of samples positive for antibodies to E. rhusiopathiae. RESULTS: The overall prevalence of samples with antibodies to E. rhusiopathiae was 39.8 (95% CI=35.68-44.06)% for the 545 samples, and 46/55 (84%) farms that were tested had at least one positive sample. Mean seroprevalence for types of housing was 44.2 (95% CI=37.79-50.70)% for free-range (n=240 birds), 23.7 (95% CI=17.83-30.38)% for shed (n=190), 73 (95% CI=56-86)% for caged (n=37) and 50 (95% CI=38-62)% for unknown (n=78). The disease was present in all seven geographical locations from which samples were obtained for this study. Seroprevalence increased with increasing age of birds (p<0.001); for birds ≤ 12 weeks of age it was 2 (95% CI=0.3-8)% (n=91), 13-24 weeks 29.1 (95% CI=23.34-35.46)% (n=230), 25-36 weeks 47 (95% CI=32-64)% (n=40), 37-48 weeks 75 (95% CI=51-91)% (n=20), >48 weeks 63.8 (95% CI=54.78-72.12)% (n=127). Neither housing type nor geographical location had a significant effect on the likelihood of samples being positive for antibodies to E. rhusiopathiae. CONCLUSIONS: The results of this study suggest the prevalence of erysipelas under New Zealand field conditions may be higher than expected and that the disease is significantly associated with increasing age. Housing type and geographical location appear to be unrelated to seroprevalence. CLINICAL RELEVANCE: Further study of the epidemiology of E. rhusiopathiae in chickens in New Zealand should be considered in order to minimise the extent of birds' exposure to the organism. These findings will assist in the design of further studies.

Development of an enzyme-linked immunosorbent assay for the serological detection of exposure of poultry in New Zealand to Erysipelothrix rhusiopathiae and their serological response to vaccination.

AIMS: To modify and validate an existing swine erysipelas ELISA for use with poultry serum and to assess the safety of a swine erysipelas vaccine for use in New Zealand layer birds. METHODS: An existing swine erysipelas ELISA was modified for use in domestic poultry and was validated using sera from birds injected with either 2 mL of a commercially available killed swine erysipelas vaccine (low-dose; n=12 birds), 4 mL of vaccine (high-dose; n=11 birds), or 2 mL saline (control; n=11 birds) on Day 0 and again on Day 21. Blood samples were collected on Days 0, 21, 42, and 63, and safety of the vaccine for use in layer birds was determined by assessing cloacal temperature and injection site reactions in birds at 0, 4, 24, 48, 72 and 96 h post-vaccination. RESULTS: The ELISA that was developed had a diagnostic sensitivity and specificity of 93% and 98%, respectively, after being optimised for a positive cut-off at an optical density (OD) ≥ 1.50 read at 450-nm wavelength. OD readings were higher on Days 21, 42, and 63 than Day 0 in both the low-dose and high-dose groups (p<0.05), and differed amongst the three groups on Days 21, 42, and 63 (p<0.05), suggesting that vaccination using either dose induced detectable levels of antibody, even after a single dose. In addition, the high-dose protocol induced higher levels of antibody production than the low-dose protocol. No local or systemic reactions to the vaccine were observed and cloacal temperatures remained in the normal biological range after vaccination. CONCLUSIONS: The ELISA that was developed had satisfactory diagnostic performance characteristics and the vaccine appeared to be safe for use in layer birds. However, the study design did not permit an assessment of the vaccine's efficacy to protect birds from clinical erysipelas. CLINICAL RELEVANCE: A diagnostic ELISA has been developed for determining the exposure of layer birds to E. rhusiopathiae. The test will be useful for monitoring flock-level erysipelas, response to vaccination, and in epidemiological studies designed to identify risk factors for exposure to the disease.

Effects of blood sample mishandling on ELISA results for infectious bronchitis virus, avian encephalomyelitis virus and chicken anaemia virus.

This study determined the effect of sample mishandling on the performance of ELISAs for detection of antibodies against infectious bronchitis virus (IBV), avian encephalomyelitis virus (AEV) and chicken anaemia virus (CAV) in the serum of chickens. The effects of five different sample mishandling treatments were assessed: heat treatment, repetitive freezing and thawing and three levels of severity of haemolysis. These mishandling treatments simulated different conditions that might occur during routine blood collection, transport or storage in a clinical practice setting. Each mishandling treatment was experimentally applied under laboratory conditions and then samples were assayed for antibodies against IBV, AEV and CAV using commercial ELISA kits. Severe haemolysis had the most consistent detrimental effect on ELISA performance, producing results that were significantly different from the reference standard in all three ELISAs, although the direction of the effect varied (less positive for the IBV and CAV assays; more positive for the AEV assay). Moderate levels of haemolysis had a similar, but less consistent, effect to that of severe haemolysis, producing results that were significantly different from the reference standard only for the IBV (less positive) and AEV (more positive) ELISAs. Repetitive freeze-thawing also produced a significant effect on ELISA results for IBV (less positive) and AEV (more positive). The IBV ELISA appeared to be most susceptible to the effects of serum maltreatment. The findings from this study suggest that unpredictable variation in the results of ELISAs can occur due to different sample mishandling treatments.