key: cord-0010589-2fg05fi5 authors: GARDNER, IA.; EAMENS, GJ.; TURNER, MJ.; HORNITZKY, CL. title: Toxigenic type D Pasteurella multocida in New South Wales pig herds—prevalence and factors associated with infection date: 2008-03-10 journal: Aust Vet J DOI: 10.1111/j.1751-0813.1989.tb09715.x sha: 906a66df6c31d1042fe472d1f0778b902fd324b1 doc_id: 10589 cord_uid: 2fg05fi5 Between March and July 1987, a study was undertaken to determine the prevalence of and factors associated with toxigenic type D Pasteurella multocida infection in New South Wales pig herds. Toxigenic type D P. multocida was isolated from the nasal cavities of pigs in one (2%) of 50 randomly selected herds. Toxigenic isolates were also recovered from 2 (8%) of a separate group of 25 herds that had purchased pigs from a known infected piggery in South Australia (herd SA). Snout abnormalities were present in 9.4%, 3.2% and 1.8% of grower pigs in the 3 affected herds. Isolation of toxigenic P. multocida was significantly associated (p < 0.0001) with the occurrence of clinically affected pigs in the herd. Purchase of at least 5 pigs from herd SA was associated with an elevated risk (p < 0.05) of isolation of toxigenic P. multocida. Toxin-producing isolates of Pasteurella multocida play an essential role in the aetiology of enzootic or progressive atrophic rhinitis (Pedersen and Barfod 1981; Rutter and Rojas 1982; Rutter 1983; Rutter et a1 1984) . P. multocida colonises the nasal cavity after some pre-existing damage (Rutter and Rojas 1982; Rutter 1987) , and produces toxins which cause turbinate atrophy. Isolation of toxigenic type D P. multocida from swine in 2 New South Wales pig herds was recently described (Eamens et al 1988) . Infection was thought to have been introduced with pigs from a known infected South Australian piggery (herd SA) but there were no definitive data to support this hypothesis. The present study was undertaken to determine the apparent prevalence of toxin-producing P. multocida in a random sample of New South Wales pig herds, to measure the association between recovery of toxigenic isolates and clinical disease, and to determine whether purchase of pigs from herd SA and other introduction practices were associated with an increased risk of isolation of P. multocida. The Swine Brand Register, a list maintained by the Department of Agriculture and Fisheries of all pig producers in New South Wales, was used for selection of the sample. The Register identified herds by geographical area (Pastures Protection Board) and included data, that was up to 2 years old, on the number of sows in each herd. From the Register, all herds with 20 or more sows (n = 1227) were selected as the where b = prevalence, estimated to be between 0-5% On the basis of these calculations and consideration of available resources, 50 herds were drawn from the reference population using computer-generated random numbers. Owners of selected herds were advised by mail of the reasons for the study, and the requirement for nasal swabbing and clinical evaluation of pigs in their herd. Ten replacement herds were required because 8 owners no longer had pigs, one owner did not have the minimum number of sows, and one owner refused to participate. Replacement herds were randomly selected from the Register from the same Pastures Protection Board as the non-participating herd. All 25 owners who purchased pigs from herd SA during 1985 or 1986 (PW Brownrigg, personal communication 1987 were also surveyed. None of the 25 herds (referred to in the balance of the report as high risk herds) has been previously selected in the random sample. The 2 study groups were therefore mutually exclusive. The 75 herds (50 random, 25 high risk) were visited by a Pastures Protection Board or Department of Agriculture and Fisheries veterinarian between March and July 1987. The farm visit consisted of 3 parts. Sucker, weaner, grower and finisher pigs were clinically evaluated for signs of enzootic atrophic rhinitis (Schoss 1983) . Second, nasal swabbing, as detailed below, was carried out to determine whether a herd was infected with toxigenic P. multocida. Third, the owner or manager completed a questionnaire which characterised herd size, production systems, housing characteristics, number and sources of introduced pigs in the 2 years prior to the survey, and quarantine and antibiotic treatment practices for introduced pigs. When clinical signs of atrophic rhinitis were observed, owners indicated when they first detected disease. where C = level of confidence and P = probability that a sampled pig was infected At a prevalence of 20% and 90% confidenceievel, 10 randomly selected pigs were needed in each herd to detect at least one infected pig. However, to increase the likelihood of recovery of toxigenic P. multocida, priority was given to sampling pigs with upper respiratory tract disease or snout deformities. When the veterinarian considered all pigs in a herd were clincially normal, pigs weaned in the previous 2 weeks were swabbed. The nasal swabbing procedure was carried out as follows: external nares were wiped with 70% alcohol to remove gross contamination. Mini-tipped aluminium shafted swabs* were inserted to the caudal end of the nasal cavity, placed in Amies charcoal transport medium and despatched at ambient temperature to Regional Veterinary Laboratories. Swabs were cultured onto Pedersen's NB medium (Eamens eta1 1988) and isolates confirmed as P. multocida by standard procedures (Cowan 1974) . Capsular type and toxigenity of P. multocida isolates to bovine turbinate cells were determined (Eamens et a1 1988) . The number of pigs to be swabbed was calculated using the n = log (I-C)/log (I-P) Apparent herd prevalence was calculated as the proportion of randomly selected herds with one or more isolates of toxigenic P. multocida during the survey period. A 95% confidence interval was estimated for the herd prevalence in New South Wales (Fleiss 1981) . For clinically affected herds, rates of snout abnormalities were calculated using as denominators the total number of weaner, grower and finisher pigs in the herd. Data from the 2 groups of herds were combined for further analysis. The association between the presence of clinical disease in a herd and isolation of toxigenic P. multocida was tested for significance by Fisher's exact test (Fleiss 1981) . Factors associated with isolation of P. multocida were tested at the 5% level of significance by Fisher's exact or chi-square tests (Fleiss 1981 ). Fisher's exact test was used for 2 x 2 tables and chi-square tests for tables with 3 or more levels of a factor. Descriptive variables for the 75 selected herds are presented in Table 1 and the geographical distribution of these herds is shown in Figure 1 . Pasteurella multocida P. multocida was isolated from 98 (13.3%) of 737 nasal swabs collected during the survey. One or more isolates were obtained from 31 (41.3%) of 75 herds (Table 2 ). Five isolates from 3 herds (1 random, 2 high risk) were classed as toxigenic type D. Toxigenic type A P. multocida was not detected. All positive herds were located in central-western New South Wales ( Figure I) . Based on the 50 randomly selected herds, the apparent prevalence of toxigenic P. multocida in New South Wales herds ith more than 20 sows was 2% (95% confidence interval = 0.1 -12.0%). Isolation of toxigenic P. multocida was significantly associated (p c 0.OOOl) with the occurrence of clinically affected pigs in the herd. Toxigenic P. multocida was recovered from 3 of 4 clinically affected herds that had snout deformities in 9.4%, 3.2% and 1.8% of weaner, grower and finisher pigs. Owners of affected herds first detected pigs with snout abnormalities in July 1986, January 1987, and April 1987. respectively. In the 3 affected herds, toxigenic P. 17 Negative herdshigh risk group A Negative herdsrandom group recovered from a fourth herd that had a single finisher pig (0.9% of pigs at risk) with a twisted snout, or from 71 herds that were free of snout deformities. The owner of the fourth herd had not noticed the pig with the twisted snout prior to the farm visit. The crude association between isolation of toxigenic P. multocida and introduction of pigs from herd SA was not significant (Table 3) . However, owners who purchased more than 5 pigs from herd SA were more likely to have toxigenic P. multocida isolated from pigs in their herds than owners who purchased less than 5 pigs from that source. Introduction of more than 50 pigs from other sources in the same period was also associated with an increased risk (p < 0.001) but the number of sources from which these pigs were derived was not significant (p = 0.06). Lack of quarantine and antibiotic treatment of introduced pigs was not associated (p = 0.87) with an increased risk of recovery of toxigenic isolates (Table TABLE 1 Characteristics t Eight instead of ten swabs were received from one sample herd $ Isolate was non-toxigenic; capsular type was neither D nor A 5 Two swabs (both positive) were collected from one herd, and, from another, only 7 were collected more than one of the 5 categories. 3). Associations with herd size (categories 20-50, 51-100, > 100 sows) were also not significant (p = 0.16), and commercial herds were no more likely to be affected than stud herds (p = 0.61). Toxigenic P. multocida was more likely to be isolated (p = 0.05) from herds in which all age groups (dry sows, lactating sows, weaners, growers and finishers) were housed in the same shed than those in which there was either partial or total segregation of age groups into separate sheds. A significant association existed between the number of pigs from herd SA and the number of introduced pigs from other sources (p < 0.OOOl). Owners of the 3 affected herds also had purchased more than 50 pigs from herds other than SA. Under the assumption that the number of pigs introduced from other sources was a potential confounding factor, a stratified analysis was conducted. Using this approach, purchase of more than 5 pigs from herd SA was not significantly associated (p = 0.28) with a higher isolation rate of toxigenic P. multocida. t Introduced pigs were those entering herd in the 2 years prior to sampling. Two owners did not introduce pigs during that period. $ Maximim number of sources was 6. The apparent herd prevalence of toxin-producing P. multocida of 2% indicated that toxigenic isolates were not widespread in breeding herds in New South Wales. All infected herds were directly or indirectly linked to introductions of pigs. The 2 herds in the high risk group had bought pigs from herd SA. The third herd, in the random sample group, 320 had purchased 45 gilts between September and November 1986 from the owner who had first seen clinical signs of atrophic rhinitis in July 1986. Recently detected infection in 2 other herds in central-western NSW has also been associated with introductions from the latter herd (I A Gardner, unpublished). Estimates of the prevalence of atrophic rhinitis in other Australian states have largely been based on notifications to veterinary authorities, abattoir monitoring and laboratory testing. Prevalence estimates based on such non-random sampling procedures may over or underestimate the true prevalence depending on the direction of the selection bias. The random sampling procedure in the present study allowed us to calculate an unbiased estimate of herd prevalence. Only herds with at least 20 sows were included in the sampling frame as these were herds which represented better the commercial pig industry. The diagnostic test procedure consisted of 2 parts: isolation of P. multocidu from nasal swabs followed by cell culture assay for toxigenicity. The specificity of both procedures is likely to approach loo%, but nasal swabbing is of unknown sensitivity. Although not specifying test sensitivity, Rutter (1987) indicated that if 30qo of pigs had snout deformities, P. multocida could be isolated from 50-60qo of 6-to 8-week-old pigs. Pedersen (1983) , in his review of the cultural and serological diagnosis of atrophic rhinitis, made no reference to the sensitivity of nasal swabbing. Because serological tests are not available for the detection of P. multocida in pigs, nasal swabbing and culturing onto a selective medium is currently the only practical method for detecting toxigenic isolates in live pigs. The mouse inoculation test, which is a more sensitive procedure, is unacceptable because of the large numbers of mice required for routine testing (Pedersen 1983) . Detection of toxin-producing P. multocida was strongly associated with clinical signs of atrophic rhinitis in the herd. This finding is consistent with evidence from the UK (Rutter 1983; Rutter et a1 1984) and Denmark (Pedersen 1983 ) which indicated that toxigenic P. multocida could be isolated from herds with progressive disease but not from herds without a history of atrophic rhinitis. Rutter (1987) also observed that as the prevalence of clinical signs increased so did the prevalence of infection in weaner pigs. To maximise the chance of detecting infected pigs in a herd, veterinarians swabbed pigs that had snout abnormalities, were sneezing, or had nasal discharges. If infection prevalence was low and pigs did not show signs of disease, much larger samples would have been required. For example, at 1% prevalence, 229 weaner pigs would need to have been sampled in a large herd to be 90% confident of detecting at least one infected animal. Such intensive sampling would have been impractical except in a small number of herds. In the present study the isolation rate of toxigenic P. multocida of 22.7% from pigs suspected clinically of having atrophic rhinitis was less than isolation rates from field cases reported by Rutter et a1 (1984) or the rate of 46% calculated from the data of Sawata et a1 (1984) . The reasons for the lower rates are speculative but include clearance of toxigenic isolates from the nasal cavities of affected pigs and suboptimal handling of swabs between collection and identification of colonies at laboratories. Evaluation of infection status on the basis of a single isolate per swab rather than multiple isolates may also have contributed t o differences. Our study showed a significant crude association between purchase of more than 5 pigs from herd SA and isolation of toxigenic P. multocida. Of 3 herds in this category, the 2 that purchased the greatest number of breeding pigs (9 and 13) from herd SA, were classed as infected. The probability of introducing infected pigs is a function of prevalence of infection in the source herd and the number of introductions. For example, if 10% of pigs in herd SA were infected, the probabilities of these 2 herds not introducing infection ((1-P)") were 0.39 and 0.25, respectively. Owners of both infected herds in the high risk group also introduced more than 50 pigs from other sources in the 2 years prior to the study but this factor was unlikely to be causal. All other herds except one, which supplied pigs to the 3 positive herds in the 2 years prior to the survey were sampled in a similar fashion and found free from clinical or cultural evidence of infection. Samples were not collected from one source, a large breeding company, but the consulting veterinarian indicated that the herd was free from infection (B Munro, personal communication 1988) . Therefore, we believe that introduction of more than 5 pigs from herd SA was causally associated with the risk of isolation of toxigenic P. multocida and that the observed statistical association between introduction of more than 50 pigs from other sources and isolation was a spurious one. Evaluation of factors other than introductions was limited by the small number of herds from which toxigenic isolates were recovered and the small number of herds in some exposure categories. Smith (1983) reported that management and housing factors such as poor ventilation, continuous throughput in farrowing and weaner accommodation, and high stocking densities were important risk factors for atrophic rhinitis. Such evaluations will be more appropriate should atrophic rhinitis become more prevalent in NSW. Enteric disease associated with the presence of virus particles in faeces is becoming more widely recognised in birds. This paper describes the clinical signs and lesions of an enteric disease in galahs (Cacatua roseicapi//a) and a sulphurcrested cockatoo (C. galerita) associated with the presence of virus particles in the faeces. cases. The clinical and pathological description is based on 17 Cowan and Sfeel's Manual for the Identifcation of Medical Bacteria Statistical methods for rates andproportions Cultural and serological diagnosis of atrophic rhinitis in pigs, Agriculture -Atrophic Rhinitis. Seminar in the CEC program of Coordination of Research on Clinical diagnosis of atrophic rhinitis, Agriculture -Atrophic Rhinitis. Seminar in the CEC program of Coordination of Research on Animal Pathology Infectious atrophic rhinitis -non-infectious determinants, Agriculture -Atrophic Rhinitis. Seminar in the CEC program of Coordination of Research on Animal Pathology The authors gratefully acknowledge the co-operation of piggery owners and the assistance of Pastures Protection Board and Department of Agriculture and Fisheries veterinarians with farm visits. Staff at Regional Veterinary Laboratories kindly cultured nasal swabs and identified P. multocida isolates. H Ridings provided valuable assistance with the statistical nalysis. The study was financially supported by the Swine Compensation Fund of New South Wales.