A cross-sectional study of colic in horses on thoroughbred training premises in the British Isles in 1997

The aim of this study was to estimate the incidence of colic in horses in Thoroughbred training premises in 1997 in the British Isles. The seasonal pattern and outcome of colic episodes were also investigated, together with any association between premises level variables and colic. Data were collected by a postal questionnaire. The results showed a colic incidence density of 7.19/100 horse years (s.e. 0.42) and a cumulative incidence of colic of 5.80% (s.e. 0.30). Premises were grouped according to whether they had more Flat than National Hunt horses (Flat premises) or more National Hunt than Flat horses (NH premises). A similar rate of colic episodes was found in each group. However, significantly higher cumulative incidences of one episode of colic were found in the Flat premises and of multiple episodes in the NH premises, respectively. The outcome of each episode of colic showed spontaneous recovery in 28.7%, medical recovery in 63.1%, surgical recovery in 2.0% and death in 6.2% of cases. This indicated an overall mortality rate from colic of 0.45 deaths/100 horse years. The seasonal pattern of episodes of colic showed a spring and autumn peak, with significant differences in the seasonal pattern between the Flat and NH premises. Relative risk analysis and logistic regression modelling with random effects showed significant associations between the number of episodes of colic and the number of horses on the premises (allowing for the number of horses on each premises, the larger premises had a decreased risk of colic). After adjusting for the number of horses, 3 other variables were associated with colic; Flat premises (with an increased risk), the owner being the sole person looking after the horses (a decreased risk) and the premises being a combined training and breeding establishment (a decreased risk).

A prospective study of the roles of clostridium difficile and enterotoxigenic Clostridium perfringens in equine diarrhoea

Faecal samples from adult horses and from foals with diarrhoea or with normal faeces were evaluated for the presence of Clostridium difficile, C. difficile toxins, C. perfringens enterotoxin (CPE) and C. perfringens spore counts. Clostridium difficile was isolated from 7/55 horses (12.7%) and 11/31 foals (35.5%) with colitis, but from 1/255 normal adults (0.4%) and 0/47 normal foals (P<0.001). Clostridium difficile toxins A and/or B were detected in 12/55 diarrhoeic adults (21.8%) and 5/30 diarrhoeic foals (16.7%) but in only 1/83 adults (1.2%) and 0/21 foals with normal faeces (P<0.001 and P<0.05, respectively). Clostridium perfringens enterotoxin was detected in 9/47 diarrhoeic adults (19%) and 8/28 diarrhoeic foals (28.6%), but was not detected in 47 adult horses (P<0.002) or 4 foals (P = 0.22) with normal faeces. The positive predictive value of isolation of C. perfringens with respect to the presence of CPE was only 60% in adult horses and 64% in foals. There was no association between total C. perfringens spore count and CPE in the faeces. The overall mortality rate from colitis was 22% for adult horses and 18% for foals. Clostridium difficile toxin-positive adult horses with colitis were less likely to survive than C. difficile-negative horses with colitis (P = 0.03). This study provides further evidence that C. difficile and enterotoxigenic C. perfringens are associated with equine enterocolitis.

Endoscopic findings of the gastric antrum and pylorus in horses: 162 cases (1996-2000)

Medical records and endoscopy images were examined for 209 horses that had gastroscopic examinations performed with a 2.5- or 3-m-long endoscope by one of the authors (MJM) during a 4-year period (1996-2000). The antrum and pylorus were viewed in 162 horses, and the duodenum was viewed in 94 horses. Of these 162 horses, the gastric squamous mucosa was seen in 157 horses and 50% or more of the glandular mucosa of the body of the stomach was seen in 156 horses. Erosions or ulcers were seen in the gastric squamous mucosa in 91 (58%) horses. Erosions or ulcers were seen in the glandular mucosa of the body of the stomach in only 8% of the horses. Lesions consisting of erosion or ulceration were seen in the antrum or pylorus in 94 (58%) horses. Lesions consisting of hyperemia and a rough or "bumpy" appearance were seen in the mucosa of the duodenum of 16 horses. An association between the presence of lesions in the squamous mucosa and the presence of lesions in the mucosa of the antrum/pylorus was examined by Fisher's exact test, and the linear association of lesion severity scores between the squamous mucosa and the mucosa of the antrum/pylorus was tested using a Monte Carlo estimate for linear-by-linear association. There was no association (P = .88) between these sites for presence of lesions or lesion severity scores. Similarly, there was no association between scores for the glandular mucosa in the gastric body and those in the antrum/pylorus. Because of the high prevalence of lesions in the antrum and pylorus of the stomachs of adult horses examined in a hospital setting, the entire stomach should be viewed during a gastroscopic examination.

Virulence and viremia characteristics of 1992 epizootic subtype IC Venezuelan equine encephalitis viruses and closely related enzootic subtype ID strains

Following a 19-year hiatus, Venezuelan equine encephalitis (VEE) reemerged in western Venezuela in December 1992. This outbreak is important in understanding VEE emergence because phylogenetic studies imply that sympatric, enzootic, subtype ID VEE viruses mutated to generate the epizootic/epidemic. Although the 1992-1993 strains belong to subtype IC, a serotype implicated in extensive outbreaks during the 1960s and in 1995, relatively small numbers of human and equine cases occurred in 1992-1993. We, therefore, evaluated the pathogenicity of these Venezuelan enzootic ID and epizootic IC viruses to determine 1) if they exhibit phenotypes like those described previously for more distantly related enzootic and epizootic strains, and 2) if the 1992-1993 outbreak was limited by the inability of these IC viruses to exploit equines as amplification hosts. All strains were virulent in mice and guinea pigs, but were benign for cotton rats, natural hosts of enzootic viruses. However, only the IC strains produced equine disease, with mean peak viremias of 10(5) suckling mouse 50% lethal doses per mL serum, and some titers exceeding 10(7). These viremias approximate those observed previously with VEE strains isolated during more extensive epizootics, suggesting that efficient equine amplification did not limit the scope and duration of the 1992-1993 outbreak. Enzootic ID virus infection protected all horses from challenge with epizootic strain P676, supporting the hypothesis that epizootics bypass regions of enzootic transmission due to natural immunization of equines by enzootic VEE viruses.

Age distributions of horses with strangulation of the small intestine by a lipoma or in the epiploic foramen: 46 cases (1994-2000)

OBJECTIVE

To test the hypothesis that strangulation of the small intestine by a lipoma or in the epiploic foramen is more common in older horses.

DESIGN

Retrospective study.

ANIMALS

46 horses.

PROCEDURE

Ages of horses with strangulation of the small intestine by a lipoma (n = 29) or in the epiploic foramen (17) were compared with ages of 79 horses with miscellaneous small intestinal lesions. Effects of increasing age on risk of the diseases of interest were examined by use of logistic regression and a 1-sided trend test for binomial proportions.

RESULTS

Mean age of the horses with strangulation in the epiploic foramen (9.6 years) was the same as that for the horses with miscellaneous small intestinal lesions (7.7), but mean age of the horses with strangulation by a lipoma (19.2) was significantly greater than that for the other groups. The proportion of horses with lipoma increased significantly with increasing age, but the proportion with strangulation in the epiploic foramen did not.

CONCLUSIONS AND CLINICAL RELEVANCE

Results refute the current suggestion that increasing age predisposes horses for strangulation of the small intestine in the epiploic foramen but support the suggestion that the risk of strangulation of the small intestine by a lipoma increases with age.

Risk factors associated with development of postoperative ileus in horses

OBJECTIVE

To determine risk factors associated with development of postoperative ileus in horses undergoing surgery for colic.

DESIGN

Case-control study.

ANIMALS

69 horses that developed ileus after surgery for colic and 307 horses that did not develop postoperative ileus.

PROCEDURE

Signalment, history, clinicopathologic data, treatment, lesions, and outcome were obtained from medical records.

RESULTS

Variables associated with increased risk of postoperative ileus included age > 10 years, Arabian breed, PCV > or = 45%, high serum concentrations of protein and albumin, anesthesia > 2.5 hours' duration, surgery > 2 hours' duration, resection and anastomosis, and lesions in the small intestine. Enterotomy reduced the risk of postoperative ileus. After multivariate logistic regression, the final model included the variables Arabian breed, PCV > or = 45%, lesion type, duration of surgery (> 2 hours vs < or = 2 hours), and pelvic flexure enterotomy.

CONCLUSIONS AND CLINICAL RELEVANCE

Results suggest that by evaluating certain factors, horses at increased risk of postoperative ileus may be recognized before the condition develops. Preventative treatment and early intervention may be instituted in these horses. Shortening surgery time and performing an enterotomy may decrease the probability of horses developing postoperative ileus.

Estimate of the national incidence of and operation-level risk factors for colic among horses in the United States, spring 1998 to spring 1999

OBJECTIVE

To estimate the national incidence of, operation-level risk factors for, and annual economic impact of colic among horses in the United States during 1998 and 1999.

DESIGN

Epidemiologic survey.

ANIMALS

21,820 horses on 1,026 horse operations in 28 states.

PROCEDURES

Horses were monitored for colic for 1 year, and results were recorded in a log that was collected quarterly. Operation-level data were collected via 4 on-site personal interviews. Associations between colic and independent variables adjusted for size of operation were determined.

RESULTS

Annual national incidence of colic in the US horse population was estimated to be 4.2 colic events/100 horses per year. Case fatality rate was 11%, and 1.4% of colic events resulted in surgery. Annual cost of colic in the United States was estimated to be $115,300,000.

CONCLUSIONS AND CLINICAL RELEVANCE

The national impact of equine colic is substantial because of the high case fatality rate.

Management of horses showing stereotypic behaviour, owner perception and the implications for welfare

A telephone survey was conducted of 100 racing stables, 100 riding schools and 100 competition establishments (8,427 horses in total) to determine what management practices were being applied to horses showing stereotypic behaviour, and to determine the underlying reasons for them by assessing the perceptions and opinions of the people working with the horses. The results indicated that horse owners are concerned about stereotypic behaviour, first, because it reduces the performance of the animal (31, 30 and 27 per cent of the owners of racing stables, riding schools and competition establishments respectively), secondly, because it has adverse clinical effects on the horse (52, 55 and 56 per cent), and thirdly, because it reduces the monetary value of the animal (45, 59 and 31 per cent). The belief that these behaviours are learnt or copied also affects the management of affected horses: they are not allowed on to the premises by 4, 32 and 17 per cent of owners of racing stables, riding schools and competition establishments, respectively; attempts are made to remove the causal factors of the stereotypy by 35, 43 and 36 per cent; the behaviours are physically prevented by 77, 67 and 79 per cent, and the affected horses are kept separate from other horses by 39, 30 and 48 per cent.

History of trichinellosis surveillance

The origin of trichinellosis, which existed in ancient times as testified by the discovery of parasite larvae on an Egyptian mummy, unfolded in several stages: discovery of encapsulated larvae (in the 1820s), identification and scientific description of these larvae (Paget & Owen, 1835), followed by experimental infestations of animals (dogs, pigs, rabbits, mice) or of humans as from 1850. The main occurrences of trichinellosis were followed with particular attention in Europe (Germany, Denmark, France, etc.) and in the United States of America at the end of the XIXth century. They affected numerous domestic animal species (pigs, horses, etc.) or wildlife and humans. Germany paid the heaviest toll with regard to the disease in humans, between 1860 and 1880, with several thousands of patients and more than 500 deaths. Different trichinellosis surveillance systems were set up in the relevant countries in the 1860s. In humans, this surveillance was carried out on affected living patients by a biopsy of the biceps muscles and subsequently by an analysis of eosinophilia (1895). In animals, surveillance was for a long time solely based on postmortem examination of the muscles of the affected animals. This method was used for the first time in 1863 in Germany, and from the 1890s, on several hundreds of thousands of pigs in Europe or in the United States of America.

Horse trichinellosis, an unresolved puzzle

In spite of routine controls to detect Trichinella larvae in horse-meat, human infections due to horse-meat consumption continue to occur in France and Italy. The epidemiology of horse trichinellosis since its discovery in 1975 is outlined, addressing the possible modes of natural transmission to horses, the need to develop more sensitive methods for detecting Trichinella larvae in horses, and the economic impact of horse trichinellosis. Investigations of human outbreaks due to horse-meat consumption have implicated single cases of inadequate veterinary controls on horses imported from non-European Union countries. In particular, most cases of human infection have been attributed to horses imported from Eastern Europe, where pig trichinellosis is re-emerging and the main source of infection in horses.

Lack of antibodies to porcine circovirus type 2 virus in beef and dairy cattle and horses in western Canada

Porcine circovirus type 2 (PCV2) is a recently recognized agent that is consistently associated with postweaning multisystemic wasting disease in swine. There are conflicting data concerning the ability of this virus to infect and cause disease in other species. To determine if normal cattle, cattle affected with various illnesses, and normal horses in endemic areas of PCV2 infection in swine have had PCV2 infections, 100 randomly selected bovine sera, 100 equine sera, and 100 colostrum samples from clinically normal dairy cattle were examined for the presence of antibodies to porcine circoviruses by using ELISAs. All samples tested were negative for antibodies to porcine circoviruses. As well, a seronegative neonatal Holstein calf and 6 seronegative, 6-month-old beef calves that were experimentally infected with PCV2 failed to develop antibodies to the virus. These results suggest that natural infection of cattle and horses with PCV2 does not occur, or is a rare event, in western Canada.

Immunodiagnosis of Trichinella infection in the horse

From 1998 to 2000, 5,267 horse sera were collected from several Trichinella regions in Romania. Sera were initially screened in laboratories in Romania, Serbia and Italy with an ELISA and a Western blot (Wb) using an excretory/secretory (ES) antigen and several conjugates (protein A, protein G, and sheep or goat anti-horse). Differences in serology results were obtained among the different conjugates and also between ELISA and Wb. Depending on the test used, specific antibodies were found at a prevalence rate of 3-6% of horses. Serum samples classified as positive were tested again by ELISA using a synthetic tyvelose glycan-BSA antigen, in Italy. All serum samples tested using this antigen were negative; in contrast, serum samples from experimentally infected horses were positive with the glycan antigen. The negative results obtained with the glycan antigen are consistent with the low prevalence of horse trichinellosis reported in the literature. Based on these results, further studies are needed to validate immunodiagnostic tests to detect Trichinella infection in horses.

A challenge of veterinary public health in the European Union: human trichinellosis due to horse meat consumption

Human trichinellosis in the European Union due to insufficiently cooked horse meat consumption has been reported in France and in Italy during the past 25 years. It occurred in several outbreaks totalling more than 3,000 patients during this period, with a low mortality and a high morbidity. Causative Trichinella species or phenotypes were determined by the International Reference Laboratory for Trichinellosis in Rome, Italy. They were: Trichinella spiralis, T. britovi and T. murelli. As the culinary habits and customs of populations cannot be changed by regulations, measures of protection of public health essentially depend on food inspection. Comprehensive studies having been conducted in the pathophysiology of Trichinella infection in horses; it was demonstrated that the localisation of larvae are quite different in horses and in pork. It resulted an instruction from the French Veterinary, Service recommending that: the sampling of muscles in horses carcasses has to be done at first in the tongue (apex), then in the diaphragm (pillars); at least 50 g have to be sampled in each site; examination for larvae has to be done with the digestion method. Such recommendations might be extended to other EU member countries then to the OIE Zoo-Sanitary Code.

Prevalence of virulent Rhodococcus equi in isolates from soil collected from two horse farms in South Africa and restriction fragment length polymorphisms of virulence plasmids in the isolates from infected foals, a dog and a monkey

The prevalence of virulent Rhodococcus equi in soil isolates from two horse farms in South Africa and nine clinical isolates from six foals, a foal foetus, a dog, and a monkey was investigated. The isolates were tested for the presence of virulence plasmid DNA and 15- to 17-kDa antigens by immunoblotting. Rhodococcus equi was isolated from almost all of the soil samples obtained from the two farms with 5.0 x 10(1) to 3.3 x 10(4) colony forming units per gram of soil. Virulent R. equi was isolated from three soil samples from one of the farms and appeared in 3.8% (three of 80 isolates), but not in any of the 182 isolates from the other farm. Of the three virulent R. equi isolates, one contained an 85-kb type I plasmid and two an 87-kb type I plasmid. Of nine clinical isolates from the foals, foal foetus, dog and monkey, five from the foals were virulent R. equi which expressed the virulence-associated antigens and contained a virulence plasmid 85-kb type I, and were all isolated from cases of pneumonia typical of that induced by R. equi in young foals living in widely separated areas in South Africa. The isolates from the other four foals, the dog and the monkey were avirulent R. equi.

[Equine protozoal myeloencephalitis in the Netherlands? An overview]

Equine protozoal myeloencephalitis (EPM) was diagnosed in a Dutch Warmblood gelding a few months after its export to the United States. The horse came back and was treated here. Additionally, an overview of the disease complex 'EPM' is given. Mode of infection, diagnosis of disease and its differential diagnoses, and general therapeutic options are presented. Although EPM due to infection with Sarcocystis neurona in Europe seems restricted to those horses that return or are imported from North America, the possibility of future cases of EPM caused by an infection with Neospora spp. is briefly discussed.

Risk factors to tick infestations and their occurrence on horses in the state of São Paulo, Brazil

From December 1998 to March 1999, 40 stud farms were studied in the state of São Paulo, Brazil. During visits to farms, horses reared under grazing conditions were examined for the presence of ticks. On each farm visit, horse pastures were closely inspected and a questionnaire was given to the farm supervisor with the purpose of gaining information about ecological and management variables (independent variables) that could be associated with the presence and infestation levels of ticks on the farm (dependent variables). Three tick species were found during the study. Anocentor nitens, Amblyomma cajennense and Boophilus microplus were present on horses from 38 (95%), 20 (50%) and four (10%) farms, respectively. All farms that had A. cajennense or B. microplus infestations also had A. nitens infestations. Only one of the four farms with B. microplus infestations on the horses also had A. cajennense infestations. Two farms had all horses free of ticks. There was a strong association between the presence of infestation by B. microplus on horses and the simultaneous use of a grazing area by cattle and horses (P = 0.000). There was no statistical association between any of the independent variables and the presence or infestation level of A. nitens on the horses (P > 0.20). The presence of A. cajennense was statistically associated with the presence of at least one mixed overgrowth pasture in the farm (P = 0.001). A mixed overgrowth pasture means the presence of undesired plants such as bushes and shrubs in the pasture. The presence of high levels of A. cajennense on horses was also associated with the presence of at least one mixed overgrowth pasture in the farm (P = 0.026). The regular use of acaricides was statistically associated with the presence of ticks on the horses (P < 0.05), making this procedure a result of the inefficacy of controlling ticks on the farms. The occurrence of human infestation by ticks was statistically associated with the presence of A. cajennense on the horses (P=0.000). The presence of at least one mixed overgrowth pasture on the farm was associated (P = 0.000) to either higher horse densities and to farms that did not mow all the pastures once a year, indicating that mowing all the pastures at least once a year can be considered a protective factor against the presence of mixed overgrowth pastures on the farm, and consequently, against the presence of A. cajennense on the horses.

Molecular epidemiology of Salmonella Heidelberg in an equine hospital

From 1992 to 1997, multi-drug resistant (MDR) Salmonella Heidelberg isolates were cultured from a number of horses hospitalised in a veterinary hospital in Victoria, Australia. To examine the relationships between the cases, 28 isolates from the hospital were compared by pulsed field gel electrophoresis (PFGE), IS200 element profiles, antimicrobial resistance patterns, plasmid profiles and phage typing. The PFGE patterns following digestion with XbaI and BlnI restriction endonucleases showed that the isolates from the veterinary hospital originated from a common source. These isolates also had indistinguishable IS200 profiles. However, PFGE was more discriminatory than IS200 profiles. All the veterinary hospital isolates and one independent isolate had the same antimicrobial resistance pattern and had at least one plasmid in common. Localisation of antimicrobial resistance genes indicated that the veterinary hospital isolates had more than one plasmid carrying resistance genes and that the genes encoding sulphathiazole and trimethoprim resistance were not on these plasmids. Phage typing was ineffective as 22 of the 28 isolates were untypeable. In conclusion, the combination of different methods used for epidemiological studies suggested that a single strain of MDR S. Heidelberg was isolated from horses admitted to the hospital for 6 years and caused salmonellosis in susceptible horses within that period with no apparent correlation between the antimicrobials used and retention of its MDR phenotype.

Cross-sectional study of the seroprevalence to Borrelia burgdorferi sensu lato and granulocytic Ehrlichia spp. and demographic, clinical and tick-exposure factors in Swedish horses

A cross-sectional study of the seroprevalence to Borrelia burgdorferi sensu lato and granulocytic Ehrlichia spp. in Swedish horses was conducted to evaluate associations with demographic, clinical and tick-exposure factors. From September 1997-1998, blood samples from 2018 horses were collected from the animals presented to veterinary clinics affiliated with the Swedish Horserace Totalizator Board (regardless of the primary cause for consultation). Standardized questionnaires with information both from owners and attending veterinarians accompanied each blood sample. The apparent seroprevalences to B. burgdorferi s. l. and granulocytic Ehrlichia spp. were 16.8 and 16.7%, respectively. The northern region had the lowest seroprevalences. Four logistic models were developed (controlling for demographic variables). In the disease model of seropositivity to B. burgdorferi s. l., age, breed, geographic region, the serologic titer to granulocytic Ehrlichia spp., season and the diagnosis coffin-joint arthritis were significant. In the tick-exposure model of B. burgdorferi s. l., pasture access the previous year and gender were significant. Age, racing activity, geographic region, season and the serologic titer to B. burgdorferi s. l. were associated with positivity to granulocytic Ehrlichia spp. In the tick-exposure model of granulocytic Ehrlichia spp., pasture access was a risk factor. An interaction between racing activity and geographic region showed that the risk of positive serologic reactions to Ehrlichia spp. was increased in the horse population in the south and middle of Sweden, but only among horses not used for racing. Except for the positive association between coffin-joint arthritis and serologic reactions to B. burgdorferi s. l., there were no significant associations in the multivariable models between non-specific or specific clinical sign or disease with seropositivity to either of these agents.

A multicenter, matched case-control study of risk factors for equine laminitis

Risk factors for equine laminitis were examined in a prospective case-control study of the 258 cases seen at six collaborating veterinary teaching hospitals over a 32-month period. Case-control pairs were matched on institution, clinician, and season of diagnosis. The 90% of case-control pairs (78 acute, 155 chronic) that had complete data for age, gender, and breed were used in separate conditional logistic-regression models for acute and chronic laminitis. There was an increase in risk for horses with acute laminitis from 5 to 7 years of age (OR 4.7, 95% CI 1.3-16) and from 13 to 31 years of age (OR 3.9, 95% CI 1.3-12) (both compared to <5 years); risk was increased for chronic laminitis from 10 to 14 years (OR 3, 95% CI 1.4-6.8) and from 15 to 38 years (OR 2.9, 95% CI 1.4-6.1) (both compared to <6 years). Mares - but not stallions - were more likely than geldings to develop acute laminitis (OR 2.6, 95% CI 1.1-6.2) and chronic laminitis (OR 2.0, 95% CI 1.1-3.6). In the small acute-laminitis data set, the breed variable was collapsed into three categories: Thoroughbred (THB, reference), the Quarter Horse (QH), and other (non-QH-THB). The non-QH-THB group was at increased risk of acute laminitis (OR 3.8, 95% CI 1.2-11.8). For the seven breed-group categories used in the chronic-laminitis model, however, all non-THB breed groups appeared significantly at risk as compared to the THB, with odds ratios ranging from 3.3 (95% CI 1.3-8.30) for the QH to 9.1 (95% CI 2.1-39.3) for ponies.

Frequency of and risk factors for epistaxis associated with exercise-induced pulmonary hemorrhage in horses: 251,609 race starts (1992-1997)

OBJECTIVE

To determine the frequency of epistaxis during or after racing among racehorses and identify factors associated with development of epistaxis.

DESIGN

Retrospective study.

SAMPLE POPULATION

247,564 Thoroughbred and 4,045 Anglo-Arab race starts.

PROCEDURE

Race start information (breed, age, sex, racing distance, and race type) was obtained for Thoroughbred and Anglo-Arab horses racing in Japan Racing Association-sanctioned races between 1992 and 1997. All horses that raced were examined by a veterinarian within 30 minutes of the conclusion of the race; any horse that had blood at the nostrils was examined with an endoscope. If blood was observed in the trachea, epistaxis related to exercise-induced pulmonary hemorrhage (EIPH) was diagnosed.

RESULTS

Epistaxis related to EIPH was identified following 369 race starts (0.15%). Frequency of EIPH-related epistaxis was significantly associated with race type, age, distance, and sex. Epistaxis was more common following steeplechase races than following flat races, in older horses than in horses that were 2 years old, following races < or =1,600 m long than following races between 1,601 and 2,000 m long, and in females than in sexually intact males. For horses that had an episode of epistaxis, the recurrence rate was 4.64%.

CONCLUSIONS AND CLINICAL RELEVANCE

Results suggested that frequency of EIPH-related epistaxis in racehorses is associated with the horse's age and sex, the type of race, and the distance raced. The higher frequency in shorter races suggests that higher intensity exercise of shorter duration may increase the probability of EIPH.

Epidemiology of Rhodococcus equi strains on Thoroughbred horse farms

Pulsed-field gel electrophoresis of restriction endonuclease-digested genomic DNA from a large collection of clinical isolates of Rhodococcus equi, an important pathogen of foals, was used to compare strain distribution between farms and over time. Forty-four strains were found among 209 isolates, with 5 of these accounting for over half the isolates and the 22 strains isolated more than once accounting for 90% of the isolates. The average genotypic diversity on each farm and in each year was found to be less than the genotypic diversity of the isolates taken as a whole, with 5.2% of the total diversity being due to differences between farms and 5.5% to differences between years. A small number of strains on each farm were found to have caused at least half the clinical cases of disease, and these varied between farms and, to a lesser extent, years. Most strains were found on more than one farm, and some very similar restriction patterns were found among isolates from different continents, indicating that strains can be very widespread. Multiple strains were isolated in five of the six cases in which more than one isolate from a single foal was examined, indicating that disease may commonly be caused by simultaneous infection with multiple strains. It was concluded that there are a number of different strains of R. equi which carry the vapA gene, and these strains tend to be widespread, but individual farms tend to have particular strains associated with them.

An outbreak of salmonellosis among horses at a veterinary teaching hospital

Between May 1996 and February 1997, 27 horses and a veterinary student at a veterinary teaching hospital developed apparent nosocomial Salmonella Typhimurium infection. The source of the multiple-drug resistant Salmonella Typhimurium was a neonatal foal admitted for treatment of septicemia. A high infection rate (approx 13% of hospitalized horses) coupled with a high case fatality rate (44%) for the initial 18 horses affected led to a decision to close the hospital for extensive cleaning and disinfection. Despite this effort and modification of hospital policies for infection control, 9 additional horses developed nosocomial Salmonella Typhimurium infection during the 6 months after the hospital reopened. Polymerase chain reaction testing of environmental samples was useful in identifying a potential reservoir of the organism in drains in the isolation facility. Coupled with clinical data, comparison of antimicrobial resistance patterns of Salmonella Typhimurium isolates provided a rapid initial means to support or refute nosocomial infection. Although minor changes in the genome of these isolates developed over the course of the outbreak, pulsed-field gel electrophoresis testing further supported that salmonellosis was nosocomial in all 27 horses.

Isolation of virulent Rhodococcus equi from native Japanese horses

R. equi was isolated from soil samples obtained from the environment of seven native Japanese horse breeds (Hokkaido, Kiso, Noma, Misaki, Tokara, Miyako and Yonaguni) and from fecal samples collected from three native horse breeds (Hokkaido, Kiso and Misaki). Virulent R. equi at various levels (ranging from 0.5 to 12.9%) was isolated from the feces or soil environment of Hokkaido, Kiso and Misaki horses. Isolates were investigated both for the presence of 15- to 17-kDa antigens (virulence-associated protein antigens; VapA) by colony blotting, using the monoclonal antibody 10G5, and the gene of VapA by PCR. Plasmid DNAs extracted from positive isolates were digested with restriction endonucleases, and the digestion patterns of the plasmids of virulent isolates were divided into three types. Two of the three types (87-kb type II and 90-kb type I) had already been reported in Japanese isolates, and a new type (tentatively designated as 90-kb type II) had been found in isolates from Kiso horses. Six virulent R. equi isolates from the Hokkaido horses contained an 87-kb type II plasmid. Eight of 24 isolates from the Kiso horses contained an 87-kb type II plasmid, and the remaining 16 contained a 90-kb type II (a new type) plasmid. Two isolates from the Misaki horses contained a 90-kb type I plasmid. These results demonstrate the geographic difference in the distribution of virulence plasmids in R. equi isolates among native Japanese horses.