was supplied with fresh chicken by a single wholesaler (wholesaler A), who obtained all of the chicken that it distributed from a single farm. This farm and its immediate surroundings were investigated to determine the source(s) of the organisms colonizing the chickens. Evidence was obtained that poultry from the farm caused sporadic human campylobacteriosis in the population served by the farm for at least 18 months after the recognition of the catering college outbreak. A report on this investigation is being prepared for publication elsewhere.

SUMMARYThree populations of small wild British rodents were studied by capture–recapture methods over a period of 3 years, a fourth group was studied for 1 year and a fifth was sampled annually for 4 years. Blood smears were taken from 3 species of rodents: the woodmouseA podemus sylvaticus, the bank voleClethrionomys glareolus(and an island sub-speciesC.g. skomerensis) and the short-tailed voleMicrotus agrestis. The smears were examined microscopically. Four genera of haemoparasitesBabesia, Hepatozoon, TrypanosomaandGrahamellawere detected.Babesiawas absent fromC.g. skomerensis, Hepatozoonwas rarely found inA. sylvaticusandM. agrestisandTrypanosomawas rare inA. sylvaticus. More males were infected than females but the difference was only statistically significant for the infection withHepatozoonin adultC.g. skomerensis. Infections withBabesiaandHepatozoonwere more prevalent in adult animals and infections withTrypanosomawere more prevalent in younger individuals. Only inC.g. skomerensiswas there a significant difference between age classes in the prevalence of infection withGrahamella– there being more adults infected. Concurrent infections were detected,Hepatozoonbeing the parasite most commonly involved. The prevalence of infections was found to be approximately proportional to the number of animals known to be alive, regardless of the season.

Four populations of small wild British rodents were studied by capture--re-capture methods over a period of three years. Samples of blood were taken from these animals and tested for antibodies to nine viruses. Animals were removed from another 11 sites around the UK, and immunosuppressed. Samples of tissue from these animals were tested for the presence of viruses by passage in laboratory mice and serum samples from some of them were tested for antibody to the nine viruses. Indications were found of the possible influence of epizootic outbreaks of certain diseases on animal populations.

A study of Campylobacter jejuni on a broiler chicken farm between 1989 and 1994 gave an estimated isolation rate of 27% (3,304 of 12,233) from a 0.9% sample of 1.44 million broiler chickens from six to eight sheds over 32 consecutive rearing flocks comprising 251 broiler shed flocks. During the study, C. jejuni was found in 35.5% of the 251 shed flocks but only 9.2% (23 of 251) had Campylobacter isolates in successive flocks, with 9 of those 23 sheds having the same serotype between consecutive flocks, indicating a low level of transmission between flocks. Analysis of a systematic sample of 484 of 3,304 (14.6%) C. jejuni isolates showed that 85% were of 10 serotype complexes but 58% were of 3 serotype complexes, indicating a high degree of strain similarity throughout the entire study. The three commonest types were detected in 8 of 32 flocks during the 5-year study period, suggesting an intermittent common external Campylobacter source. This hypothesis was tested by a retrospective cohort analysis of C. jejuni rates and types by reference to hatchery supplier of the 1-day-old chicks. Isolation rates of C. jejuni and frequency distribution of types were determined in 6-week-old broiler chickens identified by the hatchery supplying the original chicks. The isolation rate of C. jejuni in broilers, supplied by hatchery A, was 17.6%, compared to 42.9% (P < 0.0001) for broilers reared from chicks supplied by hatchery B. In two instances, when both hatcheries were used to stock the same farm flock, Campylobacter isolates were found only in those sheds with chicks supplied by hatchery B. Thus, the frequency distribution of Campylobacter types for chickens supplied by the two hatcheries over the 5-year period showed marked dissimilarity. These findings suggest that the isolation rate and type of Campylobacter isolates in broiler chickens was associated with the hatchery supplying chicks. The lack of diversity of types and the intermittent high positivity of sheds is evidence for a common source of C. jejuni introduced by vertical transmission rather than contamination at the hatchery or during transportation.

Poultry is a source of human campylobacteriosis, but a large continuous source outbreak, heretofore, has not been attributed to both a single source of poultry and single serotype of Campylobacter. Here we report an outbreak of C. jejuni affecting 6 catering college trainees and 13 patrons of a restaurant in southern England. An epidemiological investigation successfully tracked the outbreak source to the farm of origin. Frequency of occurrence of campylobacters and outbreak serotype distribution were determined in index cases, the local population, and local chicken suppliers. The source farm was investigated and the effect of interventions assessed. A single outbreak serotype of C. jejuni was isolated from trainee chefs, patrons, and chicken supplied to the college by Wholesaler A. The Campylobacter isolation rate for Wholesaler A was 89% (98% outbreak serotype), compared to 40% for non-Wholesaler A (10% outbreak serotype). The isolation rate for 14 months averaged 85% (99% outbreak serotype) in chickens grown on two farms (X and Y) supplying Wholesaler A, contributing approximately 40% to all local cases. In the research reported here, a specific strain and hygiene practice were found to be important for understanding transmission of Campylobacter from poultry to humans in this outbreak.

Seventy-two laboratory confirmed cases of Campylobacter infection were identified in people who attended a large festival in England. A case-control study was undertaken to identify the vehicle of infection. Potential risk factors included the water supply to the site, and food, bottled spring water and unpasteurised milk sold at the event. Only the consumption of unpasteurised milk showed a statistical association with having a Campylobacter infection, strongly suggesting that this was the vehicle of infection.

A survey of Skomer voles (Clethrionomys glareolus skomerensis) conducted in August of 1994 and 1995 discovered that 51% of live-trapped animals were positive for Cryptosporidium species (Coccidia). Of the positive animals, 85% were shedding C. muris in their faeces, 5% C. parvum and 10% apparently both parasite species. On the U.K. mainland, the prevalence of Cryptosporidium in the bank vole (Clethrionomys glareolus) is a quarter of that on Skomer and the species normally found is C. parvum. Interest in C. parvum stems from the often severe diarrhoeal disease cryptosporidiosis which it can cause in humans and livestock. The parasite occupies the gastro-intestinal tract and is transmitted between hosts by the faecaloral route. It has been suggested that wild rodents may be an important reservoir of infection by C. parvum for livestock. However, on Skomer island, C. muris, which is rarely found in livestock, is the dominant species infecting voles.

No abstract