The strength and diversity of antibody responses to infection are the basis for many of the most rapid and sensitive tests in virology, yet the need to obtain the patient's consent and co–operation, to collect blood from a vein and to separate the serum before the test often deters investigators, particularly if children are involved or if there is added risk, difficulty or cost in doing the venepuncture.

Thirty-eight isolates of the SII biotype of coagulase negative staphylococci were examined by phage typing, antibiograms, DNA and immunoblot fingerprinting. Multiple isolates were available from 5 patients with clinically proven infections and from 12 further patients who were epidemiologically distinct. Only 3 of the isolates were phage typable and antibiograms produced only 9 types. All isolates were typable by both DNA and immunoblot fingerprinting and reproducibility was excellent provided the conditions were standardized. Discrimination was better with immunoblot fingerprinting (17 types) than DNA fingerprinting (11 types). Both techniques were successful in assessing the significance of multiple isolates from a single patient.

From July 1985 to March 1987, the occurrence of motile Aeromonas sp. in stool, food and environmental specimens was investigated to assess their pathogenic significance and to determine sources and routes of infection. A total of 9366 stool specimens were examined; Aeromonas was isolated from 11.1% of diarrhoeal stools and 2.2% of normal stools (P 0.001). Aeromonas counts in food specimens, which included minced beef, pork and chicken, seafood and various vegetables and their products, were unexpectedly high suggesting that infection might be food–borne rather than water–borne. About 70% of the isolates from meat products were A. hydrophila and A. sobria, while A. Caviae was the most common in sea–fish, vegetables and their products. Most A. hydrophila and A. sobria strains produced haemolysin, but haemagglutinin was found more frequently in A. Sobria.

A total of 201 new cases of tuberculosis due to bovine tubercle bacilli was confirmed in South–East England between 1977 and 1987 inclusive. This represents about 1% of all cases of tuberculosis in this region. Most cases occurred amongst older individuals of indigenous white British origin, although some younger patients of Southern European and Indian subcontinent ethnic origin were also diagnosed. The lung was the most frequent site of disease, followed by the genitor–urinary tract. In view of the known risk of transmission of disease from man to cattle via the respiratory and urinary tracts, continued surveillance of this relatively uncommon form of tuberculosis is still indicated.

Captive, healthy, adult badgers have blood containing haemoglobin at 13·3 g/dl, and 8·4×1012/l red cells with an MCV of 46·2 f1 and an MCH of 15·6 pg. They have 5·1×109 white cells/1 of which 3·29×109 are polymorphs, 1·49×109 are lymphocytes, 0·26×109 are monocytes, 0·07×109 are eosinophils and 0·01×109 are basophils. These values are somewhat less in adult animals just trapped from the wild, and are lower still in wild cubs.

Changes associated with tuberculosis are a rise, and then a fall in red blood count and white blood count, an increase in the proportion of polymorphs and momocytes and a fall in lymphocytes late in the disease. This picture is similar to that seen in widespread, disseminated, tuberculin negative, tuberculosis in humans, a type of disease similar to that occurring in many badgers.

BCG vaccination of badgers did not produce any measurable change in the blood picture.

A case-controlled study of the incidence of heat-labile enterotoxin-producing Escherichia coli (LT+ETEC) and other enteric pathogens in travellers returning to the Sheffield area was conducted from May 1984 to April 1986. LT+ETEC were found in 35 (5.8%) of 600 travellers to developed countries (mainly popular Mediterranean holiday resorts), 36 (11.3%) of 320 travellers to less-developed countries, and 11 (0.9%) of 1282 control patients whose illness was not associated with recent travel abroad. A seasonal peak of LT+ETEC infection was observed only in travellers to developed countries, with infections being significantly commoner in August to October. There was no significant deviation from expected age/sex distribution of LT+ETEC infection. Strains of LT+ETEC from travellers produced more toxin than strains from control patients, strains from travellers to less-developed countries producing most of all.

Gastroenteritis due to Salmonella montevideo occurred amongst guests attending two social functions held within 24 h, food for both having been provided by the same catering firm. Salmon was the most likely vehicle of infection in each case, although cross-contamination of other foods occurred. There were no deaths; four patients were admitted to hospital, one of whom underwent appendicectomy.

A review of salmon-associated food-poisoning outbreaks suggests that fresh salmon is an infrequent cause of food poisoning in the United Kingdom. The two outbreaks described here resulted from a failure of simple kitchen hygiene measures at a time of high ambient temperatures. Some current cooking instructions for salmon are inadequate.

During much of the present century, the population of herring gulls in Britain has increased rapidly (Cramp, Bourne & Saunders, 1973; Chabrzyk & Coulson, 1976; Monaghan & Coulson, 1977). Accompanying this changes was an increased utilization by these birds of human waste as a food supply, particularly sewage and refuse emanating from our towns and cities (Monaghan, 1983; Horton et al. 1983). This, coupled with their habits of roosting on agricultural land and water storage reservoirs and of breeding on inhabited buildings, has given rise to concern over the role of these birds in the spread of disease to man and domestic animals (e.g. Fenlon, 1981; Reilly et al. 1981; Butterfield et al. 1983; Monaghan et al. 1985; Girdwood et al. 1986).

Gastric campylobacter-like organisms (CLO) were isolated from gastric tissues removed at sacrifice from 17 mature ferrets; all animals were colonized, but no macroscopic mucosal lesions or histological features of chronic gastritis were seen. The isolates resembled Campylobacter pylori in many cultural and biochemical characteristics, and produced substantial urease activity, but there were sufficient differences from C. pylori to suggest that ferret gastric CLO represents a separate species. Comparison of human C. pylori and ferret gastric CLO may help to clucidate the pathogenicity of the former in patients with gastritis, and the ferret may serve as a useful animal model for the study of C. pylori infection.

Removal of thermophilic campylobacters from sewage at three different stages of treatment at a trickling filter sewage works has been assessed. Samples of incoming sewage, primary sedimentation effluent and final effluent were taken daily from 06.00 h to 20.00 h for 5 consecutive days and the numbers of campylobacters determined by using a most probable number method. Each sample was cultured using 2 h pre–enrichment followed by enrichment in Preston broth for 48 h and detection by plating. Over 78% of the incoming campylobacters were removed after primary sedimentation and < 0.1% remained in the final effluent. Campylobacter jejun biotype I and biotype II constituted 81.5% and 15.9% respectively of the 232 isolates tested. Serotypes common in sewage were common in human faces. It appears that the trickling filter sewage works removes most of the campylobacters entering the sewage works, but large numbers, estimated to be approximately 1010, are released into the environment daily from a local sewage works.

Campylobacter organisms isolated from water samples taken weekly from ponds and land-drains in the City of Hull were compared with isolates from humans. Of 314 campylobacter organisms isolated from patients, 237 (75.5%) of the strains were identified as typical Campylobacter jejuni, whilst of 125 identified strains isolated from the water samples, 85 (68%) resembled C. jejuni in most respects but were hippurate hydrolysis negative by the Hwang and Ederer method. The ponds and land drains in the city were therefore not a source of campylobacteriosis in the people living near these water courses. The atypical C. jejuni strains isolated from the environment may be mistaken for the C. jejuni strains which cause human infection. It is therefore essential that such strains are fully identified before attributing human and animal infections to their ingestion.

Three human volunteer experiments were performed in which river water expected to contain campylobacter organisms was ingested. Despite the ingestion of over 44000 organisms in one experiment, the subject did not suffer any symptoms, nor were campylobacter organisms excreted, nor was an antibody response to the ingested strains detected. The campylobacter organisms ingested resembled Campylobacter jejuni on colonial and microscopic morphology but were hippurate negative, and were distinct from C. coli. These environmental campylobacter strains appear to be non-pathogenic, however they may be mistaken for C. Jejuni or C.coli if they are not fully identified.

The strain of methicillin-resistant Staphylococcus aureus (MRSA) prevalent in south-east England produced in one acute hospital in a year 40 infections (bacteraemia, pneumonia and surgical wound, skin and urinary tract infections) with three attributable deaths. Rigorous measures succeeded in controlling the outbreak despite its extent, but our results suggest that less stringent measures could fail to control outbreaks of this scale. Several subsequent localized outbreaks within the hospital, probably caused by separate re-introductions of MRSA from other hospitals, were controlled by re-instigation of control measures on individual wards. The overall success of the intervention was shown by the decline in the incidence of MRSA infections from 27 in the 6 months beforehand to 2 in the most recent 6 months, and by the decline in the prevalence of colonization among patients 10 or more days in hospital from 52% immediately before the intervention to 3% 7 months after it. The incidence of attributable morbidity and death without control measures warrants a concerted effort to tackle the epidemic in all affected hospitals in Britain.

Eighteen strains of Salmonella choleraesuis subsp. choleraesuis serovar Typhi (S. typhi) isolated from blood of patiens in Japan who had visited Indonesia and returned before the onset of typhoid fever were found to possess the H-antigen z66reported by Guinée et al. (1981) but had the naturally occuring H-antigen d. They were not lysed by any of the phages of the Vi phage typing system. After passage through semi-solid medium containing H-z66 antiserum, H-antigens of 11 of 18 cultures with H-z66 phage changed to H-j and those of 2 others to H-d, while the remaining 5 were immobilized. With cultures in the H-d or H-j phages, change of the H-antigens did not occur when they were cultured in semi-solid medium containing homologous H-antiserum. These phase induction experiments as well as colony examination of original cultures suggest that H-z66 phase is unstable and tends to change of H-antigen to H-d or H-j occured without difficulty probably represented a mixed population of cells in the H-z66 and H-d or H-j phases. Since none of 856 isolates from Japan or fromimported cases of typhoid fever from Southeast Asia other than Indonesia exhibited the H-z66 antigen, it was concluded that the focus of typhois fever caused by S. typhi in H–z66 phase was probably in Indonesia.

A comparison was made of different culture media and procedures for detection of Serratia marcescens from faecal, pharyngeal and ocular swabs collected from 213 neonates. MacConkey agar and MacConkey agar with sorbitol (1%) and/or colistin (200 i.u./ml) were used both for primary isolation and after enrichment using Mossel Enterobacteriaceae broth with colistin (200 i.u./ml). The use of MacConkey agar supplemented with colistin for primary isolation improved considerably the isolation rate of S. marcescens from faecal swabs but not from pharyngeal swabs; the number of ocular isolations were insufficient to demonstrate differences between procedures. Moreover the enrichment procedures consistently increased the number of S. marcescens isolates especially from pharyngeal and ocular swabs. Use of sorbitol made detection of S'. marcescens from clinical specimens easier and time– and cost–efficient.

A sterile suspension containing 950 mouse LD50 per ml of type C botulinum toxin was exposed for various periods to different temperatures. The time required for the 99% (hundred-fold) reduction of toxicity was more than 5 years at -70°C or -20°C, 6 montha at +5 °C, 3 weeks at +20 °C, 2 weeks at +28 °C, 2 days at +37 °C, 9h at +42°C, less than 30 min at 56 °C, less than 20 min at +60 °C, and below 5 min at +80 °C. The results suggest that Clostridium botulinum type C toxin, if produced in an ecosystem of the mild climatic zone, might persist there over the winter season and cause the intoxication of vertebrates next carly spring in the absence of further microbial toxigenesis.

Using DNA probes specific for the genes encoding Vero cytotoxins 1 and 2 in hybridization experiments on faecal samples. Vero cytotoxin-producing Escherichia coli (VTEC) of serogroup 0 157 were detected in 21 of 63 cases of haemorrhagic colitis, 9 of 31 cases of non-bloody diarrhoea and 14 of 68 cases of haemolytic uraemic syndrome. Compared with these results sorbitol-MacConkey agar in conjunction with a specific 0 157 antiserum gave a sensitivity of 62% in haemorrhagic colitis, 56% in non-bloody diarrhoea and 57% in haemolytic uraemic syndrome.The specificity of this method was 100% in all three groups. This demonstrates that sorbitol-MacConkey agar is a useful screening method for the detection of VTEC of serogroup O 157 when used in conjunction with a specific homologous antiserum. However, this method does not detect VTEC belonging to other serogroups and such strains were found, particularly in cases of haemolytic uraemic syndrome.

The pH of the rument contents of cattle was recorded at slaughter; pH ranged from 5.5 to 7–8 and was not correlated with the period from saleyard to slaughter. Volatile fatty acids (VFA) were measured in 43 rumen samples; acetic, propionic and butyric were the major acids present, and the total VFA ranged from 75·9 mM/1 for samples between pH 6–7, to 7·1 mM/1 for samples of pH 8–9. Ten Salmonella strains belonging to 8 serotypes were grown in these 43 rumen samples. Where acid levels of these samples were high and pH low, most Salmonella sp. were inhibited: as the pH rose (pH 7–8) all Salmonella serotypes grew, some vigorously; as the total acid declined and pH continued to rise, growth of salmonella ceased. Serotypes and strains of the same serotype differed in their ability to grow in rumen contents, particularly when the pH was low.

The specificity of human antibodies for the two major sidechain determinants of the type II group B streptococcal (GBS) polysaccharide was examined in 90 pairs of maternal and cord sera. Using an ELISA system, total antibody was measured against the complete (sialylated) type II antigen and the proportion of antibody against the galactose determinant was estimated by inhibition with free β–methylgalactopyranosided. Mothers colonized by type II or by other GBS types had higher levels of total specific antibody (means, 33 and 4–7 µg/ml, respectively) than those not colonized (mean, 2–2 µg/ml). Cord sera averaged 1–2 µg/ml lower than maternal sera. Colonization with GBS was also associated with higher levels against the galactose determinant (mean, 1.5&µg/ml, compared to 0–7µg/ml for those not colonized). The distribution of specificities favoured antibodies against the sialic acid determinant in maternal but not cord sera. Specificity as well as antibody level may play a role in the epidemiology of GBS type II.

The investigation of an outbreak of cryptosporidiosis in a hospital day-care centre is reported. Twenty-eight (27%) children and one member of the staff were infected. Most of the cases were symptomatic and the major presenting symptom was watery diarrhoea. Shedding of oocysts continued for a mean of 12 days after diarrhoea had subsided and most of the cases were still excreting oocysts when the first follow-up sample was examined. Intermittent shedding or a carrier state were not seen in any of the cases. Giardia intestinalis cysts were detected in 19 children, 7 of whom showing mixed infection with Cryptosporidium sp., but this association was not statistically significant. All cases recovered without specific therapy. No definite proof could be found to demonstrate the source of the outbreak or the route of transmission but some observations suggesting person-to-person transmission are discussed.