An epidemiological study of Sonne dysentery in Dundee during the years 1971–6 was made by examining, in respect of 1420 isolates of Shigella sonnei, the discriminating power of colicine typing, antibiogram testing, biotyping and resistotyping and the stability of the markers they provided.

Colicine typing identified nine colicine types, including four not previously described. However, because types 4 and 4 var., determined by col Ib, and type U, producing no colicines, accounted for 96 % of the isolates, discrimination with colicine typing was poor. In antibiotic sensitivity tests, 13 different antibiogram patterns were noted. Less than 1 % of the isolates were sensitive to all of the eight antibiotics tested; most were multiply drug-resistant. Resistance to kanamycin, neomycin and paromomycin (KNP) was apparently due to a single resistance determinant, widely distributed in a majority (53%) of the isolates. When definitive times were chosen for reading each biotyping test, only maltose and rhamnose of the 13 ‘sugars’ tested differentiated isolates into prompt- and late-fermenting types. Though the ability to ferment rhamnose was a stable property, it discriminated only 1·5% of the minority, late-fermenting type. Resistotyping with six chemicals discriminated eight epidemiologically valid resistotypes, including three new types. However, 93 % of the isolates belonged to only three resistotypes.

Analysis of the data for isolates from 286 epidemiologically distinct episodes showed that the variability of colicine and antibiogram characters, found among isolates within, respectively, 40 and 28 % of the episodes, was generally associated with loss or gain of a plasmid (‘col Ib-KNP’) which determined production of colicine Ib and KNP resistance. These characters varied both in vivo and in vitro. Variability of resistotype characters, on the other hand, was observed in only 28 (9%) episodes, 14 of which possibly represented examples of mixed or sequential infections.

For accurate epidemiological tracing of strains of Sh. sonnei in a community, resistotyping, the technique showing the greatest discrimination and least variability of the four tested, should be included as the principal typing method.

We have tried to detect prenatal infection in 34 infants whose mothers were re-infected with rubella virus during pregnancy and in six infants whose mothers had primary subclinical rubella during pregnancy. Two methods of assessment were used: first, serum obtained soon after birth was tested for IgM antibody; secondly, serum obtained after the age of 8 months was tested for specific IgG.

The 34 women with re-infections had increases in IgG antibody titre but no IgM response. No evidence of prenatal infection was found in 33 of their 34 infants. One infant was found to have IgG antibody at the age of 11 months. This infant was IgM-negative at birth and had a rubelliform rash at the age of 5½ months; it therefore probably contracted post- rather than pre-natal infection. Fetal infection from maternal re-infection during pregnancy is probably rare.

The six women with primary subclinical rubella produced both IgG and IgM classes of antibody. Three of their six infants showed serological evidence of intrauterine infection. One, infected when its mother was 8 weeks pregnant, had clinical evidence of congenital rubella. Primary subclinical rubella during pregnancy therefore carries a significant risk of fetal infection.

Because of the difference in outcome, great care should be taken to distinguish between primary infection and re-infection when investigating symptomless increases in antibody titre after contact with rubella during pregnancy.

The faecal flora of patients with Crohn's disease was compared with that of healthy subjects. In patients with terminal ileitis, numbers of anaerobic gramnegative and coccoid rods (species of Eubacterium and Peptostreptococcus) were higher than in the controls whereas anaerobic gram-positive rods and cocci and aerobes occurred in normal numbers. The composition of the flora was neither influenced by duration of the disease nor by ileocaecal resection. In healthy subjects and patients, a chemically defined diet induced only slight changes in the flora. Thus, the flora in terminal ileitis although stable was permanently abnormal.

In patients with Crohn's colitis, abnormally low numbers of anaerobes were found in patients with severe, bloody diarrhoea while aerobic counts were normal. The flora in patients with mild colitis was similar to that in terminal ileitis. It is suggested that the abnormal flora composition might be an expression of the genetic predisposition to Crohn's disease.

Sera from patients suffering from Crohn's and other diseases and from healthy subjects were tested for agglutinins to anaerobic, gram-positive coccoid rods belonging to species of Eubacterium and Peptostreptococcus. Four strains labelled Eubacterium contortum (two strains), Eubacterium rectale and Peptostreptococcus productus were agglutinated by a higher percentage of sera from patients with Crohn's disease than from healthy subjects and from patients with liver and intestinal diseases (including ulcerative colitis), ankylosing spondylitis, granulomatous diseases, diseases of immunity and malignancies.

The agglutinins were of the IgG and IgM classes and strain-specific; the titres were low.

The results obtained with sera from patients with Crohn's disease and healthy people were subjected to discriminant analysis to estimate the probability, based on the combined results with the four strains, that a patient suffers from Crohn's disease. When sera giving an a posteriori probability ≥ 0·95 (a priori probability = 0·5) were considered positive, the test with four strains had a sensitivity of 54% and a, specificity of nearly 100%. The results with sera submitted for diagnosis showed that positive reactions in patients with a diagnosis apparently incompatible with Crohn's disease were within acceptable limits.

Different strains of Shigella sonnei present in Dundee from 1971 to 1976 were identified by a ‘multiple typing’ method in which resistotyping, used as the main method of differentiation, was supplemented by colicine typing, antibiogram testing and biotyping. At least 19 different ‘multiple types’ (MTs) were identified by combining information from the four typing techniques. The relation of the different types and their possible derivation from each other are discussed.

The practical value of multiple typing was demonstrated in a study of 247 isolates, of three distinct MTs, recovered from 178 persons involved in an extended outbreak centred primarily on day nurseries.

A few episodes that yielded isolates of different resistotypes were analysed to determine whether the cultures were: isolates of the same strain different in resistotype as a result of in vivo or in vitro variation of resistotype characters, or isolates of distinct strains of different resistotypes.

The multiple typing approach clarified the way in which different MTs emerged, persisted, disappeared or co-existed in the community during the 6 years of the study.

A large outbreak of campylobacter enteritis associated with the consumption of free school milk is described. The outbreak had an abrupt onset, and lasted for about 3 weeks; it involved mainly school children in the 2–4 and 5–7 year old age groups. During this period it was established from epidemiological and microbiological data that some 2500 children were infected. The source of the epidemic was almost certainly contaminated milk, although bacteriological proof could not be obtained. Biotyping of isolates was of considerable epidemiological value and showed the involvement of two distinct strains, one of which was dominant. Epidemiological evidence of limited person to person transmission of the infection was obtained; febrile convulsions as a prodromal sign of the illness was recognized for the first time. Strains of Campylobacter jejuni, and samples of patients' serum collected during this outbreak have enabled subsequent studies to be initiated on the serotyping of the responsible organism, on the serological response of patients infected with the organism, and on experimental infection of the bovine udder which demonstrated its potential as a source of C. jejuni in raw milk. A careful search of the literature suggests that this is the largest documented outbreak of campylobacter enteritis.

Gas production by coliform organisms and Escherichia coli from lauryl tryptose lactose broth (LTLB) was compared with that from brilliant green (lactose) bile broth (BGB). These media were compared with lauryl tryptose mannitol broth (LTMB) with and without added tryptophan for both gas and indole production. At 37 °C, LTLB and BGB were both satisfactory for gas production, but at 44 °C, LTLB gave fewer false-negative results and was thus significantly less inhibitory than BGB. However when LTLB and LTMB were compared as single-tube confirmatory media, LTLB give a high proportion of false-negative reactions in the indole test at 44 °C. The substitution of mannitol for lactose and the addition of tryptophan yielded a satisfactory medium for both confirmation of gas production and the demonstration of indole at 44 °C.

Two outbreaks of campylobacter enteritis in residential student populations are described. In the first outbreak, 34 of 240 students in one hall of residence became ill over the course of 4 or 5 days. A single serotype of C. jejuni was isolated from samples of faeces. Sera collected demonstrated that the students had little pre-existing antibody and antibody responses were demonstrated in those who were infected. About 30% of students with serological evidence for infection with C. jejuni did not have gastro-intestinal symptoms.

In the second outbreak, 77 of 300 agricultural students developed symptoms over a period of 3 weeks. Unpasteurized milk was the source of infection. Two serotypes of C. jejuni were isolated from faeces. Antibody responses were detected to both these serotypes and to another serotype, suggesting the presence of at least 3 serotypes in the milk. Bactericidal antibody consistent with recent infection was found in all students with symptoms and in 63% of students in the absence of symptoms, indicating the possible development of immunity. The high prevalence of antibody in these students was correlated with the habitual consumption of unpasteurized milk.

Anaerobic coccoid rods belonging to species of Eubacterium and Peptostreptococcus agglutinate more frequently with sera from patients with Crohn's disease than with sera from patients suffering from other diseases and from healthy subjects. Results of agglutination tests with four strains of coccoid anaerobes were used to estimate the probability that a patient suffers from Crohn's disease. The data on healthy subjects and patients with Crohn's disease were subjected to logistic discriminant analysis. With the methods and interpretation described, 52% of the patients with Crohn's disease were recognized as ‘definite’ or ‘probable’ Crohn's disease and 14% as ‘suspected’. Only 1% of the healthy subjects were classified as ‘suspected’ and none as ‘definite’ or ‘probable’ Crohn's disease.

One hundred and eighty samples of pork sausages were examined after preenrichment in buffered peptone water (P medium), for the presence of salmonellas. From each pre-enrichment four enrichments were made: (1) 0·1 ml of P medium was inoculated into 10 ml of Rappaport's medium, formula R10 (R10/43 °C), (2) 1 ml of the P medium was added to 100 ml of R10 broth (R10/100 ml/43 °C), (3) 1 ml of P medium was inoculated into 10 ml of Muller-Kauffmann tetrathionate broth (MK medium) prepared in accordance with the International Standards Organization document ISO 3565 (MK/43 °C) and (4) 10 ml of P medium were added to 100 ml of MK broth (MK/100 ml/43 °C). All the enrichments were incubated at 43 °C for 48 h. Forty-six and 47 samples were found positive with the first two enrichment methods (R10/43 °C and R10/100 ml/43 °C), while only 16 samples were found positive with the method MK/43 °C, and 27 with the method MK/100 ml/43 °C. The superiority of either one of the two R10 procedures over either one of the two MK methods is statistically highly significant (paired χ2; P < 0·001 in all four comparisons). The superiority of procedure MK/100 ml/43 °C over the method MK/43 °C is also statistically significant (P < 0·005). Both R10/43 °C and R10/100 ml/43 °C procedures had a much stronger inhibitory effect on the competing organisms (lactose- and sucrosenegative) than the two MK methods.

Unadsorbed, regular production pertussis vaccine was treated with polymyxin B sulphate at concentrations of 25,50 and 100 μg/ml. The toxic activity of treated and untreated vaccines was compared using both the limulus amoebocyte lysate test and the mouse-weight-gain test. Protective efficacy was also assessed by the mouse protection test. No discernible effect on either toxicity or efficacy of the pertussis vaccine was observed. When the vaccine was treated with 5000 μg/ml of polymyxin, endotoxic activity assessed by the limulus lysate test appeared to be abolished.

Laboratory and field trials were conducted to determine the efficacy of the anticoagulant rodenticide bromadiolone against the house mouse (Mus musculus). In laboratory feeding tests, family groups of warfarin-resistant mice maintained in pens and conditioned to feeding on plain foods were offered pinhead oatmeal bait containing bromadiolone at 0·005%. Overall mortality in replicated 21-day poison treatments was 55/58 or 94·8%. Six field trials were carried out, using the same poison bait, against mice infesting farm buildings. Treatment success, estimated from the results of census baitings conducted before and after treatment, ranged between 60·4% and 100%, mean 92·4%. In equivalent field trials using difenacoum, another newly developed anticoagulant rodenticide, the control achieved ranged between 70·2% and 100%, mean 96·0%.

Five field trials, three involving bromadiolone and two difenacoum, were not completely successful and the surviving mice were removed for laboratory examination. In 21-day toxicity tests, each animal was fed the poison bait offered to it earlier in the field. Bromadiolone and difenacoum gave kills of 12/21 (57·1%) and 9/11 (81·8%) respectively. The possible emergence of mouse populations resistant to these anticoagulants is considered.

Antibody responses to influenza virus immunizations were examined among junior high school students. The students received two doses of a commercial split-product vaccine containing influenza A H1N1 during a 2-year period following the first appearance of H1N1 virus in the winter of 1977–78. In haemagglutination-inhibition (HI) tests, the students who had been infected with H1N1 virus in 1977–78 showed a better response and wider cross-reactivity to the drift strain than the students who had not experienced earlier H1N1 influenza infection. Neuraminidase-inhibition (NAI) antibody titres after immunization depended upon a history of natural infection with H1N1 virus, since students not previously infected showed no significant NAI antibody rise after immunization.

Brodifacoum was evaluated in the laboratory against the two gerbils, Tatera indica and Meriones hurrianae and the house rat, Rattus rattus. The acute oral LD50 for these rodents was found to be 0·10 mg/kg, 0·083 mg/kg and 0·77 mg/kg respectively. Feeding tests with 0·002% and 0·005% brodifacoum produced a 100% mortality after a 3-day feeding period in the gerbils and after a 4-day period in R. rattus. The anticoagulant is toxic at both the concentrations to all three species but is less palatable in comparison to plain baits. Results of this laboratory evaluation indicates that 0·002% brodifacoum-treated bait can be effectively used against T. indica, M. hurrianae and R. rattus.

Two recombinants (R22 and R75) of the attenuated B/USSR/69 strain Brigit and the virulent B/Hong Kong/5/72 and one recombinant (R5) of Brigit and the virulent B/Hong Kong/8/73 were selected for genotypic and phenotypic characterization. All three recombinants had the growth property of the attenuated parent Brigit. Analysis of their RNA's by polyacrylamide gel electrophoresis revealed that, the strains R22 and R75 had derived all their genes from Brigit, those coding for haemagglutinin excepted. These recombinants were clinically evaluated and found to be attenuated and immunogenic. The recombinant R5 which derived, besides the gene coding for the haemagglutinin, several other genes from B/Hong Kong/8/73 was only partly attenuated since it induced influenza-like symptoms in one out of three volunteers.

It is concluded that the strain Brigit can be used as a donor of genes for the attenuation of the B/Hong Kong/5/72 virus and that recombinants of influenza type B can be identified, like influenza type A recombinants, by their RNA pattern.

Parainfluenza viruses were isolated 165 times during 14 years surveillance of the illnesses of a general practice population of around 3700. Type 1 isolations numbered 57, type 2 isolations 22 and type 3 isolations 86, representing annual rates of 33, 13 and 50 infections respectively per 10000 of population. Type 4 parainfluenza virus was not isolated. Three major classes of illness gave the following rates: sore throats (Throats) nine, acute febrile respiratory diseases (FRD) 23, acute non-febrile respiratory diseases (non-FRD) 71. The illnesses caused by the three types isolated were similar. Type 1 infections were most abundant in November and type 2 infections in December, and only 11.4 % of these types were isolated in the warm semester April through September. Type 3 infections were seasonally bi-modal, with a winter peak in January and an even greater prevalence (66% of the total) in the warm semester. Type 3 infections in the warmer months and in the later years of the Survey were usually more severe. Type 3 virus may therefore be heterogeneous, one subtype possessing and the other lacking the genetic mechanism of ‘cold-season’ prevalence. Geographical discontinuity between summer and winter isolations strengthens the case for the existence of the two subtypes of type 3 parainfluenza virus.

Type 3 infections caused the majority of the infections in very young infants. Type 2 infections were widely distributed at all ages. Females were attacked more often than males: type 1, 68.4%; type 2,636%; type 3, 53.5%. Type 3 infections in males outnumbered those in females up to 60 years of age, whereas female predominance became apparent in types 1 and 2 before 10 years of age.

All types were widely and sparsely distributed, areas of prevalence changing from year to year. Recurrences occurred only twice, both with type 3 infections. Six persons suffered both a type 1 and a type 3 infection, and one person suffered both a type 2 and a type 3 infection.

Strains of Escherichia coli isolated from adult females with symptomatic urinary tract infection were found to possess the following properties significantly more frequently than faecal strains: (i) high K-antigen titre; (ii) haemolysin; (iii) type 1 pili; (iv) mannose-resistant haemagglutination; (v) fermentation of dulcitol and salicin; (vi) O serotype 2, 6 and 75; (vii) H serotype 1. E. coli isolated from urine specimens containing significant numbers of antibody-coated bacteria were richer in these seven properties than strains from urines without detectable antibody-coated bacteria.

The O and H serotypes of E. coli obtained from patients with urinary tract infection in two New Zealand cities were compared with those reported in the world literature and found to be similar.

Antigenic variants of A/Texas/77 (H3N2) virus were selected in vitro using monoclonal antibody to virus haemagglutinin (HA). The antigenic variants and parental A/Texas/77 viruses were used to evaluate the frequency of anti-HA antibodies in the sera of children and adults using single-radial-haemolysis (SRH) tests. Twenty to 41 % of selected sera from adults, which contained antibody to the parental A/Texas/77 virus, failed to react with the different antigenic mutant viruses. A higher proportion of sera from children (37–58%) failed to react with the antigenic variants. Certain human sera and particularly those of children would appear to possess a more limited antibody repertoire to influenza HA, potentially allowing new antigenic variants to escape neutralization and spread in the community.

Ninety-five patients with recurrent pyogenic cholangitis had a range of specimens taken at laparotomy for bacterial culture. Bacteria were isolated from 68 % of cases. Escherichia coli was the predominant pathogen. Other aerobic and anaerobic intestinal bacteria were also isolated. Liver biopsy, bile and gallstones were the most rewarding specimens for culture. Infection was usually localized, but systemic infection occurred occasionally. The site of infection is probably in the liver parenchyma; however, the route by which intestinal bacteria invade the liver is not known. Sixty per cent of the cases of recurrent pyogenic cholangitis had gallstones and 20% were infected with Clonorchis sinensis.

Thirteen definite and 3 probable cases of chlamydial eye infection were diagnosed in young adults attending the Bristol Eye Hospital between June 1978 and May 1980, an incidence of about 1 case per 44000 per year in the 15 to 44-year-old community served by this hospital, and 1 per 100000 in the total population of this community. These patients presented with a sub-acute follicular conjunctivitis or kerato-conjunctivitis, which had usually been present for several weeks and had often failed to respond to topical chloramphenicol treatment before presentation. Sera obtained from 14 patients all had chlamydial antibody titres of 64 or more. Over the same period of time, an estimated 2500 patients per year from the same community attended the Venereology Department at the Bristol Royal Infirmary with genital chlamydial infections. These figures suggest that chlamydial infection of the eye complicates no more than 1 in 300 chlamydial infections of the genital tract in adults.