The epidemiology of enterotoxigenic Escherichia coli (ETEC) was studied in children without diarrhoea in two remote Aboriginal communities in tropical north-western Australia. Serial surveys of the same individuals during different seasons showed that isolations were much more frequent in the wet monsoonal summer than in the dry winter. All E. coli were isolated from symptomless children aged 5 years or less; in addition, clearance of ETEC carriage without treatment was observed in all individuals within 3 months of isolation. Of the 58 ETEC strains isolated, 40 had either an H32 or an O126 antigen. Five O antigens which have never been associated with ETEC (O2, O41, O71, O77 and O157) were found. A recently proposed system to detect ETEC, using groups of polyvalent antisera, would have detected only 3 out of these 58 ETEC strains.

Epidemiological analysis of Klebsiella strains from nosocomial infections needs a simple, stable and reproducible technique of typing. We have evaluated three bacteriocin typing methods for Kiebsiella spp. by means of 15 bacteriocin-producing strains, with special reference to stability and reproducibility.

With the three methods indicator strains and clinical strains were retyped on different days under constant test procedures. Stability of bacteriocins was tested by titration immediately after preparation and after 4 weeks of storage at –70°C. Thereafter, reproducibility of typing was tested by means of these freshly prepared bacteriocin lysates and of portions of the same lysates stored at –70 °C. A moderate reproducibility was obtained with one method after two typing experi ments: 79.2% and 61.3% for indicator strains and clinical strains respectively. The other two methods gave a much lower reproducibility of 38 .5% and 32.5 % for indicator strains and of 11.1% and 25.5% for clinical strains after two experiments.

The reproducibility decreased after retyping three, four or five times. These methods are simple to perform but their usefulness for epidemiological studies is doubtful. Possible causes of the lack of reproducibility of the methods are discussed.

One hundred and nineteen volunteers were divided into five groups, and each volunteer inoculated subcutaneously with an aqueous subunit B/Hong Kong/73 vaccine containing 40, 20, 10, or 5 μg of HA or saline alone in a 0·5 ml volume. The incidence of reactions was recorded 24 h after inoculation. One month following immunization the serum HI antibody to B/Hong Kong/73 virus was measured; each volunteer was inoculated intranasally with live, attenuated influenza B (RB77) virus; and the incidence of infection by the challenge virus was determined by HI antibody response.

The results showed that the incidence of reactions to all doses of vaccine were relatively low, the severity mild, and the duration short. However, the incidence of reactions was highest for those given 40 μg HA and least for those given 5 μg HA. The serum HI antibody responses to vaccine showed a dose-response relationship. For volunteers given 40 μg HA, 22 (96%) showed a fourfold rise in antibody titre and all volunteers had antibody titres of > 40 following immunization: for volunteers given 5 μg HA the g.m.t. increased from 16·6 to 86·1; and for those given 10 and 20 μg HA the response was intermediate. Following challenge, the lowest incidence of infection was seen in volunteers given the highest dose of vaccine. However, all doses of vaccine induced some protection against challenge virus infection, and the incidence of infection was directly related to the serum antibody titre at the time of challenge. The 50% protection titre of serum HI antibody was estimated as 15 to 20.

Natural isolates of Escherichia coli, Salmonella typhimurium, Klebsiella pneumoniae, and Providencia stuartii were analysed to determine their plasmid content. This data allowed the identification of nosocomial strains of K. pneumoniae and P. stuartii and helped in the differentiation of epidemic strains of E. coli 0111 and S. typhimurium. Phenotypically similar isolates of S. typhimurium could be shown to be of independent origin using plasmid pattern analysis. The dissemination of a particular plasmid through different strains of S. typhimurium resulted in a simulation of a very widely distributed epidemic strain, because the plasmid interfered with the phage type of its host strain in addition to determining re sistance properties. Plasmid pattern analysis disclosed two independently existing but interacting epidemic processes: a bacterial ‘epidemic’ strain may become disseminated over a large territory and may predominate there for a long time; a single plasmid, however, may also become distributed through many different bacterial strains and may spread over a large territory. Plasmid pattern analysis provides a valuable and universal epidemiological laboratory method.

Human rotaviruses from the states of Rio de Janeiro, São Paulo and Pará of Brazil were analysed by RNA electrophoresis. At least some bands characteristic of rotavirus double-stranded RNA were detected in 138 (86·8%) of 159 faecal samples in which the presence of rotavirus had been demonstrated by enzyme immunoassay. Of the RNA-positi ve samples, 18 (13·0%) were classified as subgroup 1, 94 (68·1%) as subgroup 2, and 26 (18·8%) could not be classified due to absence of visible bands 10 and 11. Subgroup 2 was more frequent in the three states. All strains of subgroup 1 detected in Rio de Janeiro were associated with a single short-lived school outbreak. All strains of subgroup 1 resembled each other in electrophoretie pattern, irrespective of geographical origin, although minor differences could be detected by co-electrophoresis. Subgroup 2, on the other hand, showed a great degree of electrophoretic heterogeneity and could be divided into several sub-categories.

Over a period of about 1 year, otitis externa occurred in at least 300 visitors to a recreational park. The infections were associated with the presence of Pseudomonas aeruginosa in insufficiently chlorinated whirlpools.

A detailed investigation into the effect of modifying the incubation temperature of the complement-fixation (CF) test is described. For varicella-zoster virus cytomegalovirus and rubella virus, increasing the incubation temperature progressively increased the sensitivity of the CF test to reach a maximum at 15°C, at which temperature the geometric mean titre of seropositive samples was significantly greater than that found at 4°C. For these three viruses, each serum shown to contain IgG antibodies by ultrasensitive radioimmunoassay procedures was detected by CF following incubation at 15°C. No false-positive reactions occurred at 15°C, but it was our impression that anticomplementary activity was enhanced at this temperature. Significant increases in antibody titre at 15°C were also seen when measles virus, respiratory syncytial virus, adenovirus and Mycoplasma pneumoniae were employed as CF antigens. The results demonstrate that the CF test should be performed at 15°C if optimum sensitivity is to be achieved. The ability of the test to detect significant rises in antibody titre was not impaired at the higher incubation temperature.

Hybridomas secreting monoclonal antibodies to HBsAg and HBcAg were prepared from immunized mice. An antibody capture radioimmunoassay was used to detect and select appropriate hybrids for propagation and cloning. The advantages of this assay were discussed. The resulting monoclonal antibodies were compared with conventional polyclonal antisera for the detection of virus antigens in liver tissue and found to give excellent results.