There occurred during the planning and publication of this review: the twentieth anniversary of the start of the intensified WHO Smallpox Eradication Campaign and the tenth anniversary of the last endemic case of smallpox; debate about the fate of smallpox virus and possibly its irrevocable destruction; claims that mass smallpox vaccination campaigns may have helped the spread of AIDS in Africa; publication of the definitive account of Smallpox and its Eradication (Fenner et al. 1988), and the closure of the WHO Smallpox Eradication Unit (SEU). It is therefore perhaps an appropriate moment to assess the current status of human poxvirus infections and their epidemiology; this review concentrates on those viruses antigenically related to smallpox (i.e. orthopoxviruses).

The predominant causative organism of whooping cough in Australia is of a serotype which has normally been associated overseas with unvaccinated communities. Australian DTP vaccines pass the statutory mouse test for Bordetella pertussis potency but this test is now believed to be relatively insensitive to certain factors, especially the major type-specific agglutinogens, which are presumably also important in the human host-parasite relationship. Because endemic B. bronchiseptica infections make some laboratory animals unsatisfactory for testing B. pertussis agglutinin responses, we have developed a test in which young farm sheep were immunized with vaccines. Type-specific agglutinins in their sera were assayed after absorption of non-specific agglutinins by suspensions of selected bordetella strains. Three well-reputed European DTP vaccines and two recent batches of Australian DTP vaccine were tested and compared thus. All evoked significant agglutinin responses to the main agglutinogens.

One hundred and twenty-four strains of non-penicillinase-producing gonococci isolated in Munich in 1986 were characterized in terms of their plasmid content and protein I serovar. Eighty-two per cent of the strains belonged to serogroup 1-B with over half belonging to either serovar 1B-2 or 1B-3. Half of the 22 serogroup 1A strains belonged to serovar 1A-2. Nineteen strains (15·3%) were found to lack the 2·6 Md cryptic plasmid although seven of these strains contained the 24·4 Md conjugative plasmid. Nine of the 105 strains which harboured the cryptic plasmid also contained the conjugative plasmid. The 19 strains which lacked the cryptic plasmid comprised 10 different serovars, indicating the heterogeneous nature of this group of organisms.

During the 5-year (1981–5) surveillance period, 2322 salmonella isolations were recorded from animals and other non-human sources in Peninsular Malaysia. This was an increase of 356% over the preceding 5-year period. The 83 serotypes isolated were recovered from 41 sources. Of these 34 were new serotypes bringing the total number of serotypes isolated from non-human sources to date up 97. Food animals and edible animal products accounted for 92·2% of the total isolations, with cattle and beef accounting for 70% of the total. Salmonella dublin was the most frequently isolated serotype, whereas S. typhimurium had the widest zoological distribution. More than 80% of the non-human salmonella serotypes have also been reported in man in this country.

Between 1969 and 1984, 6564 non-typhoid salmonella strains were isolated at the Liverpool Public Health Laboratory of which 194 (3·0%) were from extraintestinal sites. Blood (34%) and urine (32%) isolates accounted for twothirds of these, with the remainder being recovered from pus and inflammatory tissue (23%), bone (5%), cerebrospinal fluid (5%) and sputum (3%). Certain serotypes tended to cause more invasive disease than others, i.e. Salmonella choleraesuis, S. dublin, S. london, S. virchow and S. panama: this association for S. london has not previously been described. The spectrum of disease caused by non-typhoid strains was broad. This survey confirms the importance of nontyphoid salmonellas as occasional causes of invasive disease and local sepsis outside the gastrointestinal tract.

The use of the methylene blue test for the examination of cows milk was investigated in an inter-laboratory survey. A poor relationship was found between results of these tests and total viable counts. Coliforms were detected in a large number of pasteurized milks, indicating frequent post-pasteurization contamination. No relationship was found between the results of the methylene blue test and the presence of coliforms.

Results from this survey highlight the need for reappraisal of the methylene blue test as a statutory method for the examination of milk. A total viable count and coliform test are suggested for providing information regarding dairy hygiene and the quality of the product at the point of retail sale.

Intestinal thermophilic Campylobacter species produce stable patterns when subjected to bacterial restriction endonuclease DNA analysis (BRENDA); this technique is therefore of considerable value in epidemiological studies. BRENDA was used to examine thermophilic Campylobacter species from humans, wild and domestic animals. One hundred and ninety-four (61%) of 316 isolates of Campylobacter jejuni from humans had BRENDA patterns which could be matched to those of animal isolates. Poultry appear to be a major source of nfection for C. jejuni in humans with nearly half (49·7%) of the human isolates giving patterns which were indistinguishable from those isolated from poultry. A total of GO BRENDA types were identified from 316 human isolates and 11 of these had the same pattern as those isolated from poultry. One of the three Campylobacter coli BRENDA types recovered from poultry was indistinguishable from a human isolate type. Pigs appear to be only a minor source of C. coli infection for humans in New Zealand. Rats were found to be infected with strains of C. jejuni with BRENDA patterns indistinguishable from those infecting numans, poultry and a horse. None of the 102 isolates of Campylobacter species from wild birds gave BRENDA patterns similar to those of isolates from humans.

The Brucellosis Card test (Brewers' Diagnostic Kits, Hynson, Westcott and Dunning, Inc., Baltimore, Md.) was evaluated in relation to the Brucelloslide test (bioMérieux, France), the microagglutination test (MAT) and the demonstration of brucella-specific IgG, IgM and IgA in an enzyme-linked immunosorbent assay (ELISA). A total of 573 serum specimens was tested. These included sera from patients with acute brucellosis (159), chronic brucellosis (23) and patients who had been diagnosed previously as having had brucella infection (155). Control groups consisted of patients with diseases other than brucellosis (52), others with noninfectious diseases (20), and healthy individuals (164). The Card test detected 100% of the patients with acute and 61% of the patients with chronic brucellosis. The sera from the control groups were all negative. Similar results were obtained with the Brucelloslide test and the MAT. The ELISA test detected brucellaspecific Ig of all classes in the serum of patients with acute brucellosis, and IgG and IgA in the serum of patients with chronic brucellosis. In the latter group, IgM was also detected in 32% of the sera. Twenty-three per cent of sera with titres of 20 by the MAT were positive on the Card test and had ELISA titres for IgM, IgG and IgA of 400. Characterization of the antibodies involved in the Card test showed that sera with IgM ELISA titres of 1600, or an IgM titres of 800 together with IgG and IgA titres ≥ 200 were Card test positive. Higher IgG (≥ 1600) plus IgA (≥ 400) titres were required to produce a positive Card test in the absence of IgM or when the IgM titre was ≤ 200. The Card test has a potential value as a rapid screening test for humans with acute brucellosis and shows similar results to Brucelloslide and MAT tests. ELISA, however, remains the most reliable test for diagnosis of brucellosis especially in patients with chronic and complicated stages of the disease.

Mice killed shortly after receiving c. 2000 spores of a type E strain of Clostridium botulinum per os were incubated at one of five chosen temperatures together with bottles of cooked meat medium seeded with a similar inoculum. After incubation the rotting carcasses were homogenized. Sterile membrane filtrates of the homogenates (10%, w/v) and pure cultures were then titrated for toxicity. Some of the main findings were confirmed with two further type E strains.

Toxicity produced at 37 °C was poor in both carcasses and cultures (200–20000 mouse intraperitoneal LD/g or ml). It was good in both systems at 30 and 23 °C, usually reaching 20000–200000 LD/g or ml, and in carcasses occasionally more; at 30 °C maximal toxicity was reached more quickly in carcasses than in cultures. Prolonged incubation (36–118 days) at 30 or 23 °C resulted in complete loss of toxicity in virtually all carcasses but not in cultures. At 16 °C the development of toxicity in carcasses was strikingly greater than in cultures. At 9 °C neither system produced more than slight toxicity after prolonged incubation. Trypsinization increased the toxicity of cultures but not usually of carcasses. Unfiltered carcass homogenate (10%, w/v) with maximal intraperitoneal toxicity was harmless for mice by mouth in doses of 0·25 ml. These findings differed in important respects from those made earlier with a type C strain.

Polymyxin E is frequently used as an oral drug for flora suppression of the gastrointestinal canal. The suppression effect is dose dependent because polymyxin E is moderately inactivated by faecal and food compounds. Three oral polymyxin E doses (150, 300, 600 mg daily) were given to six volunteers for 6 days. The Enterobacteriaceae suppression effect was compared by means of the suppression index i.e. ratio of total number of faecal samples free of Enterobacteriaceae to the total number of faecal samples. The impact on the indigenous (mostly anaerobic) flora was measured in four ways: (i) beta-aspartylglycine content; (ii) volatile fatty acid pattern; (iii) yeast overgrowth and (iv) Streptococcus faecalis decrease. Enterobacteriaceae suppression was most successful during 600 mg oral polymyxin E (suppression indices during 150, 300 and 600 mg were 0·32, 0·55 and 0·89 respectively). None of the four markers of indigenous flora alterations were positive. However, using this dosage half of the volunteers suffered rather severe gastrointestinal side-effects. Oral polymyxin E in a dosage of minimum 600 mg daily seems to possess the ideal properties of a flora suppression agent, if the gastrointestinal side-effects could be mitigated.

A mathematical model is developed to mimic the transmission dynamics of the measles virus in communities in the developing world with high population growth rates and high case fatality rates. The model is used to compare the impacts of different mass vaccination programmes upon morbidity and mortality arising from infection by measles virus. Analyses identify three conclusions of practical significance to the design of optimal vaccination programmes. First, there is no single optimum age at which to vaccinate children for all urban and rural communities in developing countries. For a given community the best age at which to vaccinate depends critically on the age distribution of cases of infection prior to the introduction of control measures. Second, numerical studies predict that the introduction of mass vaccination will induce a temporary phase of very low incidence of infection before the system settles to a new pattern of recurrent epidemics. Mass vaccination acts to lengthen the inter-epidemic period in the postvaccination period when compared with that prevailing prior to control. Third, numerical simulations suggest that two-phase and two-stage vaccination programmes are of less benefit than one-stage programmes (achieving comparable coverage) aimed at young children. The paper ends with a discussion of the needs for: improved programmes of data collection; monitoring of the impact of current vaccination programmes; and the development of models that take account of viral transmission dynamics, host demography and economic factors.

A solid-phase antigen enzyme-linked immunosorbent assay (ELISA) was developed for the detection of rubella-specific IgG subclasses. For rubella-specific IgG1 and IgG3 sera were quantitated in arbitrary units (au) by comparison with standard curves. A concentration of 3 au was taken as that indicating positivity for specific IgG1 and specific IgG3. No sera reactive for specific IgG2 and IgG4 have been found, and thus the assay reagents were controlled by testing dilutions of a standard calibrant serum containing known concentrations of the specific IgG subclasses.

Of 105 unselected sera negative for rubella antibody by radial haemolysis (RH), two gave concentrations of specific IgG1 > 3 au and both were positive by rubella latex agglutination (LA). The sensitivity of the assay for specific IgG1 was confirmed by examining 25 selected sera negative by RH but reactive by LA. Twenty-one gave concentrations > 3 au. None of these 130 was positive for specific IgG3. All 63 sera containing > 15 international units rubella antibody by RH from cases of rubella in the remote past contained specific IgG1 and eight contained specific IgG3.

In 79 cases of primary rubella, specific IgG1 developed in all cases by day 8. Specific IgG3 became detectable in all cases except one by day 16. Serum taken on day 21 from one case was negative for specific IgG3 but the absence of later sera precluded further investigation. One case had become negative for specific IgG3 by day 56.

Sera from 24 cases of rubella reinfection were examined and all contained specific IgG1. In three cases of symptomatic reinfection, specific IgG3 was detectable in two but not in the remaining case. In 2 of the 21 cases of asymptomatic reinfection only a very early or a very late serum was available. Of the remaining 19 cases, 7 had detectable specific IgG3. However, only one of 9 sera collected 30–50 days after contact contained specific IgG3. Thus for the asymptomatic patient for whom other serological tests suggest a recent rubella infection, the failure to detect specific IgG3 in sequential sera collected after contact suggests reinfection rather than primary rubella. The detection of specific IgG3 did not correlate with the presence of specific IgM.

Sera collected 6–8 weeks after rubella vaccination had detectable specific IgG1 in 32 of 33 cases and specific IgG3 in 9 of 33. The remaining vaccinee was seronegative.

Serum and cervical secretions were collected from patients with cervical dysplasia, carcinoma-in-situ (CIS), squamous cell carcinoma (cervical SCC), and controls with normal cervices, attending clinics within the West Lambeth Health District, London. Enzyme-linked immunosorbent assays were used to examine cervical secretory IgA (sIgA) and serum IgG and IgA antibodies to herpes simplex virus (HSV). Sexual and demographic factors were considered during data analysis, which involved fitting multiple linear or multiple logistic regressions to HSV antibody levels. Prevalence of sIgA-HSV and levels of serum antibodies to HSV in all groups were compared with those of gynaecology controls. Caucasian women with mild dysplasia had a significantly higher prevalence of sIgA-HSV. Serum IgG levels to HSV (IgG-HSV) were significantly elevated in women with mild dysplasia and severe dysplasia/ClS. Serum IgA levels to HSV1 (IgG-HSV1) were significantly higher in women with cervical SCC (after adjusting for smoking habits) and other genital tumours. Significantly higher levels of serum IgA to HSV2 (IgA-HSV2) were also found among Caucasian women with cervical SCC. The possible role of HSV as a co-factor in cervical carcinogenesis is discussed.

A computer program was written to analyse oligonucleotide patterns displayed by gel electrophoresis following restriction endonuclease digestion of human cytomegaloviral DNA, and was applied to an epidemiological study of the transmission of infection in a hospital special care baby unit, with regard to infant-to-infant and mother-to-infant transmission.

The program calculates the molecular weight of oligonucleotides from their mobilities, using a cubic spline curve based on the mobilities of oligonucleotides from the AD169 strain. A matching algorithm then calculates the number of unmatched fragments for each pair of viral isolates. This was used as a similarity measure which successfully distinguished mother and infant isolate pairs from epidemiologically unrelated pairs.

The program is not intended to provide fully automatic matching, but could be recommended as a screening device to pick out pairs of strains which are sufficiently similar to suggest a common source of infection, and which may warrant closer comparison. Other applications are discussed, and the possible use of densitometers to automate data entry is considered.

Avian rotaviruses were isolated from feral pigeon faeces treated with trypsin using roller tube cultures of mammalian cells. Two pigeon strains, designated as strains PO-8 and PO-13, produced a marked cytopathic effect (CPE), small intracytoplasmic inclusion bodies and high titres of infectious particles in infected MA-104 and MDBK cell lines without cell adaptation and roller drum apparatus. The pigeon rotaviruses shared a common group specific antigen with the Lincoln strain of bovine rotavirus by indirect immunofluorescence, but differed from both the Lincoln strain and the Wa strain of human rotavirus in neutralization tests.

The RNA segment profile of this virus on polyacrylamide gel electrophoresis differed from that of group A mammalian rotaviruses. The results of a serological survey suggested that antibody to pigeon rotaviruses was widespread in avian species in Japan.

A 4 year survey was undertaken in 1982–6 to examine the seasonal nature of bluetongue virus activity in Malawi. Bluetongue infection at Bwemba in Lilongwe district and Likasi in Mchinji district, both in the Central Region of Malawi, was detected by examining sera taken from calves at each site, at monthly intervals. The proportion of seronegative calves undergoing serocon version each month was used as a measure of virus activity. At both sides bluetongue virus activity was found to be most marked during the rainy season, with no activity detected during the dry season from July to September. Thus the pattern of bluetongue infection in Malawi is highly seasonal. Examination of type-specific neutralizing antibody showed that the prevalent serotypes varied from year to year.

Thirty-one ponies immunized with inactivated virus vaccine containing A/equine/Miami/63 (H3N8) virus and six seronegative ponies were experimentally challenged with the homologous virus strain. All G unvaccinated ponies and 11 out of 31 vaccinated ponies became infected. A clear relationship between pre-challenge antibody, measured by single radial haemolysis (SRH), and protection was demonstrated as judged by virus excretion, febrile responses and antibody responses. Those ponies with SRH antibody levels > 74 mm2 were completely protected against challenge infection by the intranasal route.

During the outbreak of influenza due to A (H3N3) viruses in Finland in 1985/6 virus pairs were isolated from the same clinical specimens in embryonated hens' eggs (CE) and in canine kidney cell cultures (MDCK). Some of these isolates, the E and M pairs, were distinguished by their reactions in haemagglutination inhibition (HI) tests carried out using polyclonal antisera, and by receptorbinding properties, as evidenced by differences in their elution activity from erythrocytes. Passage of the E- and M-virus isolates in the foreign host affected their serological characteristics, but the E virus did not convert to an M-like virus and the M virus did not convert to an E-like virus. Returning the viruses to grow in the host used for their isolation changed the serological reactions so that they were once more close to the reactions of the original isolates. This contrasts with the changes in receptor-binding properties. Rapid elution from hen erythrocytes, which has been described as a property of viruses binding to the SAα2,3Gal sequence in preference to SAα2,6Gal, characterized the virus passages grown solely in MDCK cell cultures. Cultivation of the M virus in CE, at any stage of its passage history, made the virus irreversibly incapable of elution. The M virus was more sensitive than the E virus to HI antibodies against heterologous viruses of the H3N2 subtype, and, when used as an antigen in HI serology, it more frequently (90% vs. 69%; P < 0·01) detected diagnostic antibody responses in patients infected with viruses of this subtype in 1985/6. Use of antigens with a different passage history in HI serology provided evidence that this superiority, which may be due to the ability of the virus to pick out anamnestic antibody responses, is unrelated to the receptor-binding peculiarity of the M virus under consideration. The results support the concept that the host cell can select a diversity of virus variant subpopulations from a single clinical specimen during isolation and subsequent cultivation procedures. Moreover, the MDCK-grown influenza viruses may correspond better than the egg-grown isolates to the natural epidemic viruses.

A total of 6346 swine sera collected at an abattoir in the city of Obihiro, Hokkaido during the years 1978–87 were tested for the presence of antibodies to swine and human influenza viruses. A high incidence of antibody to A/New Jersey/8/76 (swine type H1N1) virus was observed throughout the 10 years except for the occasional month and a single long period of 15 months. Antibodies to human H3N2 virus in swine appeared to be related to the epidemics of human influenza which occurred in the study area during the years 1980–3, but unrelated to the epidemics during the years 1984–7. A large number of swine were found to be antibody positive to a human H1N1 virus during the period April to June 1964, and a smaller number, during the period November 1986 to June 1987. Both were in relation to human influenza epidemics. However, there were long periods where human H1N1 antibodies in swine could not be found.

The first occurrence of swine influenza in Japan was recognized in 1977, when it was presumed that the disease was introduced via imported swine (Shibata elal. 1978). Further outbreaks of swine influenza and a high prevalence of antibody to the virus in Japanese swine populations have been reported by several workers (Yamane, Sukeno & Ishida, 1978; Sugimura elal. 1981; Ogawa elal. 1983). An outbreak of influenza virus infection due to an H3N2 strain was previously seen in a herd of swine in Osaka, Japan (Sugimura etal. 1975). Later the co-existence of swine (H1N1) and human (H3N2) influenza viruses was confirmed by serological and virological studies on Japanese swine populations (Onta et al. 1978; Sugimura et al. 1980; Arikawa et al. 1982). In a previous report (Miwa et al. 1986), we suggested that the swine became infected with a human H1N1 virus as piglets during an epidemic of influenza which occurred in the human population at the same time. The present study was undertaken to evaluate the changes in the prevalence of antibodies against swine and human influenza viruses in Japanese swine during the past 10 years.