In Galicia, Spain, a dramatic increase in the incidence of meningococcal disease was seen in the 1995–6. The annual incidence rose to 11 per 105 inhabitants, and 80% of identified strains were C[ratio ]2b[ratio ]P1.2,5. This led to the implementation of an intensive A+C vaccination campaign for the population aged 18 months to 19 years. During this campaign the prevalence of carriage in areas with high and low incidence was studied. Nasopharyngeal swabs were taken from 9796 subjects immediately before the administration of meningococcal vaccine, plated onto Thayer–Martin plates, incubated and sent for analysis to the Reference Laboratory for Neisseria in Spain. The prevalence of the C[ratio ]2b[ratio ]P1.2,5 strains was 0·6% (95% CI 0·29–0·88) in the high incidence area, and 0·41% (95% CI 0·00–1·04) in the low incidence area, and that of serogroup C (all strains) 1·36% (95% CI 0·80–1·80) and 0·89% (95% CI 0·09–1·69) respectively. The prevalence of N. meningitidis (all strains) was almost the same in both areas (8%). Carriers of the epidemic strain were not found in the 2–4 year age group, that most affected by the disease.

Our data showed a wide distribution but a low carriage rate of the epidemic strain C[ratio ]2b[ratio ]P1.2,5 in the high and low disease incidence areas studied; the difference in the carriage rates between the two areas was not statistically significant.

In October 1997, an outbreak of meningococcal disease occurred at the University of Southampton. All six cases were first year students living in halls of residence. Microbiological characterization of case and carrier strains, case interviews, and a meningococcal carriage prevalence survey were used to investigate the outbreak. Five cases were due to serogroup C strains, one case was unconfirmed. Serotyping did not distinguish between the strains but gene sequencing permitted identification of two distinct strains in the outbreak. Although none of the cases was known to each other, three had attended the same nightclub one evening 3–4 days before illness. Meningococcal carriage rates in undergraduates were within the range expected (147/587, 25%), but no carriers of outbreak strains were identified in this sample. The findings suggest that in communities with a high degree of social interaction, the introduction of highly virulent meningococcal strains may result in enhanced transmission with clustering of cases.

Gastroenteritis (GE) is among the most common illnesses of humans but the burden of disease, its epidemiology, and the distribution of pathogens in adults have not been fully examined. This information is needed to plan prevention strategies particularly for high-risk groups. This study is a retrospective analysis of data from the National Hospital Discharge Survey for the years 1979 through 1995 which describes the disease burden and epidemiology of hospitalizations associated with GE among adults in the United States. Diarrhoea was listed as a diagnosis on an average of 452000 hospital discharges per year representing 1·5% of all hospitalizations among adults. The annual number of GE hospitalizations has decreased by 20% from approximately 500000 in 1979 to 400000 in 1995. The aetiology of 78% of cases coded as GE was undetermined. Until the aetiology of disease can be better established, specific strategies for prevention cannot be developed.

For many years, invasive meningococcal disease in the Czech Republic occurred sporadically and was caused mainly by meningococci of serogroup B. In 1993, when a new clone (ET-15/37) emerged, the only phenotype found was C:2a:P1.2,5. In 1995, an antigenic variation of the ET-15/37 clone, B:2a:P1.2,5, occurred. The results of immunological surveys conducted in 1989 and 1996 were compared. A significantly higher proportion of 1996 sera than those collected in 1989 showed bactericidal antibodies against N. meningitidis B:2a:P1.2,5 (19·7 vs. 5·1%) and N. meningitidis C:2a:P1.2,5 (15·9 vs. 7·4%), consistent with increased herd immunity due to the spread of the new clone in the Czech Republic. There were differences in the age distribution of the positive sera.

Because enterotoxigenic Escherichia coli (ETEC) is not identified by routine stool culture methods, ETEC outbreaks may go unrecognized, and opportunities for treatment and prevention may be missed. To improve recognition of adult ETEC outbreaks, we compared them with reported outbreaks of viral gastroenteritis. During 1975–95, we identified 14 ETEC outbreaks in the United States and 7 on cruise ships, caused by 17 different serotypes and affecting 5683 persons. Median symptom prevalences were: diarrhoea 99%, abdominal cramps 82%, nausea 49%, fever 22%, vomiting 14%. The median incubation period was 42 h, and for 8 of 10 outbreaks, the mean or median duration of illness was >72 h (range 24–264). For 17 (81%) ETEC outbreaks, but for only 2 (8%) viral outbreaks, the prevalence of diarrhoea was [ges ]2·5 times the prevalence of vomiting. ETEC outbreaks may be differentiated from viral gastroenteritis outbreaks by a diarrhoea-to-vomiting prevalence ratio of [ges ]2·5 and a longer duration of illness.

A stochastic epidemic model was applied to meningococcal disease outbreaks in defined small populations such as military garrisons and schools. Meningococci are spread primarily by asymptomatic carriers and only a small proportion of those infected develop invasive disease. Bayesian predictions of numbers of invasive cases were developed, based on observed data using a stochastic epidemic model. We used additional data sets to model both disease probability and duration of carriage. Markov chain Monte Carlo sampling techniques were used to compute the full posterior distribution which summarized all information drawn together from multiple sources.

A new sulphonamide resistant (SR) C[ratio ]15[ratio ]P1.7,16 meningococcal strain, a variant of the ET-5 clone, dominated in an outbreak of 22 cases in western Norway commencing in 1995. The first eight patients were 15–21 years old from the Nordhordland area, initiating a carrier study in the local high schools. Carriage of SR serogroup C meningococci was detected by routine methods and treated with a single dose of ofloxacin 400 mg. Of 20 treated carriers, 14 harboured the outbreak strain C[ratio ]15[ratio ]P1.7,16. Vaccination of 4000 children, adolescents and close contacts of patients was also performed. After the intervention, 14 additional cases of meningococcal disease occurred, 8 due to the outbreak strain. However, incidence rates dropped from 180 to 30 per 100000 per year in the student population, but increased from 0 to 13 in the rest of the population in Nordhordland. Carriage eradication is not generally recommended in Norway. However, tracing and treating meningococcal carriage may have reduced transmission and disease in this outbreak situation.

Bacterial and viral salivary antibody testing is proving sensitive and specific, useful for epidemiological studies, and is simple and non-invasive. Salivary serogroup C polysaccharide-specific (SC PS-S) IgA and IgG were determined as a proportion of total salivary IgA and IgG in a group of UK infants who were recipients of a conjugated A/C meningococcal PS vaccine. Geometric mean concentrations (GMCs) of salivary SC PS-S IgG per mg of total IgG (μg/mg) were 0·1 pre-vaccination, rising to 8·2 post first, 16·1 post second and 29·3 post third dose of vaccine. For IgA, the corresponding GMCs in ng/mg were 0·1, 82·8, 69·6 and 91·2. Significant correlations (P<0·0001) were found between serum Ig and salivary IgG SC PS-S antibody for pre-vaccine and 1 month post each dose of vaccine suggesting that SC PS-S IgG in saliva was largely derived from serum. Of the five infants whose sera were analysed for isotype-specific responses, only traces of IgM and IgA were measurable suggesting that the SC PS-S IgA was locally produced. These findings suggests that the widespread use of meningococcal conjugate vaccines is likely to reduce nasopharyngeal carriage and may thereby induce herd immunity in the vaccinated population.

The utility of phage typing, pulsed-field gel electrophoresis (PFGE), and plasmid profile analysis was compared, to differentiate between Canadian Escherichia coli O157[ratio ]H7 strains of human (n = 27) and cattle (n = 24) origin. The diversity indices for phage typing, plasmid analysis and PFGE were 0·85, 0·69 and 0·93, respectively. PFGE and phage typing were also applied to study the role of direct transmission of E. coli O157[ratio ]H7 from cattle to humans on isolates collected from two separate farm outbreaks. PFGE showed that more than one E. coli O157[ratio ]H7 strain with varying PFGE DNA subtype profiles, may be responsible for an outbreak, and that more than one E. coli O157[ratio ]H7 subtype may be circulating on a particular farm at any one time. To our knowledge, this is one of the first reports where PFGE typing was used to verify the direct transmission of E. coli O157[ratio ]H7 from cattle to humans.

To investigate the dynamics of nasopharyngeal colonization with Streptococcus pneumoniae, and to determine the prevalent serogroups/types (SGT) and their antimicrobial susceptibility, we studied 100 infants attending our well-baby clinic. Nasopharyngeal swab specimens were obtained at 6, 10, 14, 18 and 22 weeks and at 9 and 18 months of age and submitted for culture, serotyping and antimicrobial susceptibility testing of S. pneumoniae. Colonization with pneumococcus was seen on at least one occasion in 81 infants. The median age of acquisition was 11 weeks and the median duration of carriage was 1·3 months. The common SGTs identified were 6, 19, 14 and 15. SGT 1, which was a common invasive isolate in children in our hospital during this period, was not isolated from these children. Sequential colonization by 2, 3 or 4 SGTs was observed in 18, 5 and 2 children, respectively. Resistance to penicillin, chloramphenicol, cotrimoxazole and erythromycin was observed in 0, 13 (6%) 11 (5%) and 5 (3%) isolates, respectively. There was a significant difference in susceptibility to cotrimoxazole between colonizing and invasive isolates (5% vs. 40%, P<0·0001).

Restriction patterns obtained with EcoRI and Southern hybridization were used for the differentiation of tetracycline-resistant (Tetr) R plasmids in enterobaemorrhagic Escherichia coli (EHEC) O157[ratio ]H7 isolates from a mass outbreak at a kindergarten in Obihiro-City, Hokkaido, Japan, 1996. Two kinds of Tetr R plasmids of 50 and 95 kb were detected. The 50-kb plasmids were identical to each other, while the 93-kb plasmids were of three types that were very similar to each other. The tet genes of both 50- and 95-kb R plasmids were 100% identical to the tet gene of pSC101 and all plasmids hybridized to a probe for tet. Because food-origin O157 strains were sensitive to tetracycline, we concluded that such Tetr R-plasmids might transfer to drug-sensitive O157 strains in the infected individuals.

Although diphtheria is declining in Delhi, case fatality rates (CFRs) are rising. In 1997, of 143 clinically suspected cases admitted to the Infectious Diseases Hospital 45 (32%) died. We examined their records to understand the epidemiology and reasons for high CFRs. About 53% of cases were from Delhi; they were not limited to any particular area. All the deaths and 92% (131/143) of cases occurred in children below 10 years of age. Only 12% of cases had received one or more doses of DPT. Muslims contributed significantly more cases than Hindus. CFRs were significantly higher in young (P=0·03) and unvaccinated (P=0·01) children and in those who received antitoxin on the third day of illness or later (P=0·03). The study highlights the importance of improved vaccine coverage and early diagnosis and prompt administration of antitoxin in reducing CFRs for diphtheria in Delhi.

In March 1997, an outbreak of Vero cytotoxin-producing Escherichi coli O157 (VTEC) infection occurred amongst holidaymakers returning from Fuerteventura, Canary Islands. For the investigation, a confirmed case was an individual staying in Fuerteventura during March 1997, with either E. coli O157 VTEC isolated in stool, HUS or serological evidence of recent infection; a probable case was an individual with bloody diarrhoea without laboratory confirmation. Local and Europe-wide active case finding was undertaken through national centres, Salm-Net and the European Programme of Intervention Epidemiology, followed by a case-control study.

Fourteen confirmed and one probable case were identified from England (7), Finland (5), Wales (1), Sweden (1) and Denmark (1) staying in four hotels. Three of the four hotels were supplied with water from a private well which appeared to be the probable vehicle of transmission. The case-control study showed illness was associated with consumption of raw vegetables (OR 8·4, 95% CI 1·5–48·2) which may have been washed in well water. This investigation shows the importance of international collaboration in the detection and investigation of clusters of enteric infection.

In May 1991 an immunization programme against Haemophilus influenzae type b (Hib) infection began within the Oxford region. We validate a deterministic mathematical model of Hib by comparison with the incidence of disease in the Oxford region, 1985–97. The comparison of model results with observed outcome allows an exploration of some of the poorly understood properties of the immunization programme. Model results and observed incidence are consistent with a vaccine that blocks the acquisition of carriage. Similarly, the data suggest that factors other than experience of Hib carriage are likely to have generated acquired immunity to Hib disease prior to the introduction of vaccination. Hence it is unlikely that waning of vaccine-derived protection will result in a resurgence of disease. The inclusion in the immunization schedule of a booster dose, as used in other countries, would have provided very little extra benefit.

We investigated an outbreak of Salmonella enteritidis involving at least 19 British tourists returning from one hotel in another European country. A retrospective cohort study of 47 hotel guests identified lasagne as the most likely vehicle of transmission (RR 11·5; 95% CI 3·0–44·1; P<0·0001). However, difficulties in information exchange and lack of formal mechanisms to agree on the aims of the cross-national investigation hampered efficient management of the outbreak. The factors leading to contamination of the food vehicle were not identified and therefore specific action to prevent reoccurrence could not be taken. There is need to develop protocols for cross-national investigations of outbreaks in Europe which should include specifying objectives, roles and responsibilities of investigators and control agencies, with formal reporting of the outcome of the investigation.

Available data sources on disease due to Bordetella pertussis, including notifications, hospital admissions, deaths, and an enhanced laboratory-based surveillance system commenced in January 1994, were reviewed for the period 1995–7. Pertussis notifications continued their approximately 3-year cycle although at historically reduced levels. A slight seasonal increase in late summer/early autumn existed over and above a relatively constant background rate. Over time, the proportion of pertussis cases in younger, unvaccinated children, and to a lesser extent, adolescents and young adults, is increasing. There is a continuing significant and under- reported mortality associated with pertussis in the very young age group. Disease due to serotype 1,2 is on the increase despite persistent high vaccination levels and this serotype causes more severe disease. The provision of preventative antibiotics prior to disease onset reduced the severity of the disease but its use remains uncommon in England and Wales. While overall levels of pertussis notifications have declined in recent times, vaccination efficacy wanes with increasing age, and pertussis remains a significant cause of mortality and severe morbidity in the very young. This could be reduced by timely booster vaccination and increased recognition of mild disease in older cases followed by early antibiotic therapy for the very young household contacts.

Sixty-seven strains of the five described Salmonella serotypes having antigens 6,7[ratio ]c[ratio ]1,5, that is S. enterica serotype Choleraesuis sensu stricto, Choleraesuis var. Kunzendorf, Choleraesuis var. Decatur, Paratyphi C, and Typhisuis, were examined for 16S rrn profile ribotype, presence of IS200 and phenotypic characters, including rate of change of flagellar-antigen phase and nutritional character. Choleraesuis sensu stricto and its Kunzendorf variant had related but distinct ribotypes. Therefore, ribotyping appears to be a suitable method for differentiating Choleraesuis non-Kunzendorf from Choleraesuis var. Kunzendorf. Some strains of Paratyphi C had 16S profiles that resembled that of Choleraesuis non-Kunzendorf, while others resembled that of Choleraesuis var. Kunzendorf. The Typhisuis profiles were like those of Choleraesuis non-Kunzendorf, while the Choleraesuis var. Decatur profiles were unlike those of any of the other four groups. Furthermore, IS200 fingerprinting discriminated between Choleraesuis var. Decatur and the other strains with antigenic formula O6,7[ratio ]c[ratio ]1,5, and comparison of IS200 patterns showed a high degree of genetic divergence within Choleraesuis var. Decatur. Our findings show that ribotyping and IS200 fingerprinting, combined with classical microbiological methods, distinguish the groups Choleraesuis non-Kunzendorf, Choleraesuis var. Kunzendorf, Choleraesuis var. Decatur, Paratyphi C and Typhisuis.

A total of 31 strains of Vibrio cholerae O1 (10 from outbreak cases and 7 from surface water) and non-O1 (4 from clinical and 10 from surface water sources) isolated between 1993 and 1997 were examined with respect to presence of cholera enterotoxin (CT) gene by PCR-based assays, resistance to antibiotics, plasmid profiles and random amplified polymorphic DNA (RAPD) analysis. All were resistant to 9 or more of the 17 antibiotics tested. Identical antibiotic resistance patterns of the isolates may indicate that they share a common mode of developing antibiotic resistance. Furthermore, the multiple antibiotic resistance indexing showed that all strains tested originated from high risk contamination. Plasmid profile analysis by agarose gel electrophoresis showed the presence of small plasmids in 12 (7 non-O1 and 5 O1 serotypes) with sizes ranging 1·3–4·6 MDa. The CT gene was detected in all clinical isolates but was present in only 14 (6 O1 serotype and 8 non-O1 serotype) isolates from environmental waters. The genetic relatedness of the clinical and environmental Vibrio cholerae O1 and non-O1 strains was investigated by RAPD fingerprinting with four primers. The four primers generated polymorphisms in all 31 strains of Vibrio cholerae tested, producing bands ranging from <250 to 4500 bp. The RAPD profiles revealed a wide variability and no correlation with the source of isolation. This study provides evidence that Vibrio cholerae O1 and non-O1 have significant public health implications.

Twenty-two strains of enteroaggregative Escherichia coli (EAggEC), isolated from four outbreaks of diarrhoeal disease in England, were examined for a range of phenotypic attributes including the ability to produce fimbriae, haemolysins and siderophores, and cell-surface properties such as surface charge and hydrophobicity. Strains of EAggEC isolated from two of these outbreaks belonged to a diverse range of serotypes and were heterogeneous in phenotype. Strains of EAggEC isolated from the other two outbreaks belonged predominantly to serotypes O86[ratio ]H34 and O98[ratio ]H-, respectively. Only two strains expressed fimbriae and two strains produced an 18 kDa membrane associated protein (MAP), suggesting that EAggEC express a range of adhesion mechanisms to produce the cell arrangement recognized as the ‘stacked brick’ formation. The possible explanation for the diversity of EAggEC serotypes is discussed.

Campylobacter is the most commonly reported cause of gastro-intestinal infection in England and Wales, with over 50000 reported cases in 1997. The majority of human campylobacter isolates in England and Wales are C. jejuni (c. 90%) with most of the remainder being C. coli. We describe the use of phage typing as an extension to serotyping for more detailed characterization within these two species. The scheme was piloted during a study of 2407 C. jejuni and 182 C. coli strains isolated in Wales between April 1996 and March 1997. Fifty- seven C. jejuni phage types were identified, with the ten most prevalent phage types accounting for 60% of isolates tested; 16% of isolates were untypable. The most common phage type was PT 1 which represented c. 20% of isolates. A further 7% of isolates reacted with the phages but did not conform to a designated type (RDNC). Only 12 phage types were identified among C. coli, with the two most common types, PT 2 and PT 7 accounting for 75·2% of isolates. When used in conjunction with serotyping, the ability of phage typing to identify between 6 and 29 subtypes within each of the predominant HS types has enabled a further level of discrimination to be achieved that enhances the epidemiological typing of C. jejuni and C. coli.