Using molecular techniques, we investigated whether the clone of Vibrio cholerae O1 biotype El Tor which appeared in Calcutta, India, in 1994 has spread to other cholera endemic areas in the country. The ribotype of 31 of the 33 strains isolated from different parts of India during 1996 and 1997 was identical to the ribotype displayed by the new clone of V. cholerae O1 which emerged in Calcutta in 1994. Likewise, 12 of the 15 strains examined by pulsed-field gel electrophoresis (PFGE) showed identical profile to that exhibited by the new clone of O1. The restriction fragment length polymorphism (RFLP) of CTX genetic element of these strains also matched with the new clone of O1 which emerged after the outbreak of V. cholerae O139 in Calcutta. However, two strains (AH042 and AH046) isolated from an outbreak in Ahmedabad (western India) showed different CTX RFLP but had the same ribotype and PFGE profile as the new clone, whereas one strain from Goa (G2) showed distinct ribotype and PFGE profile and the CTX RFLP was identical to the O1 strains which prevailed before the genesis of O139 in Calcutta. The drug resistance pattern of most of the O1 strains examined in this study, except strain G2, was similar to that of the new clone of V. cholerae O1. None of the strains in this study carried plasmids. Molecular studies clearly show that the new expanded drug resistant clone of V. cholerae O1 has spread to all cholera endemic areas in India and also provide evidence for the evolution of new clones of the O1 serogroup.

The Ukraine cholera epidemic of 1994 and 1995 was caused by Vibrio cholerae O1, serotype Ogawa, biotype El Tor. This epidemic was centred in the area around Respublika Krim (Crimea) and Mykolajiv, and spread to include parts of southern Ukraine. Cases of cholera occurred between September and November of 1994 and between June and October of 1995. The 32 fatalities among 1370 recorded cases (case fatality ratio, 2·3%) occurred throughout the course of the epidemic. V. cholerae from patients with cholera produced cholera toxin and were resistant to multiple antibiotics, though no resistance plasmids were found. Conjugation experiments suggested that resistance to multiple antibiotics may be present on a self-transmissible genetic element. Environmental sources of V. cholerae O1 El Tor included sewage, sea and surface water, and fresh water and marine fish. All but one of the environmental V. cholerae isolated during the epidemic were very similar to selected isolates from patients at the same time, supporting the role of these environmental sources in the spread of disease.

The effect of a community intervention programme of antibiotics and meningitis vaccine on pharyngeal carriage of Neisseria meningitidis was investigated. Carriage rates were determined in pupils at both secondary schools (ages 11–18 years) included in the community intervention programme and compared with two schools outside the area matched for socio-economic status. A total of 1869 pupils were studied 6 months after the programmes, and 2457 pupils after 11 months.

Six months after the programme was completed there was a 72% reduction in pharyngeal carriage of Neisseria meningitidis in pupils attending the schools in the intervention area compared with pupils in the control schools. After 11 months this difference persisted in the 11–14 age group but not in the 15–18 age group. No resistance to the antibiotics used in the programme was found.

A community intervention programme of antibiotics and vaccine for the control of meningococcal disease led to a long-term reduction in Neisseria meningitidis carriage in some age groups.

Thirty-seven Vibrio cholerae and four non-cholera Vibrio isolates from Ukraine, including strains from the epidemic of 1994–5, were analysed by molecular methods. Results from PFGE and ribotyping indicated that all Ukrainian toxigenic V. cholerae were closely related to each other and to an isolate from a patient from Pakistan. A non-toxigenic river water strain obtained during the height of the epidemic was more distantly related to these V. cholerae strains, while the Vibrio parahaemolyticus isolates and Vibrio alginolyticus isolate were not closely related to V. cholerae or each other. ERIC- and REP-PCR allowed the differentiation of strains identical by other methods. The results obtained confirm that the epidemic Ukrainian strains are most closely related to seventh pandemic strains from Asia and support a hypothesis that the Ukrainian epidemic of 1994–5 was caused by toxigenic environmental strains surviving since the time of the 1991 Ukrainian epidemic or before.

Six hundred and twenty-four Vibrio cholerae O1 strains, 623 serotype Ogawa and one serotype Inaba, isolated in Romania between 1977–95 were tested to detect all changing traits concerning serogroup, serotype, biotype, phage type and resistotype patterns and subsequently, the possible epidemiological relationship among these strains. Biotyping revealed one classical, 580 eltor strains and 43 intermediary variants. When tested with Mukerjee phages, 546 (87%) strains were sensitive and 78 (13%) resistant. One phage type (M4) dominated during 1977–90, two phage types (M4 and M5) exhibited the same high frequencies during 1991, a diversity of types occurred during 1993–4 whereas in 1995, two phage types (M4 and M5) showed similar distributions again. Five patterns of drug susceptibility were successively described during 1977–95. The most prominent changes in Vibrio cholerae O1 strains were noticed during 1993–4: the highest number of non-typable strains and intermediary variants, the widest spectrum of phage types and of multidrug resistance. In 1995, the strains reverted to the previous typable forms but a new drug resistance pattern was noticed.

Meningococcal acquisition is a prerequisite for invasive disease. Three hundred and eleven male marine commando recruits were studied throughout 29 weeks of basic training to identify factors influencing meningococcal carriage and acquisition including troop number, season, smoking, respiratory infection, antibiotic usage and nasopharyngeal bacterial interference flora.

A high carriage rate on entry to training (118/311, 37·9%) and subsequent sustained high rates of meningococcal acquisition were found. Of the potential factors examined, only active and passive smoking were found to be associated significantly with meningococcal carriage on entry. The association between active smoking and meningococcal carriage was dose-dependent, with odds ratios (OR) of 2·2 (95% CIs 1·0–4·8) and 7·2 (95% CIs 2·3–22·9) for light and heavy smokers respectively. Passive smoking predisposed independently to carriage (OR 1·8, 95% CIs 1·1–3·0). Active and passive smoking combined to give an attributable risk for meningococcal carriage of 33%. In contrast, despite a high and sustained rate of meningococcal acquisition in the study population, none of the risk factors investigated, including active smoking, was associated significantly with meningococcal acquisition. No cases of meningococcal disease occurred during the 16-month study period. Therefore smoking may increase the duration of meningococcal carriage rather than the rate of acquisition, consistent with the increased risk of meningococcal disease from passive as opposed to active smoking. Public health measures that reduce the prevalence of smoking should reduce the risk of meningococcal disease.

In a comparative study, isolates of methicillin-resistant Staphylococcus aureus (MRSA) with known pulsed-field gel electrophoresis (PFGE) and bacteriophage type were analysed by polymerase chain reaction (PCR) and restriction fragment length polymorphisms (RFLP) for additional discriminatory subtyping information. PFGE was previously performed using standardized, commercially available kits and pre-programmed software. Isolates were examined for coagulase (coa) and protein A (spa) gene polymorphisms following PCR amplification of the coa hypervariable and spa repeat regions. Coa gene RFLPs produced a total of 38 distinct combined patterns after digestion with HaeIII and AluI and identified the predominant epidemic (EMRSA) types 15 and 16. A unique HaeIII restriction site was identified by RFLP and sequence analysis in the coa gene for EMRSA 15 but not EMRSA 16. The spa gene PCR yielded a total of 14 different profiles ranging from 3–18 repeats with the 2 predominant EMRSA types falling into 2 distinct groups. PCR detection of coa and spa polymorphisms offer a rapid preliminary strain identification and discriminatory subtyping information for surveillance of MRSA.

Seventy clinical strains of Vibrio cholerae O1 isolated from 1982–96 in Samutsakorn, a port city 30 km southwest of Bangkok where cholera occurs at low levels with regular seasonality, were characterized to investigate if there were any differences among the O1 strains isolated before, during and after the O139 epidemic. Pulsed-field gel electrophoresis (PFGE) typing, ribotyping and southern blot hybridization with a cholera toxin probe (CT genotyping) demonstrated several genotypes among O1 strains isolated before the emergence of V. cholerae O139. However, O1 strains isolated during and after the advent of O139 showed identical ribotypes which were distinctly different from the types identified in strains isolated before the emergence of O139. Ribotypes identified in strains during and after the advent of O139 were also demonstrated by O1 strains isolated immediately before the emergence of O139. Considering the seasonality of cholera in Samutsakorn, the identical ribotype and CT genotype and the closely related PFGE types shown by all O1 strains isolated during and after the appearance of O139 is remarkable and suggest that the V. cholerae O1 strain may reemerge from an environmental source. A subgroup of V. cholerae O1 strains isolated before the emergence of the O139 epidemic had a ribotype identical to a type demonstrated by O139 strains isolated in Thailand. Our results support similar findings in Bangladesh and India that a distinct O1 strain appeared during the O139 epidemic. However, compared with the apparent identical strain which replaced O139 in Bangladesh and India, the emerged O1 strain in Samutsakorn showed a different ribotype and CT genotype.

Enterotoxigenic Escherichia coli (ETEC) serotype O169[ratio ]H41 organisms have become the most prevalent ETEC in Japan since the first outbreak in 1991. It was assumed that the outbreaks were due to clonal spread of this new ETEC serotype. The relationship of 32 strains isolated from 6 outbreaks were examined for biotype, antibiotic susceptibility, enterotoxigenicity, protein banding pattern, lipopolysaccharide banding pattern, plasmid analysis, and ribotyping. Further, the strains were examined by haemagglutination, surface hydrophobicity, and the ability to adhere to HEp-2 cells. The present study suggests that the outbreaks were caused by multiple clones of STp-producing O169[ratio ]H41 since they showed differences in ribotype and outer membrane protein banding patterns. The strains did not agglutinate human or bovine red blood cells in a mannose-resistant manner. They adhered to HEp-2 cells in a manner resembling enteroaggregative E. coli. Five strains were examined by dot-blot tests for the colonization factor antigens CFA/I, CS1, CS2, CS3, CS4, CS5, CS6, CS7, PCFO159, PCFO166 and CFA/III. Although four strains expressed CS6, no structure for CS6 was identified. A strain that the anti-CS6 MAbs did not react with could adhere to HEp-2 cells in mannose resistant manner; thus, it is unlikely that CS6 play an important role in the adhesion to the cells. Electron microscopy studies of the O169[ratio ]H41 strains suggested that curly fimbriae, a possible new colonization factor, may be playing an important role in the adhesion of the bacteria to HEp-2 cells. In conclusion, outbreaks due to ETEC O169: H41 were caused by multiple clones, and the strains should be examined in detail for a possible new colonization factor.

An increase in the incidence of invasive and non-invasive infections caused by group A β-haemolytic streptococci (GAS) was noted in and around the town of Glynneath (population approx. 4000) in West Glamorgan, South Wales between 1 January and 30 June 1995. A total of 133 cases was ascertained with 127 (96%) occurring between 1 March and 30 June 1995. Six patients had invasive disease (one died) and all presented at the peak of the outbreak. There were 127 non-invasive cases of whom 7 were hospitalized. The outbreak was investigated to determine its extent and whether it was caused by a single M-serotype of GAS. Serotyping showed that 13 different M-serotypes were involved with the M1 serotype predominating. The overall incidence of GAS invasive disease in West Glamorgan (population 365000) increased sevenfold from a crude incidence of 0·5/105 per year in 1994 to 3·5/105 per year in 1995, but fell back to 0·75/105 per year in 1996. Eighty-two (80%) out of 102 individuals affected by GAS replied to a health questionnaire; sore throat was the commonest symptom reported (97%). Thirty-nine of these index cases identified at least one other member of their household who had experienced similar symptoms. The interval between the onset of illness in members of a single household was 0·83 days with a mean of 22 days. The mean duration of illness was 13·5 days and 61% of patients were treated with penicillin V for a mean duration of 9·3 days. Twenty-one per cent of GAS isolates were erythromycin-resistant and the M4 and M6 serotypes were especially resistant to erythromycin (87·5 and 100% resistance, respectively). Penicillin V failed to eradicate GAS from the throats of 25% of assessable patients. In this community, an outbreak of non-invasive disease caused by GAS was linked in time and place with an outbreak of serious invasive disease.

We conducted a 1-year case-control study of sporadic vibrio infections to identify risk factors related to consumption of seafood products in two coastal areas of Louisiana and Texas. Twenty-six persons with sporadic vibrio infections and 77 matched controls were enrolled. Multivariate analysis revealed that crayfish (P<0·025) and raw oysters (P<0·009) were independently associated with illness. Species-specific analysis revealed an association between consumption of cooked crayfish and Vibrio parahemolyticus infection (or 9·24, P<0·05). No crayfish consumption was reported by persons with V. vulnificus infection. Although crayfish had been suspected as a vehicle for foodborne disease, this is the first time to our knowledge that consumption of cooked crayfish has been demonstrated to be associated with vibrio infection.

Shigellosis is common among children in the Andaman and Nicobar islands. Our experience showed two distinct features of shigellosis within a span of 3 years in 1994–6: (i) changing patterns of serotype or subtype specific shigellosis and (ii) emergence of multidrug resistant isolates with changing R-patterns. The rate of isolation was 10·4–27·9% with the rate of isolation of Shigella flexneri interchanging with S. dysenteriae alternately. In 1994, S. flexneri superseded S. dysenteriae (48·6% vs. 33·3%; P<0·05) while S. dysenteriae dominated over S. flexneri in 1995 (54·7% vs. 34·0%; P<0·05). The picture reversed again in 1996 (63·0% vs. 22·2%; P<0·05). Among shigellae isolates, the commonest serotypes were S. dysenteriae type 1 and S. flexneri type 2a. Isolated shigellae were of multidrug resistant type. Seven R-patterns were observed in 1994, while 8R-patterns were observed during the next year with the emergence of nalidixic acid resistance. In 1996, emergence of gentamicin resistance was also observed. All isolates were resistant to ampicillin and sensitive to quinolones. The MIC of nalidixic acid and gentamicin are [egs ]128 μg/ml and [egs ]64 μg/ml respectively. These changing trends in shigellosis has important public health significance.

An outbreak of gastrointestinal illness affecting 12 of 29 customers of a ‘Hawaiian’ theme restaurant specializing in stir-fried food occurred in Cardiff, Wales in February 1997. Campylobacter jejuni serotype HS50 phage type 49 (PT49) was isolated from 5 cases. A total of 47 isolates of C. jejuni HS50 PT49 were identified from Wales during 1997, of which 11 were isolated in late February or early March and from the Cardiff area. In the outbreak, illness was associated with eating stir-fried chicken pieces (relative risk 4·81, 95% confidence interval (CI) 0·76–30·44, P=0·03) and a dose-response relationship between risk of illness and amount of chicken consumed was observed (χ2-test for linear trend 3·96, P=0·047). Undercooking of chicken was probably due to a combination of inadequate cooking time and use of large chicken pieces. This is the first time that stir-fried food has been associated with a campylobacter outbreak. The incident also illustrates the value of routine campylobacter subtyping in supporting outbreak investigation.

Previous respiratory tract infections (RTI) and antibiotics consumption as possible risk factors for extended duration of PRP carriage were investigated in 24 children (cases) with previous carriage of penicillin-resistant pneumococci (PRP) for a duration exceeding 120 days (median 168 days) and a control group of 53 children with a duration of PRP carriage less than 90 days (median 21 days). The cases had experienced 0·99 episodes of acute otitis media (AOM) per life-year compared to 0·79 episodes in the controls (P=0·32). For antibiotic-treated RTI other than AOM, the corresponding numbers were 0·49 and 0·29 episodes per life-year, respectively (P=0·01). No differences in antibiotic consumption in the 3 months preceding the carriage, nor during the carriage period were noted. Other factors than impaired host defence to respiratory tract pathogens or antibiotics consumption seem to be more important in determining the duration of PRP carriage.

An apparent increase in the incidence of S. Brandenburg in New Zealand, coupled with the possibility that the virulence of the organism may also be changing, prompted this study.

Three typing methods: macro-restriction fragment length polymorphism (MRFLP) profiling using pulsed field gel electrophoresis (PFGE), plasmid profiling and antimicrobial susceptibility profiling were used to determine strain diversity amongst 115 recent and historical isolates of S. Brandenburg from both human cases and non-human sources.

Antimicrobial resistance was noted only in three isolates. Plasmids of varying sizes were found in 31 isolates. MRFLP analysis resulted in 13 different patterns. Combining the three sets of typing data yielded 24 composite types. Comparison of composite type, isolation date and geographical location of case allowed the retrospective recognition of seven potential clusters during the 5-year study period. Composite types of 24 (80%) of the non-human isolates tested were indistinguishable from human isolates, suggesting that human infection may be via a number of vehicles.

Although not cost-effective for routine use on all salmonella isolates, the methods used in this study are an important adjunct to serotyping for discrimination within an emerging serotype.

We report a foodborne outbreak of Campylobacter jejuni infection in a summer camp. Outbreak-related cases occurred in 79 persons including 3 secondary cases in campers. Campylobacter jejuni was isolated from stool specimens from 16 of 21 patients who submitted a sample; 13 viable isolates were serotyped and all were serotype O[ratio ]33 (somatic O scheme) or HL[ratio ]18 (heat-labile scheme), and biotype III (Lior scheme). This serotype is widely distributed geographically but rarely isolated from humans. Samples of water from the wells supplying the camp were negative for faecal coliforms, and raw milk had not been served in the camp. A matched (1[ratio ]1) case-control study identified tuna salad served for lunch on 19 July as the likely food item associated with illness (matched odds ratio=22; 95% confidence intervals (CI)=3·6–908). Swimming in the camp pool and other recreational water use in area lakes by the campers were not statistically associated with illness. The precise mechanism of introduction of the organism into the tuna salad remains unknown; contamination most likely occurred through cross-contamination with another food product, the hands of a food handler, or a work surface. Several deficiencies in the operation of the camp kitchen were identified. In Wisconsin, kitchens of such camps are subject to different inspection rules than restaurants. Camp staff, administrators, counselors, food managers, and infirmary staff, should fulfil important roles in their respective areas to prevent future outbreaks.

Transmission routes of Campylobacter spp. in broilers and possibilities for prevention of infections were studied on two Dutch broiler farms. The occurrence of Campylobacter spp. was studied in successive broiler flocks, in the environment of the farms and in some of the parent flocks involved. Isolates of Campylobacter spp. were typed by using randomly amplified polymorphic DNA (RAPD) analysis. The results indicate that broiler flocks become infected from environmental sources. The typing results suggest that on one farm transmission of Campylobacter spp. occurred from cattle to broilers via the farmer's footwear. After several campylobacter positive broiler cycles hygiene measures, including thorough cleaning and disinfection procedures, change of footwear at the entrance of each broiler house, control of vermin and other hygienic precautions, were introduced on both farms in order to prevent transmission of Campylobacter spp. from the farm environment to the broilers. The results indicate that the application of hygiene measures significantly reduced campylobacter infections of broiler flocks on both farms.

In recent decades, several epidemiological studies have been published on acute otitis media (AOM), indicating that the occurrence of AOM is increasing. However, the comparison between the surveys is complicated and biased by several factors, e.g. variable study demography and design and dissimilar diagnostic criteria. The present study was performed with an identical set-up in 1978–9 and 1994–5 to find out potential changes in the occurrence of AOM. All the attacks of AOM among children under 10 years diagnosed by a physician during the 12-month periods 1 June, 1978 to 31 May 1979 and 1 June 1994 to 31 May 1995 were registered retrospectively in two Finnish municipalities. The incidence rate (total number of AOM attacks per 100 child years) was 19 (95% CI 18–21) in 1978–9 and 32 (95% CI 30–34) in 1994–5. The increase in the occurrence of AOM was 68% (95% CI 53–79%, P<0·001).

Molecular analyses based on plasmid profile typing and pulsed-field gel electrophoresis have defined a strain of Salmonella enterica serotype Anatum associated with the consumption of a particular brand of formula-dried milk responsible for an outbreak in late 1996/early 1997 involving 15 infants and 2 relatives in the UK, and 2 infants in France. The study has demonstrated the value of laboratory-based surveillance involving identification of the outbreak strain at the molecular level coupled with food microbiology and targeted epidemiological investigations, and has highlighted the importance of rapid communication and subsequent international collaboration through the European Union-funded Salm-Net salmonella surveillance network.

The prevalence of Vibrio cholerae in drinking water, lakes and sewage outfalls during July and August 1996 in Vellore, India was determined. Drinking water samples were collected on single occasions from 12 sites in different geographic areas of the town where cholera had been reported. Samples of water, plankton and sediment were collected from fixed sites at three lakes on three occasions separated by at least 3 days during the course of the study. Samples from open sewers were taken from two representative sites in four areas of the town. Bacteria isolated from samples were identified by standard biochemical tests and isolated strains of V. cholerae tested for their ability to agglutinate O1 and O139 antisera. Water samples from lakes were also tested for the presence of V. cholerae O1 and O139 by fluorescent antibody staining. Non-O1, non-O139 strains of V. cholerae were detected in 41% of drinking water samples and 100% of water, sediment and plankton samples from the test lakes. Eighty-seven per cent of open sewers sampled contained viable non-O1, non-O139 V. cholerae. Fluorescent antibody staining gave positive results for V. cholerae O1 and O139 for all water samples from the three lake sites. Strains of Aeromonas spp. were isolated from 58% of drinking water samples and from 66% of sediment, 77% of plankton and 55% of water samples from lakes. All open sewers sampled contained Aeromonas spp. PCR amplification employing specific primers demonstrated that none of the non-agglutinating V. cholerae isolates contained the ctx operon. The non-O1, non-O139 V. cholerae isolates showed different patterns of antibiotic resistance to ampicillin, ciprofloxacin, chloramphenicol, tetracycline and trimethoprim.