During the period 1989–94, 4620 strains of Streptococcus pneumoniae (4063 from blood and 557 from cerebrospinal fluid), from cases of invasive disease in Denmark, were received for capsular typing and penicillin susceptibility testing. During the study period the incidence of bacteraemic pneumococcal disease increased from 10 to 18 cases per 100000 inhabitants per year. The highest rates were seen in the very young, age less than 5 years (23/100000/year, in 1994), and in the elderly, age greater than 60 years (55/100000/year, in 1994). The annual number of cases of meningitis did not vary.

Overall, 92% (93% blood, 87% CSF) of isolates and 94% of all childhood isolates belonged to the 23 vaccine types. The capsular types occurring most commonly among the 4123 pneumococcal strains from adults were types 1, 4, 14, 6A+6B, 7F, 9V, 3, 12F, and 8 (in order of frequency). The ten most frequently occurring types from children (6A+6B, 18C, 14, 1, 7F, 19F, 9V, 4, and 23F) covered 84% of the cases of bacteraemia and meningitis. Reduced susceptibility to penicillin was rare (< 1%).

Restriction fragment length polymorphism of the gene encoding rRNA (ribotyping) was used in combination with conventional epidemiological markers to study phenotypic variations among Streptococcus agalactiae of bovine origin and the possible epidemiological interrelationship between the bovine and human reservoirs of Streptococcus agalactiae.

The bovine material constituted 53 strains (9 antigen combinations) isolated from 11 herds. Herds with a uniform as well as heterogenic antigenic pattern were included. Furthermore, strains isolated in the course of time from the same persistently infected quarters were examined. The human material constituted 16 strains, 4 each of 4 serotypes, isolated from healthy carriers. Finally, nine serotype- and the group reference strains were examined. All strains were serotyped by double diffusion in agarose gel, biotyped (lactose ±), and ribotyped using two restriction enzymes, Hind III and Hha I.

All isolates could be typed by ribotyping and seven ribotypes were identified among the reference strains. The restriction enzymes used alone or in combination gave typing results that allowed discrimination between and within serotype. Combined use of serotype, HindIII and HhaI ribotypes produced 11 types among the 16 human strains. Ribotype analysis discriminated between herds infected with the same serotype. Strains of varying antigenic patterns from the same herd had the same ribotype. Phenotypic variations in serotype observed in persistent intramammary infection were not related to genetic changes as monitored by ribotype. Two ribotypes were represented among both bovine and human strains. The discriminating capability of lactose fermentation was of limited value.

A retrospective study of 191 cases of septic arthritis was undertaken at Royal Darwin Hospital in the tropical north of Australia. Incidence was 9·2 per 100000 overall and 29·1 per 100000 in Aboriginal Australians (RR 6·6; 95% CI 5·0–8·9). Males were affected more than females (RR 1·6; 95% CI 1·2–2·1). There was no previous joint disease or medical illness in 54%. The commonest joints involved were the knee (54%) and hip (13%). Significant age associations were infected hips in those under 15 years and infected knees in those over 45 years. Seventy-two percent of infections were haematogenous. Causative organisms included Staphylococcus aureus (37%), Streptococcus pyogenes (16%) and Neisseria gonorrhoeae (12%). Unusual infections included three melioidosis cases. Polyarthritis occurred in 17%, with N. gonorrhoeae (11/23) more likely to present as polyarthritis than other organisms (22/168) (OR 6·0; 95% CI 2·1–16·7). Univariate and multivariate analysis showed the hip to be at greater risk for S. aureus than other joints. Open arthrotomy was a more successful treatment procedure than arthroscopic washout or needle aspiration.

An outbreak of gastroenteritis occurred in Italy among 39 persons who had attended a private supper. All guests were previously healthy, young, non-pregnant adults; 18 (46%) had symptoms, mostly gastrointestinal (78%), with a short incubation period. Four were hospitalized with acute febrile gastroenteritis, two of whom had blood cultures positive for Listeria monocytogenes. No other microorganisms were recovered from the hospitalized patients' specimens. Epidemiological investigation identified rice salad as the most likely vehicle of the food-borne outbreak. L. monocytogenes was isolated from three leftover foods, the kitchen freezer and blender. Isolates from the patients, the foods and the freezer were indistinguishable: serotype l/2b, same phage type and multilocus enzyme electrophoretic type. Eight (36%) of 22 guests tested were found to have antibodies against L. monocytogenes, compared with none of 11 controls from the general population. This point source outbreak was probably caused by infection with L. monocytogenes. Unusual features included the high attack rate among immunocompetent adults and the predominance of gastrointestinal symptoms.

Using a direct isolation method Listeria spp. were detected in 101 (47·4%) of 213 houses investigated. L. monocytogenes was present in 45 houses (21·1%). Listeria spp. occurred at all sampling sites. Dish-cloths (37%) and surface samples round the drain in the bathroom (27·2%) were most frequently contaminated. Highest numbers (c. 104 c.f.u./object) were found in dish-cloths and washing-up brushes. Lower levels (up to 103 c.f.u./object) were obtained from kitchen sinks, refrigerator vegetable compartment samples and tooth brushes. In total, 132 isolations of Listeria spp. were made from 871 samples. L. innocua (53%) and L. monocytogenes (41%) were the predominant species in the positive samples. Other Listeria spp. were found in only 6% of the positive samples.

The occurrence of epidemics of vaccine-preventable diseases, and the immunization coverage required to prevent them, is affected by the presence of households and heterogeneity in the community. We consider a community where individuals live in households and are of different types, according to infectivity and/or susceptibility to infection. We describe a method for computing the critical immunization coverage to prevent epidemics in such communities and discuss the effectiveness of immunization strategies. In a heterogeneous community where individuals live in households several immunization strategies are possible and we examine strategies targeting households, randomly selected individuals, or groups with highly intense transmission, such as school children. We compare estimates of the critical immunization coverage if we assume that disease is spread solely by random mixing with estimates which result if we assume the effects of the household structure. Estimates made under these two sets of assumptions differ. The results provide insights into the community effects of vaccination, and the household structure of the community should be taken into account when designing immunization policies.

After a nursery school trip to a dairy farm, 20 (53%) of 38 children and 3 (23%) of 13 adult helpers developed gastrointestinal infection. Campylobacter jejuni was isolated from 15 primary cases and from 3 of 9 secondary household cases. A cohort study of the school party found illness to be associated with drinking raw milk (relative risk 5·4, 95 % confidence interval 1·4–20·4, P = 0·001). There was a significant dose response relationship between amount of raw milk consumed and risk of illness (X2-test for linear trend 12·1, P = 0·0005) but not with incubation period, severity of symptoms or duration of illness. All 18 human campylobacter isolates were C. jejuni resistotype 02 and either biotype I (number 16) or biotype II (number 2). Campylobacter was also isolated from samples of dairy cattle and bird faeces obtained at the farm but these were of different resisto/biotypes. Educational farm visits have become increasingly popular in recent years and this outbreak illustrates the hazard of exposure to raw milk in this setting.

The significance of non-culturable forms of Campylobacter spp., especially with regard to the epidemiology of this organism in poultry flocks, was explored. Two different experiments were conducted to produce non-culturable Campylobacter spp. and test their ability to colonize the animal gut. In the first experiment a mixture of 28 different strains of Campylobacter spp. from various sources was inoculated in both sterilized surface water and potassium phosphate buffer and stored at 4 °C. After Campylobacter spp. were no longer detectable by culture in the microcosms, the mixtures of non-culturable cells were used to challenge both chicks and mice. Recovery of non-culturable Campylobacter spp. from the animals was not successful at 4 weeks after administration. In the second experiment the survival of six individual strains of Campylobacter spp. in sterilized surface water at 4 °C was studied and the resulting non-culturable cells were used to challenge chicks. None of the Campylobacter strains could be recovered from the chicks at 2 weeks after administration. We conclude that occurrence of non-culturable forms of Campylobacter spp. capable of colonizing chicks is not a common phenomenon and that non-culturable forms of Campylobacter spp. are likely to be insignificant for importantly to the epidemiology of the organism in Dutch broiler flocks.

A total of 279 Vibrio anguillarum strains were serotyped and examined for plasmid content. Plasmids were subjected to digestion with restriction enzymes. Most strains belonged to serogroup O1 (39%) and O2 (16%). In total 164 strains (53%) carried plasmids. Of the O1 and O2 isolates, 92% and 30%, respectively, carried one or more plasmids. Restriction fragment length polymorphism (RFLP) analysis of plasmid DNA indicated that plasmids belonged to several groups. Each group seemed to be restricted to a single O-serovar. The largest group was the pJM1-like plasmids among most serovar O1 strains. Most of these plasmids were about 67 kb like the pJM1 plasmid, but various derivatives ranged from 26–77 kb. RFLP studies of the 67 kb plasmids revealed 17 different restriction patterns. Some patterns were dominant among European strains whereas others were dominant among North American strains. The results confirmed the applicability of O-serotyping together with plasmid profile and restriction analysis of plasmids for typing of V. anguillarum. They also indicated that plasmids among strains which belonged to the traditional fish pathogenic serogroups, O1 and O2, showed more homology than did strains from most other serogroups, that were usually non-pathogenic, environmental bacteria.

The incidence of acute gastroenteritis and self suspected food poisoning in general practice populations was compared with consultation rates in four group practices during a 3-month winter period and a 2-month autumn period. The average monthly consultation rate for acute gastroenteritis and self suspected food poisoning was 0·3 % and 0·06 % respectively. However, over the same period, on average, an estimated 7% of the practice population per month reported an acute gastroenteritis illness, and 0·7% suspected a food poisoning illness. Only about one in 26 people who suffer an acute episode of gastroenteritis consult their general practitioner (GP). In two practices, faecal samples were sought from all patient cases; the isolation rate for salmonellas was 2% (3/191) and for campylobacters it was 12% (23/191). In the other two practices following routine management, the isolation rate for salmonellas was 9% (6/64) and for campylobacters it was 2% (1/64). Isolation of faecal pathogens was not associated with patients' suspicion of food poisoning. A history of eating out in the week before onset was associated with a significantly increased yield of salmonellas and campylobacters.

In 1994 a Texas prison containing a population of mentally retarded inmates experienced a large tuberculosis outbreak. Fifteen cases of tuberculosis were identified (8 confirmed by positive cultures for Mycobacterium tuberculosis) and more than 100 inmates became infected. The culture-confirmed patients were infected with an identical strain of tuberculosis as demonstrated by polymerase chain reaction (PCR) based DNA fingerprinting technique. The prison followed standard tuberculosis infection control policies, but these controls were inadequate to prevent tuberculosis transmission in this special population. Two hundred and thirty inmates (119 inmates showing evidence of new tuberculosis infection or active disease and 111 healthy controls) were enrolled in the investigation. Inmate cell assignments, job duties, and educational classes were identified and medical chart reviews were conducted on all inmates. Tuberculosis transmission was associated with residing on the D Wing of the prison (OR = 25·84, P < 0·01), attending school in Classroom A (OR = 8·34, P = 0·01) and working on the prison utility work crew (OR = 2·52, P < 0·01). The index case in the outbreak had been prescribed 6 months of isoniazid (INH) chemoprophylaxis in 1988.

There was a marked rise in scarlet fever mortality in England and Wales in the mid-nineteenth century and spectral analysis of the registration details, 1847–80, shows that the interepidemic interval was 5–6 years, but after 1880 the endemic level fell and the fatal epidemics disappeared. The dynamics of the scarlet fever epidemics can be represented by a linearized mathematical model and because the system is lightly damped, it could be driven by an oscillation in susceptibility. Epidemics were significantly correlated with dry conditions in spring/summer (P < 0·001), suggesting that these produced a low amplitude oscillation in susceptibility which drove the system. Epidemics also correlated (P < 0·001) with an oscillation in wheat prices but at a lag of 3 years, suggesting that malnutrition during pregnancy caused increased susceptibility in the subsequent children which interacted synergistically with seasonal dry conditions. Scarlet fever mortality was sharply reduced after 1880 in parallel with falling wheat prices suggesting that the remarkable period of high scarlet fever mortality (1840–80) was dependent on poor nutritive levels during that time.

The macrophage infectivity potentiator gene (mip) from Legionella pneumophila is a major virulence factor of the species. Thus, mip-detection by amplification has been proposed to assess the presence of L. pneumophila in clinical and environmental samples. The distribution of mip-related sequences within the Legionellaceae was studied by DNA amplification using mip-specific primers followed by Southern blot hybridization with an internal probe. Thirty nine species (48 serogroups) of Legionellaceae were screened in this attempt. Using this approach, sequences related to mip were observed in 89% of the tested species including the most recently described L. fairfieldensis, L. lansingensis and L. shakespearei. In several cases, cloning and sequencing of the amplified products confirmed the high levels of similarity between the sequence found in non-pneumophila species with that of the L. pneumophila mip gene. This confirms previous reports that mip related genes are widespread among Legionellaceae and therefore specific detection of the species L. pneumophila cannot be based on mip-targeted amplification.

During the autumn of 1992, Western Australia experienced a large viral meningitis outbreak of dual aetiology. Of the 161 cases, 64% were children under 15 years of age, with the highest notification rate being in children less than 5 years of age. Echovirus 9 caused 41% of cases and occurred mainly in the metropolitan areas of Western Australia whereas echovirus 6, which caused 37% of cases, was more widespread. An enterovirus was cultured from 70% of CSF specimens, 88% of faecal specimens and 68% of upper respiratory tract specimens. High CSF white cell counts and neutrophil predominance were common. Seven cases had normal CSF white cell counts even though an enterovirus was isolated from the CSF. Therefore, the CSF findings were of restricted value in excluding viral meningitis, and did not reliably distinguish between bacterial and viral meningitis.

To determine whether recently diagnosed adult–onset asthma is associated with serologic evidence of chronic Chlamydia pneumoniae infection, we performed a case-control study in a primary care clinic of cases with asthma (25 adults reporting first symptoms of asthma within 2 years of enrollment) and 45 concurrently enrolled sex and age (±10 years) matched non-asthmatic controls with normal pulmonary function. C. pneumoniae-specific IgA, IgG and IgG4 antibodies, and circulating immune complexes (CIC) were measured by microimmunofluorescence testing. Results showed that frequencies of IgG litres > 16 (92%), IgG4 litres > 16 (20%) and CIC > 4 (60%) in asthma patients were not significantly different from those of controls. However, asthmatics had a significantly higher prevalence of C. pneumoniae-specific IgA litres > 10 (72% of cases vs 44% of controls, P < 0·05). After adjustment for the effects of age, sex and smoking, the odds ratio for an association of IgA and asthma was 3·7 (95% confidence interval 1·2–11·5). We conclude that recently symptomalic reversible airway obstruction in adults is associated with the presence of C. pneumoniae-specific IgA antibodies, a proposed indicator of chronic respiratory C. pneumoniae infection.

Maternal, cord and infant measles antibody levels were measured and compared in a group of 411 vaccinated mothers and 240 unvaccinated mothers, and their babies, between 1983 and 1991. Maternal and cord sera were tested by haemagglutination inhibition and/or enzyme-linked immunosorbent assay, and plaque reduction neutralization tests were also used to test infant sera. Geometric mean litres were significantly higher in the unvaccinated than in the vaccinated mothers (P < 0·001). Infants born to mothers with a history of measles had higher antibody levels at birth than infants of vaccinated mothers and, although the difference narrowed over time, continued to have higher levels up to 30 weeks of age. Between 5 and 7 months of age significantly more of the children of vaccinated mothers had plaque reduction neutralization antibody levels below that which would interfere with vaccination. As the boosting effect of circulating natural measles disappears, earlier measles vaccination may need to be considered, perhaps as part of a two-dose policy.

Community studies in Guinea-Bissau have found that exposure to measles prior to 6 months of age is associated with delayed mortality later in childhood. In an attempt to understand the underlying mechanism, we examined the role of pre-exposure nutritional status and the impact of exposure to measles on growth and subsequent mortality in these outbreaks. Though exposed children were lighter than controls, there was no association between pre-exposure weight-for-age and subsequent mortality adjusting for age. Although exposure was strongly associated with excess mortality, it did not have a negative impact on growth. Adjustment for state of nutrition did not alter the mortality ratio (MR) between 6 and 59 months of age for exposed children and controls; exposed children examined anthropometrically between 6–17 months had a MR of 3·70 compared with controls. This trend was the same for anthropometric measurements obtained at 18–59 months of age. Among the controls, there was a significant association between weight-for-age and subsequent mortality to the age of 5 years. However, for exposed children there was no association; the relation between weight-for-age and subsequent mortality was significantly different for exposed children compared with controls (tests for interaction between exposure and anthropometric measurements at 6–17 months: P = 0·05). Growth faltering as a consequence of early exposure to measles does not explain the marked excess mortality among these children. Further studies of the process underlying delayed mortality after early exposure to measles are warranted.

The genotype of HCV was determined in 161 chronic HCV-infected patients. The patients were classified into three groups on the basis of the origin of the HCV infection: 50 patients had a history of intravenous drug use (IVDU) but no HIV infection; 41 patients had received blood transfusions, and 70 patients had no known exposure. The distribution of HCV genotypes was associated with the origin of infection and age of patients: genotype 1b was predominant among patients who had received blood transfusions and those without evidence of parenteral exposure (84·6% and 67·7%, respectively), whereas genotype 3a was present in 65·3% of IVDUs. Patients with genotype 1b were older than those with genotypes la or 3a: 50·3 ± 12 vs. 34·1±9·9 and 31 ± 5.4 years, respectively. These findings suggest that the pattern of HCV genotypes in our region is changing and that genotype 1b may be substituted by 3a as the dominant genotype in the future.

Progress in the development of schistosomiasis models for use in control programmes is limited by the considerable uncertainty in many of the biological parameters. In this paper, this problem is addressed by a comprehensive sensitivity analysis of a schistosomiasis model using the Latin Hypercube method. Fifty simulations with different parameter contributions are run for 50 years with treatment during the first 20 years and reinfection thereafter. The analysis shows only a relatively small divergence between simulations during the chemotherapy treatment programme but considerable divergence in reinfection levels after treatment is stopped. A skewed distribution of outcomes was seen with most simulations showing effective control and a few where control had less impact. The most important uncertainty source was due to the unknown levels of acquired immunity and also uncertainty in the true worm burden. In particular, the strength of the immune response was most important in determining whether control was effective with higher immunity leading to less effective control. Among those simulations in which control was not very effective, those in which the mean worm burden was high showed the least effective control. Since both these are areas of genuine uncertainty, it is proposed that uncertainty analysis should be an integral part of any projection of control programmes.

Hospital discharge data from 1990–4 for 26 hospitals were used to estimate and compare the cost and describe the epidemiology of severe giardiasis and cryptosporidiosis in Scotland, UK. The incidence of severe cryptosporidiosis was almost double that of severe giardiasis and the median duration of hospitalization was longer for cryptosporidiosis than giardiasis. Impaired immunity was frequently listed as co-diagnosis with cryptosporidiosis and associated with extended hospitalization. Although both infections were associated with infants, the median age was lower for cryptosporidiosis (5 years compared with 30 years). Whereas hospitalization was not significantly longer for infants with cryptosporidiosis, hospitalization for this age group with giardiasis was longer (4 days compared with 3 days). Comparison with similar data for giardiasis from USA revealed various differences and similarities. These points are discussed in relation to the epidemiology of these infections and published data were used to estimate costs of hospitalization.