Epidemiology and Infection became 100 years old in 2001. Its powers and influence however do not appear to have diminished, and, as the Senior Editor appointed in its 101st year, I am greatly honoured to be so closely associated with it. Just over 100 years after it was founded, Epidemiology and Infection is ‘going electronic’. To mark both occasions we have put together a small selection of some of the most important or thought-provoking papers published in our first century, and have invited eminent epidemiologists and microbiologists to assess and comment on them critically in the light of modern knowledge. Naturally, to mark our coming of age in terms of modern methods of communication also, all the papers chosen for our centenary edition will be available – free – online. I am most grateful to my predecessor, Professor Bo Drasar, for putting together this unique tribute to Epidemiology and Infection, its Editors, its Editorial Board, and, not least the authors of our papers. Enjoy and marvel at our predecessors' wisdom, perspicacity and scientific rigour.

A recent European Union Directive required member states to put monitoring and control programmes in place, of which vaccination is a central component. Live Salmonella vaccines generally confer better protection than killed vaccines, because the former stimulate both cell-mediated and humoral immunity. Administering Salmonella bacteria orally to newly hatched chickens results in extensive gut colonization and a strong adaptive immune stimulus but broiler chickens are immunologically immature. However, colonization exerts a variety of rapid (within 24 h) protective effects. These include specific colonization-inhibition (competitive exclusion) in which the protective bacteria exert a profound resistance to establishment and colonization by other related bacteria. This is thought to be primarily a metabolic attribute of the vaccinating bacteria but may also involve competition for attachment sites. The presence of large numbers of bacteria originating from a live Salmonella vaccine in the intestine can also induce infiltration of polymorphonuclear cells into the intestinal wall, which confers resistance to invasion and systemic spread by virulent Salmonella strains. This opens new perspectives for vaccine usage in broilers, layers and breeding poultry but also in other animals which show increased susceptibility to infection because of their young age or for other reasons, such as oral chemoprophylaxis or chemotherapy, where the lack of established normal gut flora is an issue. We recommend that all live vaccines considered for oral administration should be tested for their ability to induce the two protective effects described above. Further developments in live Salmonella vaccines are, however, currently hindered by fears associated with the use and release of live vaccines which may be genetically modified.

This study documents the changes in mortality in England and Wales over the last 100 years as a possible explanation for our increasingly allergy-prone Western society. A total of 53 million computerized recorded deaths, which occurred from 1901 to 2000 were analysed retrospectively. Childhood mortality decreased by 98%, from 40·6% of total annual deaths in 1901 to 0·9% in 2000. In 1901, 36·2% of all deaths and 51·5% of childhood deaths were from infectious diseases. By contrast in 2000, 11·6% of all deaths and only 7·4% of childhood deaths were from infectious diseases. Infectious diseases were a significant cause of childhood mortality in British cities until about 40 years ago. Several factors, including vaccination, antibiotics and improved sanitation have contributed to this trend. Survival of individuals with heightened immunity to infections may have led to natural selection of allergy-prone individuals in England and Wales. However, the relationship between changes in rates of infection and allergy is complex and not fully understood.

We aimed to describe the incidence of new episodes of molluscum contagiosum, scabies and lichen planus presenting to general practitioners in England and Wales. We examined data collected in a sentinel practice network (the Weekly Returns Service of the Royal College of General Practitioners) in which about half a million persons were observed each year over the period 1994–2003. The incidence of molluscum contagiosum in males was 243/100000 person-years and in females 231; of scabies, males 351, females 437; of lichen planus, males 32, females 37. Incidence varied by year and age. Ninety per cent of molluscum contagiosum episodes were reported in children aged 0–14 years, where incidence in 2000 (midpoint of a 6-year period of stable incidence) was 1265/100000 (95% CI 1240–1290). Scabies affected all ages and annual incidence ranged between 233 (95% CI 220–246) in 2003 and 470 (95% CI 452–488) in 2000. Lichen planus occurred chiefly in persons aged over 45 years: incidence (all ages) ranged between 27 (95% CI 23–31) in 2003 and 43 (95% CI 37–49) in 1998. The relative risk of female to male incidence (all ages) of molluscum contagiosum was 0·95 (95% CI 0·91–0·99); of scabies 1·25 (95% CI 1·21–1·28); and of lichen planus 1·19 (95% CI 1·08–1·13).

Statistical power in group-randomized vaccine trials is complex: group randomization may increase power by capturing more transmission effects but may decrease power as more individuals are affected by a common source of variance. The former effect dominates when most infections arise from within the group and the latter when most arise outside. This process is complicated further when individuals possess partial natural immunity. Vaccine trials typically compare infection frequency (as measured by agent or antibody detection) in vaccinated vs. unvaccinated groups. To assess the relative contributions to statistical power by direct agent detection vs. serological detection of recent infection, we constructed stochastic compartmental models using differential equations describing all possible discrete states of the group. We contrasted models where natural immunity was absent, altered only the susceptible state, or altered both the susceptible and infected states. We observed the effects of endemic infection levels, the fraction of infection arising from outside the group, infectiousness vs. susceptibility vaccine effects and waning rates. There was significant enhancement of statistical power when serological testing separated infected and susceptible classes into subsets by natural immunity status.

Mathematical modelling is an established tool for planning and monitoring vaccination programmes. However, the matrices describing contact rates are based on subjective choices, which have a large impact on results. This paper reviews published models and obtains prior model probabilities based on publication frequency and expert opinion. Using serological survey data on rubella and mumps, Bayesian methods of model choice are applied to select the most plausible models. Estimates of the basic reproduction number R0 are derived, taking into account model uncertainty and individual heterogeneity in contact rates. Twenty-two models are documented, for which publication frequency and expert opinion are negatively correlated. Using the expert prior with individual heterogeneity, R0=6·1 [95% credible region (CR) 4·3–9·2] for rubella and R0=19·3 (95% CR 4·0–31·5) for mumps. The posterior modes are insensitive to the prior for rubella but not for mumps. Overall, assortative models with individual heterogeneity are recommended.

This paper uses a graph-theoretical approach to investigate the properties of the observed network of disease transmission in the 2001 foot-and-mouth epidemic in the United Kingdom. This analysis revealed both global and local heterogeneity in the contact pattern between the infected premises in the first 3 weeks of the disease. In particular, the global heterogeneity contributed to the failure of the culling strategy imposed by the UK government. However, a more effective strategy targeting selective deletion of key premises in the network was not available once the epidemic had begun. We recommend that post-hoc analyses of this sort should become part of preventative and proactive policy rather than part of a reaction to an ongoing crisis.

In Scotland, between 1995 and 2000 there were between 4 and 10 cases of illness per 100000 population per year identified as being caused by Escherichia coli O157, whereas in England and Wales there were between 1 and 2 cases per 100000 population per year. Within Scotland there is significant regional variation. A cluster of high rate areas was identified in the Northeast of Scotland and a cluster of low rate areas in central-west Scotland. Temporal trends follow a seasonal pattern whilst spatial effects appeared to be distant rather than local. The best-fit model identified a significant spatial trend with case rate increasing from West to East, and from South to North. No statistically significant spatial interaction term was found. In the models fitted, the cattle population density, the human population density, and the number of cattle per person were variously significant. The findings suggest that rural/urban exposures are important in sporadic infections.

Two elementary schools were served lunches that were cooked in the same kitchen. An outbreak of Escherichia coli O157:H7 occurred at one school where the dishes that were prepared for the school were lukewarm and kept for 33 min at an average temperature of 45 °C before serving. However, no outbreak occurred at the other school where dishes were hot and were kept for 60 min at an average temperature of 50 °C before serving. In a series of experiments on the survival of E. coli O157:H7 in the liquid portion of similarly prepared food, the population of E. coli O157:H7 was reduced by 10−3 by heating at 50 °C for 60 min and by only 10−1 by heating at 45 °C for 40 min. Further, E. coli O157:H7 survived at 45 °C for 40 min but not at 50 °C for 60 min at pH 4·0 with a 4·0% salt concentration that was similar to that of the liquid part of the food. These results indicate that pH and salt concentration of cooked food markedly affect the survival of E. coli O157:H7 and help to explain the occurrence of the disease outbreak at only one of the schools.

A survey of the microbial quality of table eggs sold in Trinidad was conducted. For 23 poultry layer farms each visited twice approximately 1 month apart, 25 pooled eggs constituted a composite sample, for 14 shopping malls each visited twice approximately 1 month apart, six pooled eggs made a composite sample and for a total of 102 other retailers across the country each visited once over a 4-month period, six pooled eggs constituted a composite sample. Swabs of egg shells and egg content were tested for selected bacteria. Twenty-four (13·0%), 68 (37·0%), and two (1·1%) of a total of 184 composite eggs (shells, egg content or both) sampled were positive for Salmonella, Escherichia coli, and Campylobacter respectively. All 184 samples tested were negative for Listeria spp. Salmonella was recovered from seven (3·8%) egg shell samples only compared with 14 (7·6%) egg content samples only positive for the pathogen. Fifty-two (28·3%) egg shell samples and seven (3·8%) egg content samples were positive for E. coli. Both isolates of Campylobacter coli originated from egg contents. Of a total of 24 composite egg samples positive for Salmonella, eight different serotypes of Salmonella were isolated from a total of 24 Salmonella-positive composite eggs of which S. Enteritidis was the most prevalent, 58·3% (14/24). Salmonella Georgia was isolated for the first time in Trinidad. Failure to properly handle or heat table eggs sold in Trinidad poses a potential health hazard to consumers because of their poor microbial quality.

We sought to determine the source of a norovirus outbreak among attendees of 46 weddings taking place during a single weekend. Norovirus-compatible illness was experienced by 332 (39%) of wedding guests surveyed; the outbreak affected up to 2700 persons. Illness was associated with eating wedding cake provided by a bakery common to the weddings (adjusted RR 4·5, P<0·001). A cake requiring direct hand contact during its preparation accounted for the majority of illness. At least two bakery employees experienced norovirus-compatible illness during the week preceding the weddings. Identical sequence types of norovirus were detected in stool specimens submitted by two wedding guests, a wedding hall employee, and one of the ill bakery employees. It is likely that one or more food workers at the bakery contaminated the wedding cakes through direct and indirect contact. These findings reinforce the necessity of proper food-handling practices and of policies that discourage food handlers from working while ill.

In November 2002, the first of three outbreaks of Salmonella Montevideo infection in Australia and New Zealand was identified in New South Wales, Australia. Affected persons were interviewed, and epidemiologically linked retail outlets inspected. Imported tahini was rapidly identified as the source of infection. The contaminated tahini was recalled and international alerts posted. A second outbreak was identified in Australia in June–July 2003 and another in New Zealand in August 2003. In a total of 68 S. Montevideo infections, 66 cases were contacted. Fifty-four (82%) reported consumption of sesame seed-based foods. Laboratory analyses demonstrated closely related PFGE patterns in the S. Montevideo isolates from human cases and sesame-based foods imported from two countries. On the basis of our investigations sesame-based products were sampled in other jurisdictions and three products in Canada and one in the United Kingdom were positive for Salmonella spp., demonstrating the value of international alerts when food products have a wide distribution and a long shelf life. A review of the controls for Salmonella spp. during the production of sesame-based products is recommended.

The annual incidence of typhoid fever in Taiwan was 2·1–3·6 cases per 1000000 population from 1995 to 2002. More than 80% of 45 patients with typhoid fever treated at National Taiwan University Hospital from 1996 to 2002 had elevated serum aminotransferase levels at presentation. Ten of these patients were treated during an outbreak in Taipei County in 2002, and seven of them who did not have pre-existing liver disease developed hepatitis, which was unrelated to other aetiologies. All Salmonella typhi isolates were susceptible to extended-spectrum cephalosporins and fluoroquinolones. Multidrug resistance (intermediate resistance to ampicillin, trimethoprim–sulphamethoxazole, and chloramphenicol) was found in one (2·5%) of the 40 isolates studied. Pulsed-field gel electrophoresis analysis demonstrated a high genetic diversity among S. typhi isolates and identified a novel clone associated with the 2002 outbreak. Physicians should be alert to the possibility of typhoid fever when patients, without other gastrointestinal symptoms, present with sustained fever and hepatitis.

The primary aim of this study was to detect and genotype Campylobactercoli strains from humans over a period of one year from November 2002 to October 2003. Campylobacter spp. were isolated from patients with symptoms of enteric infection. Amplified fragment length polymorphism (AFLP) analysis was used to identify the genetic diversity of C. coli strains by cluster analysis. A total of 18·6% of all Campylobacter isolates were identified as C. coli. These data show, that C. coli is indeed of importance to human campylobacter infections. Heterogeneous patterns were detectable among the human C. coli pool by AFLP analysis suggesting different sources of infection. A continuous seasonal shift of genotypes was detectable.

The incidence, phenotypic characteristics and antimicrobial resistance patterns of 193 Shigella strains isolated from 2489 hospitalized children with acute diarrhoea were studied during January 2001 to August 2004. S. flexneri (60%) was the most prevalent serogroup, followed by S. sonnei (23·8%), S. dysenteriae (9·8%) and S. boydii (5·7%). Since 2002, S. flexneri 2a was the most dominant serotype. Almost all S. flexneri strains exhibited resistance to ampicillin, co-trimoxazole, tetracycline, nalidixic acid and fluoroquinolones. After a lapse of almost 14 years, S. dysenteriae type 1 strains reemerged for the first time during 2002 and these strains were resistant to more than two antibiotics (multidrug resistance), including fluoroquinolones. An upsurge of similar resistance patterns was also noted among S. flexneri type 2a since December 2003. Resistance to fluoroquinolone increased year on year among S. dysenteriae type 1 and S. flexneri, but not in S. boydii or S. sonnei. Monitoring of antimicrobial susceptibility through a surveillance programme is recommended to select appropriate antibiotics for the effective treatment of shigellosis in this region.

The purposes of the study were (i) to describe the abundance and epidemiology of Legionellaceae in the man-made environment in a northern Italian area, (ii) to assess the concordance between pulsed-field gel electrophoresis (PFGE) and automated ribotyping (AR) techniques for genotyping L. pneumophila and (iii) to investigate the correlation between serogrouping and genotyping data. Water was sampled from reservoirs in 12 buildings across an area of 80-km radius. Despite the water temperature always being maintained above 55 °C, all of the buildings sampled were contaminated with Legionellaceae on at least one occasion and 63 L. pneumophila isolates representing nine different serogroups were collected. The two DNA methods revealed a high degree of genetic heterogeneity, even though identical L. pneumophila clones were recovered at different sites. The AR technique provided a fairly reliable approximation of PFGE results (73% concordance), however there was poor correlation between serogrouping and genotyping data as identical DNA fingerprints were shared by isolates of different serogroups.

A retrospective study of group A streptococcal (GAS) infections was performed for the period 1985–2002 in an area of central Italy. Although very severe diseases such as streptococcal toxic shock syndrome (STSS) were observed, a general increase in invasive infections was not found. Isolates of GAS were classified by M protein genotyping (emm typing) and analysed according to their origin from invasive and non-invasive infections. The predominant emm types were types 1, 4 and 12, followed by types 3, 6 and 28. During the study period the proportion of isolates of types 1 and 12 fell, while other types (3, 6, 22, 28 and 77) appeared. Isolates from invasive and non-invasive infections shared several emm types; however, most invasive strains belonged to five types only (types 1, 4, 12, 28 and 77), while non-invasive isolates were generally more heterogeneous.

There are few studies from developing countries on the epidemiology of Haemophilus influenzae (Hib) infections among infants and children. We set out to determine the prevalence of oropharyngeal Hib colonization among Turkish children younger than two years of age and to identify antimicrobial resistance among the isolates. A cross-sectional study was conducted on 818 healthy children and oropharyngeal secretions were sampled. The carriage rate of Hib was found to be 7·2% and this increased significantly with age. Carriage of Hib among 3- to 6-month-old children (3·5%) was higher than expected and was significantly higher among children who were passive smokers (P=0·04). Logistic regression analysis showed that breastfeeding status was the sole significant factor for colonization (OR 2·2, 95% CI 1·26–3·82). Antimicrobial susceptibility tests on 56 isolates of H. influenzae showed that 51·8% and 21·4% were resistant to trimethoprim–sulphamethoxazole and ampicillin respectively. Other notable resistances were to cefalexin (10·7%) and chloramphenicol (3·6%); no isolates were resistant to ceftriaxone.

Between March and July 2003, 671 cases of severe acute respiratory syndrome (SARS) were diagnosed in Taiwan with a total of 84 fatalities. After the epidemic, a serological survey was conducted involving the asymptomatic household contacts. Household contacts of 13 index patients were enrolled in the study. Contact history and clinical symptoms of the household contacts were recorded by standardized questionnaires. Blood samples of patients and household contacts were collected at least 28 days after symptom onset in the index patients or household exposure in the contacts for SARS-associated coronavirus (SARS-CoV) IgG testing. On the basis of this investigation, 29 persons (25 adults and 4 children) were identified as having had unprotected exposure to the index cases before infection-control practices were implemented. Laboratory evaluation of clinical specimens showed no evidence of transmission of SARS-CoV infection to any contacts. This investigation demonstrated that subclinical transmission among household contacts was low in the described setting.

A summer camp was followed by an outbreak of illness involving around 90 children. Investigations included individual questionnaires, inspection of the camp facilities, and laboratory analysis of water and clinical samples. Contamination of drinking and swimming water was demonstrated. An enterovirus was detected by polymerase chain reaction (PCR) and/or culture in 4/4 cerebrospinal fluid samples, 9/15 (60%) stool samples from symptomatic children and 2/9 (22%) stool samples from asymptomatic children. The virus was identified as an echovirus 3 by sequencing and phylogenetic analysis of a short 5′ non-coding region (NCR) PCR product. Viruses from the outbreak clustered closely and an echovirus 3 from a temporally associated non-outbreak case could be readily distinguished. Despite the lack of a standardized approach, direct molecular detection and identification of enteroviruses is an efficient epidemiological tool. Here the 5′-NCR was successfully used for both detection and ‘serotyping’, and the close genetic relatedness of isolates was proven.