The depopulation of Pacific islands during the 16th to 19th centuries is a striking example of historical mass mortality due to infectious disease. Pacific Island populations have not been subject to such cataclysmic infectious disease mortality since. Here we explore the processes which could have given rise to this shift in infectious disease mortality patterns. We show, using mathematical models, that the population dynamics exhibited by Pacific Island populations are unlikely to be the result of Darwinian evolution. We propose that extreme mortality during first-contact epidemics is a function of epidemiological isolation, not a lack of previous selection. If, as pathogens become established in populations, extreme mortality is rapidly suppressed by herd immunity, Pacific Island population mortality patterns can be explained with no need to invoke genetic change. We discuss the mechanisms by which this could occur, including (i) a link between the proportion of the population transmitting infectious agents and case-fatality rates, and (ii) the course of infection with pathogens such as measles and smallpox being more severe in adults than in children. Overall, we consider the present-day risk of mass mortality from newly emerging infectious diseases is unlikely to be greater on Pacific islands than in other geographical areas.

Measles elimination goals have been adopted in a range of countries, sub-regions, and regions since the WHO declared an elimination goal by 2015 or 2020. All countries attempt to achieve and maintain high coverage through routine immunization programmes. This routine strategy, however, does not ensure the elimination goal of measles. Many developed countries, such as the United States, that have succeeded in interrupting measles transmission earlier, are now experiencing outbreaks with an increasing number of cases. Using a stochastic, age-structured model of measles vaccination dynamics, we explore and characterize the transient dynamics of measles susceptibility in the years following the implementation of routine vaccination at the herd immunity threshold. We demonstrate how a population could face risks of potentially large outbreaks even within few years of vaccination. We characterize different risk profiles depending on the incidence pattern in the years prior to vaccination. These results suggest that the classic critical vaccination threshold is necessary to achieve herd immunity, but not sufficient to prevent long periods of transient, super-critical dynamics. Our results suggest the need of future work for more careful monitoring of the impacts of current immunization programmes, and developing models that take into account more complicated vaccination strategies, demographic factors, and population movements.

Over 300 cases of acute toxoplasmosis are confirmed by reference testing in England and Wales annually. We conducted a case-control study to identify risk factors for Toxoplasma gondii infection to inform prevention strategies. Twenty-eight cases and 27 seronegative controls participated. We compared their food history and environmental exposures using logistic regression to calculate odds ratios (OR) and 95% confidence intervals in a model controlling for age and sex. Univariable analysis showed that the odds of eating beef (OR 10·7, P < 0·001), poultry (OR 6·4, P = 0·01) or lamb/mutton (OR 4·9, P = 0·01) was higher for cases than controls. After adjustment for potential confounders a strong association between beef and infection remained (OR 5·6, P = 0·01). The small sample size was a significant limitation and larger studies are needed to fully investigate potential risk factors. The study findings emphasize the need to ensure food is thoroughly cooked and handled hygienically, especially for those in vulnerable groups.

In this study we determined the presence of IgM/IgG antibodies to Toxoplasma gondii in sera of 155 and 300 pregnant women from Lisbon (Portugal) and Luanda (Angola), respectively, and evaluated the potential risk factors associated with this infection. DNA detection was performed by PCR assays targeting T. gondii regions (RE/B1). Overall, 21·9% (10·9% IgG, 10·9% IgG/IgM) of the Lisbon women and 27·3% (23·7%, IgG, 2% IgM, 1·7% IgG/IgM) of the Luanda women had antibodies to T. gondii. Single variable and binary logistic regression analyses were conducted. Based on the latter, contacts with cats (family/friends), and having more than two births were identified as risk factors for Toxoplasma infection in Lisbon women. In Luanda, the risk factors for T. gondii infection suggested by the single variable analysis (outdoor contact with cats and consumption of pasteurized milk/dairy products) were not confirmed by binary logistic regression. This study shows original data from Angola, and updated data from Portugal in the study of infection by T. gondii in pregnant women, indicating that the prevalence of anti-Toxoplasma antibodies is high enough to alert the government health authorities and implement appropriate measures to control this infection.

The risk of malaria outbreak surfaced in Vanuatu after Tropical Cyclone (TC) Pam in March 2015. In June and July 2015 we conducted malariometric surveys on the islands of Tanna, Aneityum, and Erromango in Tafea Province, where malaria elimination had been targeted, to determine if malaria incidence had increased after TC Pam. No Plasmodium infection was detected by microscopy and PCR in 3009 survey participants. Only 6·3% (190/3007) of participants had fever. Spleen rates in children aged ⩽12 years from Aneityum and Tanna were low, at 3·6% (14/387) and 5·3% (27/510), respectively. Overall bed net use was high at 72·8% (2175/2986); however, a significantly higher (P < 0·001) proportion of participants from Aneityum (85·9%, 796/927) reported net use than those from Tanna (67·1%, 751/1119) and Erromango (66·8%, 628/940). A recent decrease in malaria incidence in Tafea Province through comprehensive intervention measures had reduced the indigenous parasite reservoir and limited the latter's potential to spur an outbreak after TC Pam. The path towards malaria elimination in Tafea Province was not adversely affected by TC Pam.

Dengue presents a wide clinical spectrum of signs and symptoms, with characteristics of the host potentially influencing the disease evolution. Therefore, the purpose of this study was to evaluate the influence of gender and age on dengue clinical outcomes in a recent outbreak situation in Brazil, applying a cross-sectional design and including 6703 dengue cases with laboratory confirmation, occurring in Vitória, Espírito Santo State, Brazil, between 2007 and 2013. Data were obtained from the Information System for Notifiable Diseases. Overall, 11·3% of the sample presented with severe dengue, which affected 13·0% of males, 10·0% of females, 8·8% of children, 12·5% of adolescents, 10·5% of adults and 15·5% of the elderly. Age was higher in the severe dengue group (P = 0·03). Severe dengue was associated with males and the elderly (P < 0·01); however, considering only severe cases, children presented haemorrhage and plasma leakage more frequently than older age groups. The results emphasize the importance of a differentiated protocol for management of dengue cases, taking into consideration host factors like age. These findings also suggest the elderly and children as priority groups for immunization in a future implementation of a vaccine.

A Bayesian Belief Network (BBN) for assessing the potential risk of dengue virus emergence and distribution in Western Australia (WA) is presented and used to identify possible hotspots of dengue outbreaks in summer and winter. The model assesses the probabilities of two kinds of events which must take place before an outbreak can occur: (1) introduction of the virus and mosquito vectors to places where human population densities are high; and (2) vector population growth rates as influenced by climatic factors. The results showed that if either Aedes aegypti or Ae. albopictus were to become established in WA, three centres in the northern part of the State (Kununurra, Fitzroy Crossing, Broome) would be at particular risk of experiencing an outbreak. The model can also be readily extended to predict the risk of introduction of other viruses carried by Aedes mosquitoes, such as yellow fever, chikungunya and Zika viruses.

Dengue fever is an arthropod-borne viral infection that has become endemic in several parts of India including Delhi. We studied occurrence of co-infection with dengue viruses during an outbreak in New Delhi, India in 2014. For the present study, blood samples collected from symptomatic patients were analysed by RT–PCR. Eighty percent of the samples were positive for dengue virus. The result showed that DENV-1 (77%) was the predominant serotype followed by DENV-2 (60%). Concurrent infection with more than one serotype was identified in 43% of the positive samples. Phylogenetic analysis clustered DENV-1 strains with the American African and DENV-2 strains in Cosmopolitan genotypes. Four common amino-acid mutations were identified in the envelope gene of DENV-1 sequences (F337I, A369T, V380I and L402F) and one common mutation (N390S) in the DENV-2 sequences. Further analysis revealed purifying selection in both the serotypes. A significant number of patients were co-infected with DENV-1 and DENV-2 serotypes. Although we do not have direct evidence to demonstrate co-evolution of these two stereotypes, nonetheless their simultaneous occurrence does indicate that they are favoured by evolutionary forces. An ongoing surveillance and careful analysis of future outbreaks will strengthen the concept of co-evolution or otherwise. Whether the concurrent dengue viral infection is correlated with disease severity in a given population is another aspect to be pursued. This study is envisaged to be useful for future reference in the context of overall epidemiology.

Dengue fever (DF) and leptospirosis are serious public problems in tropical regions, especially in Manila, the Philippines. In attempting to understand the causes of DF and leptospirosis seasonality, meteorological factors have been suspected, but quantitative correlation between seasonality and meteorological factors has not been fully investigated. In this study, we investigated correlation of temporal patterns of reported numbers of laboratory-confirmed cases of both DF and leptospirosis with meteorological conditions (temperature, relative humidity, rainfall) in Manila. We used time-series analysis combined with spectral analysis and the least squares method. A 1-year cycle explained underlying variations of DF, leptospirosis and meteorological data. There was a peak of the 1-year cycle in temperature during May, followed by maxima in rainfall, relative humidity and number of laboratory-confirmed DF and leptospirosis cases. This result suggests that DF and leptospirosis epidemics are correlated not only with rainfall but also relative humidity and temperature in the Philippines. Quantifying the correlation of DF and leptospirosis infections with meteorological conditions may prove useful in predicting DF and leptospirosis epidemics, and health services should plan accordingly.

Leptospirosis is the most widespread zoonosis in humans worldwide. In the United States, widespread detection of antibodies to leptospirosis have been identified in feral swine (Sus scrofa) with the highest detection of serovars, Bratislava, Icterohaemorrhagiae, and Pomona. Over the past few years, feral swine populations have expanded their geographical range and distribution in the United States with reports in at least 39 of 50 states. Since feral swine serve as reservoirs for serovars that can infect humans, it is important to understand the risk of transmission. In order to learn more about the probability that feral swine shed infectious leptospires, we collected kidneys and paired serum when possible from 677 feral swine in 124 counties of 29 states. These counties had previously been identified as antibody positive for Leptospira interrogans serovars Bratislava, Canicola, Grippotyphosa, Hardjo, Icterohaemorrhagiae or Pomona. Although exposure to these same six serovars of leptospirosis continued to be high (53% overall) in the counties we sampled, we detected leptospiral DNA in only 3·4% of feral swine kidneys tested. Based on these results, it appears that although feral swine can serve as a source of infection to humans, especially in those who are more likely to encounter them directly such as wildlife biologists, veterinarians, and hunters, the risk may be relatively low. However, further studies to examine the relationship between leptospiral shedding in the urine and kidneys in addition to culturing the organism are recommended in order to better understand the risk associated with feral swine.

Ireland reports the highest incidence of verotoxigenic Escherichia coli (VTEC) infection in Europe. This study investigated potential risk factors for confirmed sporadic and outbreak primary VTEC infections during 2008–2013. Overall, 989 VTEC infections including 521 serogroup O157 and 233 serogroup O26 were geo-referenced to 931 of 18 488 census enumeration areas. The geographical distribution of human population, livestock, unregulated groundwater sources, domestic wastewater treatment systems (DWWTS) and a deprivation index were examined relative to notification of VTEC events in 524 of 6242 rural areas. Multivariate modelling identified three spatially derived variables associated with VTEC notification: private well usage [odds ratio (OR) 6·896, P < 0·001], cattle density (OR 1·002, P < 0·001) and DWWTS density (OR 0·978, P = 0·002). Private well usage (OR 18·727, P < 0·001) and cattle density (OR 1·001, P = 0·007) were both associated with VTEC O157 infection, while DWWTS density (OR 0·987, P = 0·028) was significant within the VTEC O26 model. Findings indicate that VTEC infection in the Republic of Ireland is particularly associated with rural areas, which are associated with a ubiquity of pathogen sources (cattle) and pathways (unregulated groundwater supplies).

Sapoviruses (SaVs) are responsible for sporadic cases and outbreaks of acute gastroenteritis. Despite this, few studies in India have focused on the epidemiological investigation of SaV in cases of acute gastroenteritis. The aim of this study was to understand the molecular epidemiology, genetic diversity and clinical impact of SaV in diarrhoeic children from Pune, Western India. Between 2007 and 2011, a total of 985 faecal samples from diarrhoeic cases and non-diarrhoeic controls were collected and examined for the presence of SaV by nested RT–PCR. SaV was detected in 2·7% (21/778) of the cases and 1·9% (4/207) of the controls. We observed that the majority of SaV mono-infections caused severe gastroenteritis (67%) with clinical manifestations of diarrhoea (100%), vomiting (73%) and dehydration (80%). All known human SaV genogroups were detected in the study. At least eight genotypes were identified from cases and controls. Genogroups GIV and GV, along with genotypes GI.5, GII.4 and GII.6, were discovered for the first time in India. Two GII.4 study strains were found to be 98·5–99% identical, having a novel intra-genogroup recombinant (GII.1/GII.4) recently reported from the Philippines, suggesting probable evidence of recombination. The circulation pattern of SaV genotypes varied during the study period, with GII.1 being predominant in 2007 and 2009, GIV.1 in 2008, and GV.1 in 2011.

Among pathogens shed by cattle, Escherichia coli O157 ranks highest in those causing human illness. To date, prevalence and risk factors for O157 shedding have been assessed in feedlot, but not dairy cattle. The study aimed to determine prevalence levels and risk factors for O157 atypical enteropathogenic E. coli (aEPEC) and enterohaemorrhagic E. coli (EHEC) shedding in dairy cattle. Dairy cattle (n = 899) within the first 21 days of lactation were sampled monthly over the course of 1 year, on three dry lot dairies surrounding Fort Collins, CO. During visits multiple factors were measured (disease history, pharmaceutical use, climate measures, etc.), and cattle faeces were collected and assessed for presence of O157 and virulence genes. Logistic regression analysis was performed using O157 outcomes and measured factors. Prevalence of O157 aEPEC was 3·7%, while EHEC was 3·0%. Many potential risk factors were highly correlated, and used to build separate multivariable models. An increase in humidity was positively associated with aEPEC, while fluid faeces and history of disease showed a negative association. Meanwhile, an increase in temperature and antibiotic treatment was positively associated with EHEC, while more days in milk, higher hygiene score and cow contact were negatively associated. These results may guide mitigation strategies that reduce O157 shedding, and contamination of the human food chain.

This cross-sectional prevalence study investigates meningococcal carriage for the first time in a Southeast Asian population. Posterior pharyngeal swabs were collected between August 2013 and March 2014 from 937 healthy Filipinos aged 5–24 years attending school or university in Manila. Of these, 35 were found to be carriers giving an overall carriage prevalence of 3·7% [95% confidence interval (CI) 2·6–5·2]. Carriage was associated with age (P < 0·001) and was highest (9·0%, 95% CI 5·5–13·8) in subjects aged 10–14 years, but was comparatively low (<3%) in all other age groups considered. This suggests that an immunization programme in the Philippines designed to reduce carriage acquisition and induce herd immunity may require a vaccine dose before the age of 10 years. Serogroup B was most commonly carried (65·7%, 95% CI 47·8–80·9), with a small number of carriers for serogroups C, Y and W also present. Two individuals (5·7%, 95% CI 0·7–19·2) who were simultaneously carrying multiple serogroups were identified. This exploratory study provides valuable insight into the asymptomatic carriage of Neisseria meningitidis in a healthy subset of the Filipino population and illustrates the importance of generating local carriage data.

H9 avian influenza virus played a key role during generation of the novel H7N9 virus. A surveillance programme was conducted to assess the H9 virus in relation to the risk of H7N9 virus contamination in the environment. Risk of H7N9 virus contamination in the presence of H9 virus was higher than without (adjusted odds ratio 4·49, 95% confidence interval 3·79–5·31). Adjusted odds ratios of the H7N9 virus associated with co-presence of H9 virus and interacting factors were 4·93 (rural vs. urban area), 46·80 (live poultry markets vs. other premises), 6·86 (Huzhou vs. Hangzhou prefecture), 40·67 (year 2015 vs. 2013), and 9·63 (sewage from cleaning poultry vs. poultry faeces). Regular surveillance on gene variability of H7N9 and H9 viruses should be conducted and extra measures are needed to reduce co-circulation of H7N9 and H9 viruses in the environment.

We explored age-dependent patterns in haemagglutination inhibition (HI) titre to seasonal [1956 A(H1N1), 1977 A(H1N1), 2007 A(H1N1)] and pandemic [A(H1N1)pdm09] influenza strains using serological data collected from an adult French influenza cohort. Subjects were recruited by their general practitioners from 2008 to 2009 and followed until 2010. We explored age-related differences between strain-specific HI titres using 1053 serological samples collected over the study period from 398 unvaccinated subjects. HI titres against the tested seasonal and pandemic strains were determined using the HI technique. Geometric mean titres (GMTs) were estimated using regression models for interval-censored data. Generalized additive mixed models were fit to log-transformed HI estimates to study the relationship between HI titre and age (age at inclusion and/or age at initial strain circulation). GMT against one strain was consistently highest in the birth cohort exposed to that strain during childhood, with peak titres observed in subjects aged 7–8 years at the time of initial strain circulation. Our results complete previous findings on influenza A(H3N2) strains and identify a strain-dependent relationship between HI titre and age at initial strain circulation.

We compared the rates of fever in adult subjects with laboratory-confirmed influenza and other respiratory viruses and examined the factors that predict fever in adults. Symptom data on 158 healthcare workers (HCWs) with a laboratory-confirmed respiratory virus infection were collected using standardized data collection forms from three separate studies. Overall, the rate of fever in confirmed viral respiratory infections in adult HCWs was 23·4% (37/158). Rates varied by virus: human rhinovirus (25·3%, 19/75), influenza A virus (30%, 3/10), coronavirus (28·6%, 2/7), human metapneumovirus (28·6%, 2/7), respiratory syncytial virus (14·3%, 4/28) and parainfluenza virus (8·3%, 1/12). Smoking [relative risk (RR) 4·65, 95% confidence interval (CI) 1·33–16·25] and co-infection with two or more viruses (RR 4·19, 95% CI 1·21–14·52) were significant predictors of fever. Fever is less common in adults with confirmed viral respiratory infections, including influenza, than described in children. More than 75% of adults with a viral respiratory infection do not have fever, which is an important finding for clinical triage of adult patients with respiratory infections. The accepted definition of ‘influenza-like illness’ includes fever and may be insensitive for surveillance when high case-finding is required. A more sensitive case definition could be used to identify adult cases, particularly in event of an emerging viral infection.

Accurate forecasting of seasonal influenza epidemics is of great concern to healthcare providers in temperate climates, since these epidemics vary substantially in their size, timing and duration from year to year, making it a challenge to deliver timely and proportionate responses. Previous studies have shown that Bayesian estimation techniques can accurately predict when an influenza epidemic will peak many weeks in advance, and we have previously tailored these methods for metropolitan Melbourne (Australia) and Google Flu Trends data. Here we extend these methods to clinical observation and laboratory-confirmation data for Melbourne, on the grounds that these data sources provide more accurate characterizations of influenza activity. We show that from each of these data sources we can accurately predict the timing of the epidemic peak 4–6 weeks in advance. We also show that making simultaneous use of multiple surveillance systems to improve forecast skill remains a fundamental challenge. Disparate systems provide complementary characterizations of disease activity, which may or may not be comparable, and it is unclear how a ‘ground truth’ for evaluating forecasts against these multiple characterizations might be defined. These findings are a significant step towards making optimal use of routine surveillance data for outbreak forecasting.

South Africa's paediatric antiretroviral therapy (ART) programme is managed using a monitoring and evaluation tool known as TIER.Net. This electronic system has several advantages over paper-based systems, allowing profiling of the paediatric ART programme over time. We analysed anonymized TIER.Net data for HIV-infected children aged <15 years who had initiated ART in a rural district of South Africa between 2005 and 2014. We performed Kaplan–Meier survival analysis to assess outcomes over time. Records of 5461 children were available for analysis; 3593 (66%) children were retained in care. Losses from the programme were higher in children initiated on treatment in more recent years (P < 0·0001) and in children aged ≤1 year at treatment initiation (P < 0·0001). For children aged <3 years, abacavir was associated with a significantly higher rate of loss from the programme compared to stavudine (hazard ratio 1·9, P < 0·001). Viral load was suppressed in 48–52% of the cohort, with no significant change over the years (P = 0·398). Analysis of TIER.Net data over time provides enhanced insights into the performance of the paediatric ART programme and highlights interventions to improve programme performance.

In Norway, no published data on seroprevalence of hepatitis E virus (HEV) in humans and swine exists. Serum samples from blood donors, veterinarians, swine farm workers and swine were analysed by ELISA to estimate the seroprevalence of HEV in Norway and to investigate the association between direct contact with swine and HEV seroprevalence in humans. The seroprevalence of HEV IgG antibodies was 30% (24/79) in farm workers, 13% (21/163) in veterinarians, 14% (162/1200) in blood donors and 90% (137/153) in swine. Our results show a high seroprevalence of HEV in humans and swine in Norway. HEV seroprevalence in farm workers and blood donors increased with age, and veterinarians working with swine were twice as likely to be HEV seropositive compared to other veterinarians. High HEV seroprevalence in farm workers and veterinarians working with swine support previous reports suggesting swine as a reservoir for HEV infections in humans in Europe.