House-to-house human movement drives dengue virus transmission

Differences in Prevalence of Symptomatic Zika Virus Infection, by Age and Sex-Puerto Rico, 2016

Background

During the outbreak of Zika virus (ZIKV) disease in Puerto Rico in 2016, nonpregnant women aged 20-39 years were disproportionately identified with ZIKV disease. We used household-based cluster investigations to determine whether this disparity was associated with age- or sex-dependent differences in the rate of ZIKV infection or reported symptoms.

Methods

Participation was offered to residents of households within a 100-m radius of the residences of a convenience sample of 19 laboratory-confirmed ZIKV disease cases. Participants answered a questionnaire and provided specimens for diagnostic testing by reverse transcription-polymerase chain reaction (RT-PCR) and enzyme-linked immunosorbent assay (ELISA).

Results

Among 367 study participants, 114 (31.1%) were laboratory positive for ZIKV infection, of whom 30% reported a recent illness (defined as self-reported rash or arthralgia) attributable to ZIKV infection. Age and sex were not associated with ZIKV infection. Female sex (adjusted prevalence ratio [aPR], 2.28; 95% confidence interval [CI], 1.40, 3.67), age <40 years (aPR, 2.39; 95% CI, 1.55, 3.70), and asthma (aPR, 1.63; 95% CI, 1.12, 2.37) were independently associated with symptomatic infection.

Conclusions

Although neither female sex nor age were associated with an increased prevalence of ZIKV infection, both were associated with symptomatic infection. Further investigation to identify a potential mechanism of age- and sex-dependent differences in reporting symptomatic ZIKV infection is warranted.

Impact of population displacement and forced movements on the transmission and outbreaks of Aedes-borne viral diseases: Dengue as a model

Population displacement and other forced movement patterns following natural disasters, armed conflicts or due to socioeconomic reasons contribute to the global emergence of Aedes-borne viral disease epidemics. In particular, dengue epidemiology is critically affected by situations of displacement and forced movement patterns, particularly within and across borders. In this respect, waves of human movements have been a major driver for the changing epidemiology and outbreaks of the disease on local, regional and global scales. Both emerging dengue autochthonous transmission and outbreaks in countries known to be non-endemic and co-circulation and hyperendemicity with multiple dengue virus serotypes have led to the emergence of severe disease forms such as dengue hemorrhagic fever and dengue shock syndrome. This paper reviews the emergence of dengue outbreaks driven by population displacement and forced movements following natural disasters and conflicts within the context of regional and sub-regional groupings.

Spatial and temporal patterns of dengue incidence in northeastern Thailand 2006-2016

Background

Dengue, a viral disease transmitted by Aedes mosquitoes, is an important public health concern throughout Thailand. Climate variables are potential predictors of dengue transmission. Associations between climate variables and dengue have usually been performed on large-scale first-level national administrative divisions, i.e. provinces. Here we analyze data on a finer spatial resolution in one province, which is often more relevant for effective disease control design. The objective of this study was to investigate the effect of seasonal variations, monthly climate variability, and to identify local clusters of symptomatic disease at the sub-district level based on reported dengue cases.

Methods

Data on dengue cases were retrieved from the national communicable disease surveillance system in Thailand. Between 2006 and 2016, 15,167 cases were recorded in 199 sub-districts of Khon Kaen Province, northeastern Thailand. Descriptive analyses included demographic characteristics and temporal patterns of disease and climate variables. The association between monthly disease incidence and climate variations was analyzed at the sub-district level using Bayesian Poisson spatial regression. A hotspot analysis was used to assess the spatial patterns (clustered/dispersed/random) of dengue incidence.

Results

Dengue was predominant in the 5–14 year-old age group (51.1%). However, over time, dengue incidence in the older age groups (> 15 years) gradually increased and was the most affected group in 2013. Dengue outbreaks coincide with the rainy season. In the spatial regression model, maximum temperature was associated with higher incidence. The hotspot analysis showed clustering of cases around the urbanized area of Khon Kaen city and in rural areas in the southwestern portion of the province.

Conclusions

There was an increase in the number of reported dengue cases in older age groups over the study period. Dengue incidence was highly seasonal and positively associated with maximum ambient temperature. However, climatic variables did not explain all the spatial variation of dengue in the province. Further analyses are needed to clarify the detailed effects of urbanization and other potential environmental risk factors. These results provide useful information for ongoing prediction modeling and developing of dengue early warning systems to guide vector control operations.

Electronic supplementary material

The online version of this article (10.1186/s12879-019-4379-3) contains supplementary material, which is available to authorized users.

What Is the Value of Different Zika Vaccination Strategies to Prevent and Mitigate Zika Outbreaks?

BACKGROUND

While the 2015-2016 Zika epidemics prompted accelerated vaccine development, decision makers need to know the potential economic value of vaccination strategies.

METHODS

We developed models of Honduras, Brazil, and Puerto Rico, simulated targeting different populations for Zika vaccination (women of childbearing age, school-aged children, young adults, and everyone) and then introduced various Zika outbreaks. Sensitivity analyses varied vaccine characteristics.

RESULTS

With a 2% attack rate ($5 vaccination), compared to no vaccination, vaccinating women of childbearing age cost $314-$1664 per case averted ($790-$4221/disability-adjusted life-year [DALY] averted) in Honduras, and saved $847-$1644/case averted in Brazil, and $3648-$4177/case averted in Puerto Rico, varying with vaccination coverage and efficacy (societal perspective). Vaccinating school-aged children cost $718-$1849/case averted (≤$5002/DALY averted) in Honduras, saved $819-$1609/case averted in Brazil, and saved $3823-$4360/case averted in Puerto Rico. Vaccinating young adults cost $310-$1666/case averted ($731-$4017/DALY averted) in Honduras, saved $953-$1703/case averted in Brazil, and saved $3857-$4372/case averted in Puerto Rico. Vaccinating everyone averted more cases but cost more, decreasing cost savings per case averted. Vaccination resulted in more cost savings and better outcomes at higher attack rates.

CONCLUSIONS

When considering transmission, while vaccinating everyone naturally averted the most cases, specifically targeting women of childbearing age or young adults was the most cost-effective.

Strategies of host resistance to pathogens in spatially structured populations: An agent-based evaluation

There is growing theoretical evidence that spatial structure can affect the ecological and evolutionary outcomes of host-parasite interactions. Locally restricted interactions have been shown in particular to affect host resistance and tolerance. In this study we investigate the evolution of several types of host disease resistance strategies, alone or in combination, in spatially structured populations. We construct a spatially explicit, individual-based stochastic model where hosts and parasites interact with each other in a spatial lattice, and interactions are restricted to a given neighbourhood of varying size. We investigate several host resistance strategies, including constitutive (expressed in all resistant hosts), induced (expressed only upon infection), and combinations thereof. We show that a costly constitutive resistance cannot reach fixation, whereas an inducible resistance strategy may become fixed in the population if the cost remains low, particularly if it impacts host recovery. We also demonstrate that mixed strategies can be maintained in the host population, and that a higher investment in a recovery-boosting inducible resistance allows for a higher investment in a constitutive response. Our simulations reveal that the spatial structure of the population impacts the selection for resistance in a complex fashion. While single strategies of resistance are generally favoured in less structured populations, mixed strategies can sometimes prevail only in highly structured environments, e.g. when combining constitutive and transmission-blocking induced responses Overall these results shed new light on the dynamics of disease resistance in a spatially-structured host-pathogen system, and advance our theoretical understanding of the evolutionary dynamics of disease resistance, a necessary step to elaborate more efficient and sustainable strategies for disease management.

Recent advances in understanding dengue

This is a selective review of recent publications on dengue clinical features, epidemiology, pathogenesis, and vaccine development placed in a context of observations made over the past half century. Four dengue viruses (DENVs) are transmitted by urban cycle mosquitoes causing diseases whose nature and severity are influenced by interacting factors such as virus, age, immune status of the host, and human genetic variability. A phenomenon that controls the kinetics of DENV infection, antibody-dependent enhancement, best explains the correlation of the vascular permeability syndrome with second heterotypic DENV infections and infection in the presence of passively acquired antibodies. Based on growing evidence in vivo and in vitro, the tissue-damaging DENV non-structural protein 1 (NS1) is responsible for most of the pathophysiological features of severe dengue. This review considers the contribution of hemophagocytic histiocytosis syndrome to cases of severe dengue, the role of movement of humans in dengue epidemiology, and modeling and planning control programs and describes a country-wide survey for dengue infections in Bangladesh and efforts to learn what controls the clinical outcome of dengue infections. Progress and problems with three tetravalent live-attenuated vaccines are reviewed. Several research mysteries remain: why is the risk of severe disease during second heterotypic DENV infection so low, why is the onset of vascular permeability correlated with defervescence, and what are the crucial components of protective immunity?

Autocidal gravid ovitraps protect humans from chikungunya virus infection by reducing Aedes aegypti mosquito populations

Background

Public health responses to outbreaks of dengue, chikungunya, and Zika virus have been stymied by the inability to control the primary vector, Aedes aegypti mosquitos. Consequently, the need for novel approaches to Aedes vector control is urgent. Placement of three autocidal gravid ovitraps (AGO traps) in ~85% of homes in a community was previously shown to sustainably reduce the density of female Ae. aegypti by >80%. Following the introduction of chikungunya virus (CHIKV) to Puerto Rico, we conducted a seroprevalence survey to estimate the prevalence of CHIKV infection in communities with and without AGO traps and evaluate their effect on reducing CHIKV transmission.

Methods and findings

Multivariate models that calculated adjusted prevalence ratios (aPR) showed that among 175 and 152 residents of communities with and without AGO traps, respectively, an estimated 26.1% and 43.8% had been infected with CHIKV (aPR = 0.50, 95% CI: 0.37–0.91). After stratification by time spent in their community, protection from CHIKV infection was strongest among residents who reported spending many or all weekly daytime hours in their community:10.3% seropositive in communities with AGO traps vs. 48.7% in communities without (PR = 0.21, 95% CI: 0.11–0.41). The age-adjusted rate of fever with arthralgia attributable to CHIKV infection was 58% (95% CI: 46–66%). The monthly number of CHIKV-infected mosquitos and symptomatic residents were diminished in communities with AGO traps compared to those without.

Conclusions

These findings indicate that AGO traps are an effective tool that protects humans from infection with a virus transmitted by Ae. aegypti mosquitos. Future studies should evaluate their protective effectiveness in large, urban communities.

Aedes species mosquitos transmit pathogens of public health importance, including dengue, Zika, and chikungunya viruses. No tools exist to control these mosquitos that sustainably and effectively prevent human infections. Autocidal gravid ovitraps (AGO traps) have been shown to sustainably reduce Aedes populations by >80%. After chikungunya virus was introduced into Puerto Rico, we conducted serosurveys in communities with and without AGO traps. We observed a two-fold lower prevalence of chikungunya virus infection among residents of communities with AGO traps compared to communities without. Among infected residents of communities with traps, a significant proportion likely had been infected while outside their community. These findings indicate that AGO traps are an effective tool that protects humans from infection with pathogens transmitted by Aedes mosquitos.

Spatial and temporal analysis of dengue infections in Queensland, Australia: Recent trend and perspectives

Dengue is a public health concern in northern Queensland, Australia. This study aimed to explore spatial and temporal characteristics of dengue cases in Queensland, and to identify high-risk areas after a 2009 dengue outbreak at fine spatial scale and thereby help in planning resource allocation for dengue control measures. Notifications of dengue cases for Queensland at Statistical Local Area (SLA) level were obtained from Queensland Health for the period 2010 to 2015. Spatial and temporal analysis was performed, including plotting of seasonal distribution and decomposition of cases, using regression models and creating choropleth maps of cumulative incidence. Both the space-time scan statistic (SaTScan) and Geographical Information System (GIS) were used to identify and visualise the space-time clusters of dengue cases at SLA level. A total of 1,773 dengue cases with 632 (35.65%) autochthonous cases and 1,141 (64.35%) overseas acquired cases were satisfied for the analysis in Queensland during the study period. Both autochthonous and overseas acquired cases occurred more frequently in autumn and showed a geographically expanding trend over the study period. The most likely cluster of autochthonous cases (Relative Risk, RR = 54.52, p<0.001) contained 50 SLAs in the north-east region of the state around Cairns occurred during 2013-2015. A cluster of overseas cases (RR of 60.81, p<0.001) occurred in a suburb of Brisbane during 2012 to 2013. These results show a clear spatiotemporal trend of recent dengue cases in Queensland, providing evidence in directing future investigations on risk factors of this disease and effective interventions in the high-risk areas.

Contrasting the value of targeted versus area-wide mosquito control scenarios to limit arbovirus transmission with human mobility patterns based on different tropical urban population centers

Vector control is still our primary intervention for both prevention and mitigation of epidemics of many vector-borne diseases. Efficiently targeting control measures is important since control can involve substantial economic costs. Targeting is not always straightforward, as transmission of vector-borne diseases is affected by various types of host movement. Here we assess how taking daily commuting patterns into consideration can help improve vector control efforts. We examine three tropical urban centers (San Juan, Recife, and Jakarta) that have recently been exposed to Zika and/or dengue infections and consider whether the distribution of human populations and resulting commuting flows affects the optimal scale at which control interventions should be implemented. We developed a stochastic, spatial model and investigated four control scenarios. The scenarios differed in the spatial extent of their implementation and were: 1) a response at the level of an individual neighborhood; 2) a response targeted at a neighborhood in which infected humans were detected and the one with which it was most strongly connected by human movement; 3) a limited area-wide response where all neighborhoods within a certain radius of the focal area were included; and 4) a collective response where all participating neighborhoods implemented control. The relative effectiveness of the scenarios varied only slightly between different settings, with the number of infections averted over time increasing with the scale of implementation. This difference depended on the efficacy of control at the neighborhood level. At low levels of efficacy, the scenarios mirrored each other in infections averted. At high levels of efficacy, impact increased with the scale of the intervention. As a result, the choice between scenarios will not only be a function of the amount of effort decision-makers are willing to invest, but largely epend on the overall effectiveness of vector control approaches.

Control and prevention of Aedes-transmitted viruses, such as dengue, chikungunya, or Zika relies heavily on vector control approaches. Given the effort and cost involved in implementation of vector control, targeting of control measures is highly desirable. However, it is unclear to what extent the effectiveness of highly focal and reactive control measures depends on the commuting and movement patterns of humans. To investigate this question, we developed a model and four control scenarios that ranged from highly focal to area-wide larval control. The distribution of humans and their commuting patterns were modelled after three major tropical urban centers, San Juan, Recife, and Jakarta. We show that as implementation is applied across a wider area, a greater number of infections is averted. Critically, this only occurs if the efficacy of control at the neighborhood level is sufficiently high. A consistent outcome across the three settings was that the focal strategy was most likely to provide the best outcome at lower levels of effort, and when the efficacy of control was low. These outcomes suggest that optimal control strategies will likely have to be tailored to individual settings by decision makers and would benefit from localized cost-effectiveness modelling studies.

Impact of adult mosquito control on dengue prevalence in a multi-patch setting: A case study in Kolkata (2014-2015)

Dengue is one of the deadliest mosquito-borne disease prevalent mainly in tropical and sub-tropical regions. Controlling the spread of this disease becomes a major concern to the public health authority. World Health Organization (WHO) adopted several mosquito control strategies to reduce the disease prevalence. In this work, a general multi-patch non-autonomous dengue model is formulated to capture the temporal and spatial transmission mechanism of the disease and the effectiveness of different adult mosquito control strategies in reducing dengue prevalence is evaluated. During the period (2014-2015) the dengue situation of Kolkata which is one of the most dengue affected city in India is considered in our study. Depending on geographical location, Kolkata is divided into five regions and our model is fitted to the monthly dengue cases of these five regions during the above-mentioned period. By considering control specific characteristics (e.g. efficacy, environment persistence) of the mosquito control strategies, we study the efficiency of three adult mosquito controls and their combined effect in reducing dengue prevalence. From our study, it is observed that control with higher environment persistence performs better in comparison to the controls having low environment persistence. It is also observed that, connectedness between the regions play a key role in the effectiveness of the control strategies.

Inferring who-infected-whom-where in the 2016 Zika outbreak in Singapore-a spatio-temporal model

Singapore experienced its first known Zika outbreak in 2016. Given the lack of herd immunity, the suitability of the climate for pathogen transmission, and the year-round presence of the vector—Aedes aegypti—Zika had the potential to become endemic, like dengue. Guillain–Barré syndrome and microcephaly are severe complications associated elsewhere with Zika and the risk of these complications makes understanding its spread imperative. We investigated the spatio-temporal spread of locally transmitted Zika in Singapore and assessed the relevance of non-residential transmission of Zika virus infections, by inferring the possible infection tree (i.e. who-infected-whom-where) and comparing inferences using geographically resolved data on cases' home, their work, or their home and work. We developed a spatio-temporal model using time of onset and both addresses of the Zika-confirmed cases between July and September 2016 to estimate the infection tree using Bayesian data augmentation. Workplaces were involved in a considerable fraction (64.2%) of infections, and homes and workplaces may be distant relative to the scale of transmission, allowing ambulant infected persons may act as the ‘vector’ infecting distant parts of the country. Contact tracing is a challenge for mosquito-borne diseases, but inferring the geographically structured transmission tree sheds light on the spatial transmission of Zika to immunologically naive regions of the country.

Use of Twitter data to improve Zika virus surveillance in the United States during the 2016 epidemic

BACKGROUND

Zika virus (ZIKV) is an emerging mosquito-borne arbovirus that can produce serious public health consequences. In 2016, ZIKV caused an epidemic in many countries around the world, including the United States. ZIKV surveillance and vector control is essential to combating future epidemics. However, challenges relating to the timely publication of case reports significantly limit the effectiveness of current surveillance methods. In many countries with poor infrastructure, established systems for case reporting often do not exist. Previous studies investigating the H1N1 pandemic, general influenza and the recent Ebola outbreak have demonstrated that time- and geo-tagged Twitter data, which is immediately available, can be utilized to overcome these limitations.

METHODS

In this study, we employed a recently developed system called Cloudberry to filter a random sample of Twitter data to investigate the feasibility of using such data for ZIKV epidemic tracking on a national and state (Florida) level. Two auto-regressive models were calibrated using weekly ZIKV case counts and zika tweets in order to estimate weekly ZIKV cases 1 week in advance.

RESULTS

While models tended to over-predict at low case counts and under-predict at extreme high counts, a comparison of predicted versus observed weekly ZIKV case counts following model calibration demonstrated overall reasonable predictive accuracy, with an R2 of 0.74 for the Florida model and 0.70 for the U.S.

MODEL

Time-series analysis of predicted and observed ZIKV cases following internal cross-validation exhibited very similar patterns, demonstrating reasonable model performance. Spatially, the distribution of cumulative ZIKV case counts (local- & travel-related) and zika tweets across all 50 U.S. states showed a high correlation (r = 0.73) after adjusting for population.

CONCLUSIONS

This study demonstrates the value of utilizing Twitter data for the purposes of disease surveillance. This is of high value to epidemiologist and public health officials charged with protecting the public during future outbreaks.

Forest cover and climate as potential drivers for dengue fever in Sumatra and Kalimantan 2006-2016: a spatiotemporal analysis

OBJECTIVES

To describe and quantify spatiotemporal trends of dengue fever at district level in Sumatra and Kalimantan, Indonesia in relation to forest cover and climatic factors.

METHODS

A spatial ecological study design was used to analyse monthly surveillance data of notified dengue fever cases from January 2006 to December 2016 in the 154 districts of Sumatra and 56 districts of Kalimantan. A multivariate, zero-inflated Poisson regression model was developed with a conditional autoregressive prior structure with posterior parameters estimated using Bayesian Markov chain Monte Carlo simulation with Gibbs sampling.

RESULTS

There were 230 745 cases in Sumatra and 132 186 cases in Kalimantan during the study period. In Sumatra, the risk of dengue fever decreased by 9% (95% credible interval [CrI] 8.5-9.5%) for a 1% increase in forest cover and by 12.2% (95% CrI 11.9-12.6%) for a 1% increase in relative humidity. In Kalimantan, dengue fever risk fell by 17.6% (95% CrI 17.1-18.1%) for a 1% increase in relative humidity and rose by 7.6% (95% CrI 6.9-8.4%) for a 1 °C increase in minimum temperature. There was no significant residual spatial clustering in Sumatra after accounting for climate and demographic variables. In Kalimantan, high residual risk areas were primarily centred in North and East of the island.

CONCLUSIONS

Dengue fever in Sumatra and Kalimantan was highly seasonal and associated with climate factors and deforestation. Incorporation of climate indicators into risk-based surveillance might be warranted for dengue fever in Indonesia.

What can urban mobility data reveal about the spatial distribution of infection in a single city?

Background

Infectious diseases spread through inherently spatial processes. Road and air traffic data have been used to model these processes at national and global scales. At metropolitan scales, however, mobility patterns are fundamentally different and less directly observable. Estimating the spatial distribution of infection has public health utility, but few studies have investigated this at an urban scale. In this study we address the question of whether the use of urban-scale mobility data can improve the prediction of spatial patterns of influenza infection. We compare the use of different sources of urban-scale mobility data, and investigate the impact of other factors relevant to modelling mobility, including mixing within and between regions, and the influence of hub and spoke commuting patterns.

Methods

We used journey-to-work (JTW) data from the Australian 2011 Census, and GPS journey data from the Sygic GPS Navigation & Maps mobile app, to characterise population mixing patterns in a spatially-explicit SEIR (susceptible, exposed, infectious, recovered) meta-population model.

Results

Using the JTW data to train the model leads to an increase in the proportion of infections that arise in central Melbourne, which is indicative of the city’s spoke-and-hub road and public transport networks, and of the commuting patterns reflected in these data. Using the GPS data increased the infections in central Melbourne to a lesser extent than the JTW data, and produced a greater heterogeneity in the middle and outer regions. Despite the limitations of both mobility data sets, the model reproduced some of the characteristics observed in the spatial distribution of reported influenza cases.

Conclusions

Urban mobility data sets can be used to support models that capture spatial heterogeneity in the transmission of infectious diseases at a metropolitan scale. These data should be adjusted to account for relevant urban features, such as highly-connected hubs where the resident population is likely to experience a much lower force of infection that the transient population. In contrast to national and international scales, the relationship between mobility and infection at an urban level is much less apparent, and requires a richer characterisation of population mobility and contact.

Electronic supplementary material

The online version of this article (10.1186/s12889-019-6968-x) contains supplementary material, which is available to authorized users.

Estimating absolute indoor density of Aedes aegypti using removal sampling

BACKGROUND

Quantification of adult Aedes aegypti abundance indoors has relied on estimates of relative density (e.g. number of adults per unit of sampling or time), most commonly using traps or timed collections using aspirators. The lack of estimates of the sensitivity of collections and lack of a numerical association between relative and the absolute density of adult Ae. aegypti represent a significant gap in vector surveillance. Here, we describe the use of sequential removal sampling to estimate absolute numbers of indoor resting Ae. aegypti and to calculate calibration coefficients for timed Prokopack aspirator collections in the city of Merida, Yucatan State, Mexico. The study was performed in 200 houses that were selected based on recent occurrence of Aedes-borne viral illness in residents. Removal sampling occurred in 10-minute sampling rounds performed sequentially until no Ae. aegypti adult was collected for 3 hours or over 2 consecutive 10-minute periods.

RESULTS

A total of 3439 Ae. aegypti were collected. The sensitivity of detection of positive houses in the first sampling round was 82.5% for any adult Ae. aegypti, 78.5% for females, 75.5% for males and 73.3% for blood-fed females. The total number of Ae. aegypti per house was on average ~5 times higher than numbers collected for the first sampling round. There was a positive linear relationship between the relative density of Ae. aegypti collected during the first 10-min round and the absolute density for all adult metrics. Coefficients from the linear regression were used to calibrate numbers from 10-min collections into estimates of absolute indoor Ae. aegypti density for all adults, females and males.

CONCLUSIONS

Exhaustive removal sampling represents a promising method for quantification of absolute indoor Ae. aegypti density, leading to improved entomological estimates of mosquito distribution, a key measure in the assessments of the risk pathogen transmission, disease modeling and the evaluation of vector control interventions.

Using Multiple Scale Space-Time Patterns in Variance-Based Global Sensitivity Analysis for Spatially Explicit Agent-Based Models

Sensitivity analysis (SA) in spatially explicit agent-based models (ABMs) has emerged to address some of the challenges associated with model specification and parameterization. For spatially explicit ABMs, the comparison of spatial or spatio-temporal patterns has been advocated to evaluate models. Nevertheless, less attention has been paid to understanding the extent to which parameter values in ABMs are responsible for mismatch between model outcomes and observations. In this paper, we propose the use of multiple scale space-time patterns in variance-based global sensitivity analysis (GSA). A vector-borne disease transmission model was used as the case study. Input factors used in GSA include one related to the environment (introduction rates), two related to interactions between agents and environment (level of herd immunity, mosquito population density), and one that defines agent state transition (mosquito extrinsic incubation period). The results show parameters related to interactions between agents and the environment have great impact on the ability of a model to reproduce observed patterns, although the magnitudes of such impacts vary by space-time scales. Additionally, the results highlight the time-dependent sensitivity to parameter values in spatially explicit ABMs. The GSA performed in this study helps in identifying the input factors that need to be carefully parameterized in the model to implement ABMs that well reproduce observed patterns at multiple space-time scales.

Estimating the impact of city-wide Aedes aegypti population control: An observational study in Iquitos, Peru

During the last 50 years, the geographic range of the mosquito Aedes aegypti has increased dramatically, in parallel with a sharp increase in the disease burden from the viruses it transmits, including Zika, chikungunya, and dengue. There is a growing consensus that vector control is essential to prevent Aedes-borne diseases, even as effective vaccines become available. What remains unclear is how effective vector control is across broad operational scales because the data and the analytical tools necessary to isolate the effect of vector-oriented interventions have not been available. We developed a statistical framework to model Ae. aegypti abundance over space and time and applied it to explore the impact of citywide vector control conducted by the Ministry of Health (MoH) in Iquitos, Peru, over a 12-year period. Citywide interventions involved multiple rounds of intradomicile insecticide space spray over large portions of urban Iquitos (up to 40% of all residences) in response to dengue outbreaks. Our model captured significant levels of spatial, temporal, and spatio-temporal variation in Ae. aegypti abundance within and between years and across the city. We estimated the shape of the relationship between the coverage of neighborhood-level vector control and reductions in female Ae. aegypti abundance; i.e., the dose-response curve. The dose-response curve, with its associated uncertainties, can be used to gauge the necessary spraying effort required to achieve a desired effect and is a critical tool currently absent from vector control programs. We found that with complete neighborhood coverage MoH intra-domicile space spray would decrease Ae. aegypti abundance on average by 67% in the treated neighborhood. Our framework can be directly translated to other interventions in other locations with geolocated mosquito abundance data. Results from our analysis can be used to inform future vector-control applications in Ae. aegypti endemic areas globally.

Vector-borne transmission and evolution of Zika virus

Zika virus (ZIKV), discovered in the Zika Forest of Uganda in 1947, is a mosquito-borne flavivirus related to yellow fever, dengue and West Nile viruses. From its discovery until 2007, only sporadic ZIKV cases were reported, with mild clinical manifestations in patients. Therefore, little attention was given to this virus before epidemics in the South Pacific and the Americas that began in 2013. Despite a growing number of ZIKV studies in the past three years, many aspects of the virus remain poorly characterized, particularly the spectrum of species involved in its transmission cycles. Here, we review the mosquito and vertebrate host species potentially involved in ZIKV vector-borne transmission worldwide. We also provide an evidence-supported analysis regarding the possibility of ZIKV spillback from an urban cycle to a zoonotic cycle outside Africa, and we review hypotheses regarding recent emergence and evolution of ZIKV. Finally, we identify critical remaining gaps in the current knowledge of ZIKV vector-borne transmission.

Prediction of dengue outbreaks based on disease surveillance, meteorological and socio-economic data

Background

The goal of this research is to create a system that can use the available relevant information about the factors responsible for the spread of dengue and; use it to predict the occurrence of dengue within a geographical region, so that public health experts can prepare for, manage and control the epidemic. Our study presents new geospatial insights into our understanding and management of health, disease and health-care systems.

Methods

We present a machine learning-based methodology capable of providing forecast estimates of dengue prediction in each of the fifty districts of Thailand by leveraging data from multiple data sources. Using a set of prediction variables, we show an increase in prediction accuracy of the model with an optimal combination of predictors which include: meteorological data, clinical data, lag variables of disease surveillance, socioeconomic data and the data encoding spatial dependence on dengue transmission. We use Generalized Additive Models (GAMs) to fit the relationships between the predictors (with a lag of one month) and the clinical data of Dengue hemorrhagic fever (DHF) using the data from 2008 to 2012. Using the data from 2013 to 2015 and a comparative set of prediction models, we evaluate the predictive ability of the fitted models according to RMSE and SRMSE as well as using adjusted R-squared value, deviance explained and change in AIC.

Results

The model allows for combining different predictors to make forecasts with a lead time of one month and also describe the statistical significance of the variables used to characterize the forecast. The discriminating ability of the final model was evaluated against Bangkok specific constant threshold and WHO moving threshold of the epidemic in terms of specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV).

Conclusions

The out-of-sample validation showed poorer results than the in-sample validation, however it demonstrated ability in detecting outbreaks up-to one month ahead. We also determine that for the predicting dengue outbreaks within a district, the influence of dengue incidences and socioeconomic data from the surrounding districts is statistically significant. This validates the influence of movement patterns of people and spatial heterogeneity of human activities on the spread of the epidemic.

Using Human Movement Data to Identify Potential Areas of Zika Transmission: Case Study of the Largest Zika Cluster in Singapore

Singapore experienced its first Zika virus (ZIKV) cluster in August 2016. To understand the implication of human movement on disease spread, a retrospective study was conducted using aggregated and anonymized mobile phone data to examine movement from the cluster to identify areas of possible transmission. An origin–destination model was developed based on the movement of three groups of individuals: (i) construction workers, (ii) residents and (iii) visitors out of the cluster locality to other parts of the island. The odds ratio of ZIKV cases in a hexagon visited by an individual from the cluster, independent of the group of individuals, is 3.20 (95% CI: 2.65–3.87, p-value < 0.05), reflecting a higher count of ZIKV cases when there is a movement into a hexagon from the cluster locality. A comparison of independent ROC curves tested the statistical significance of the difference between the areas under the curves of the three groups of individuals. Visitors (difference in AUC = 0.119) and residents (difference in AUC = 0.124) have a significantly larger difference in area under the curve compared to the construction workers (p-value < 0.05). This study supports the proof of concept of using mobile phone data to approximate population movement, thus identifying areas at risk of disease transmission.

Fogarty International Center collaborative networks in infectious disease modeling: Lessons learnt in research and capacity building

Due to a combination of ecological, political, and demographic factors, the emergence of novel pathogens has been increasingly observed in animals and humans in recent decades. Enhancing global capacity to study and interpret infectious disease surveillance data, and to develop data-driven computational models to guide policy, represents one of the most cost-effective, and yet overlooked, ways to prepare for the next pandemic. Epidemiological and behavioral data from recent pandemics and historic scourges have provided rich opportunities for validation of computational models, while new sequencing technologies and the 'big data' revolution present new tools for studying the epidemiology of outbreaks in real time. For the past two decades, the Division of International Epidemiology and Population Studies (DIEPS) of the NIH Fogarty International Center has spearheaded two synergistic programs to better understand and devise control strategies for global infectious disease threats. The Multinational Influenza Seasonal Mortality Study (MISMS) has strengthened global capacity to study the epidemiology and evolutionary dynamics of influenza viruses in 80 countries by organizing international research activities and training workshops. The Research and Policy in Infectious Disease Dynamics (RAPIDD) program and its precursor activities has established a network of global experts in infectious disease modeling operating at the research-policy interface, with collaborators in 78 countries. These activities have provided evidence-based recommendations for disease control, including during large-scale outbreaks of pandemic influenza, Ebola and Zika virus. Together, these programs have coordinated international collaborative networks to advance the study of emerging disease threats and the field of computational epidemic modeling. A global community of researchers and policy-makers have used the tools and trainings developed by these programs to interpret infectious disease patterns in their countries, understand modeling concepts, and inform control policies. Here we reflect on the scientific achievements and lessons learnt from these programs (h-index = 106 for RAPIDD and 79 for MISMS), including the identification of outstanding researchers and fellows; funding flexibility for timely research workshops and working groups (particularly relative to more traditional investigator-based grant programs); emphasis on group activities such as large-scale modeling reviews, model comparisons, forecasting challenges and special journal issues; strong quality control with a light touch on outputs; and prominence of training, data-sharing, and joint publications.

Efficacy of novel indoor residual spraying methods targeting pyrethroid-resistant Aedes aegypti within experimental houses

Challenges in maintaining high effectiveness of classic vector control in urban areas has renewed the interest in indoor residual spraying (IRS) as a promising approach for Aedes-borne disease prevention. While IRS has many benefits, application time and intrusive indoor applications make its scalability in urban areas difficult. Modifying IRS to account for Ae. aegypti resting behavior, named targeted IRS (TIRS, spraying walls below 1.5 m and under furniture) can reduce application time; however, an untested assumption is that modifications to IRS will not negatively impact entomological efficacy. We conducted a comparative experimental study evaluating the residual efficacy of classically-applied IRS (as developed for malaria control) compared to two TIRS application methods using a carbamate insecticide against a pyrethroid-resistant, field-derived Ae. aegypti strain. We performed our study within a novel experimental house setting (n = 9 houses) located in Merida (Mexico), with similar layouts and standardized contents. Classic IRS application (insecticide applied to full walls and under furniture) was compared to: a) TIRS: insecticide applied to walls below 1.5 m and under furniture, and b) Resting Site TIRS (RS-TIRS): insecticide applied only under furniture. Mosquito mortality was measured eight times post-application (out to six months post-application) by releasing 100 Ae. aegypti females /house and collecting live and dead individuals after 24 hrs exposure. Compared to Classic IRS, TIRS and RS-TIRS took less time to apply (31% and 82% reduction, respectively) and used less insecticide (38% and 85% reduction, respectively). Mortality of pyrethroid-resistant Ae. aegypti did not significantly differ among the three IRS application methods up to two months post application, and did not significantly differ between Classic IRS and TIRS up to four months post application. These data illustrate that optimizing IRS to more efficiently target Ae. aegypti can both reduce application time and insecticide volume with no apparent reduction in entomological efficacy.

Vector control is the primary strategy for managing Aedes aegypti and reducing transmission of Aedes-borne diseases; however, the indoor resting behavior of Ae. aegypti and the evolution of insecticide resistance reduces the effectiveness of many vector control tactics. Indoor residual spraying (IRS) is effective against Ae. aegypti, but lengthy application time makes IRS difficult to scale within urban environments. We compared the application and entomological efficacy of Classic IRS against two novel Aedes-targeting IRS application methods (Targeted IRS [TIRS]- insecticide applied to walls below 1.5 m and under furniture and Resting Site TIRS [RS-TIRS]- insecticide applied only under furniture) within experimental houses using a carbamate insecticide. Both TIRS and RS-TIRS took less time to apply and used less insecticide compared to Classic IRS. Mortality of pyrethroid-resistant Ae. aegypti did not differ among treatments out to two months post-application, and there was no difference in mortality between Classic IRS and TIRS out to four months post-application. These data provide evidence that IRS application methods can be improved to take less time and insecticide yet not lose entomological efficacy, making TIRS more scalable within urban environments. However, larger field studies with epidemiologic endpoints are needed to further assess the efficacy of these modified TIRS techniques.

Socio-Ecological Factors Associated with Dengue Risk and Aedes aegypti Presence in the Galápagos Islands, Ecuador

Dengue fever is an emerging infectious disease in the Galápagos Islands of Ecuador, with the first cases reported in 2002 and subsequent periodic outbreaks. We report results of a 2014 pilot study conducted in Puerto Ayora (PA) on Santa Cruz Island, and Puerto Baquerizo Moreno (PB) on San Cristobal Island. To assess the socio-ecological risk factors associated with dengue and mosquito vector presence at the household level, we conducted 100 household surveys (50 on each island) in neighborhoods with prior reported dengue cases. Adult mosquitoes were collected inside and outside the home, larval indices were determined through container surveys, and heads of households were interviewed to determine demographics, self-reported prior dengue infections, housing conditions, and knowledge, attitudes, and practices regarding dengue. Multi-model selection methods were used to derive best-fit generalized linear regression models of prior dengue infection, and Aedes aegypti presence. We found that 24% of PB and 14% of PA respondents self-reported a prior dengue infection, and more PB homes than PA homes had Ae. aegypti. The top-ranked model for prior dengue infection included several factors related to human movement, household demographics, access to water quality issues, and dengue awareness. The top-ranked model for Ae. aegypti presence included housing conditions, mosquito control practices, and dengue risk perception. This is the first study of dengue risk and Ae. aegypti presence in the Galápagos Islands.

Effects of human mobility, temperature and mosquito control on the spatiotemporal transmission of dengue

Dengue transmission exhibits evident geographic variations and seasonal differences. Such heterogeneity is caused by various impact factors, in which temperature and host/vector behaviors could drive its spatiotemporal transmission, but mosquito control could stop its progression. These factors together contribute to the observed distributions of dengue incidence from surveillance systems. To effectively and efficiently monitor and response to dengue outbreak, it would be necessary to systematically model these factors and their impacts on dengue transmission. This paper introduces a new modeling framework with consideration of multi-scale factors and surveillance data to clarify the hidden dynamics accounting for dengue spatiotemporal transmission. The model is based on compartmental system which takes into account the biting-based interactions among humans, viruses and mosquitoes, as well as the essential impacts of human mobility, temperature and mosquito control. This framework was validated with real epidemic data by applying retrospectively to the 2014 dengue epidemic in the Pearl River Delta (PRD) in southern China. The results indicated that suitable condition of temperature could be responsible for the explosive dengue outbreak in the PRD, and human mobility could be the causal factor leading to its spatial transmission across different cities. It was further found that mosquito intervention has significantly reduced dengue incidence, where a total of 52,770 (95% confidence interval [CI]: 29,231-76,308) dengue cases were prevented in the PRD in 2014. The findings can offer new insights for improving the predictability and risk assessment of dengue epidemics. The model also can be readily extended to investigate the transmission dynamics of other mosquito-borne diseases.

Feasibility of feeding Aedes aegypti mosquitoes on dengue virus-infected human volunteers for vector competence studies in Iquitos, Peru

BACKGROUND

Transmission of dengue virus (DENV) from humans to mosquitoes represents a critical component of dengue epidemiology. Examinations of this process have generally been hampered by a lack of methods that adequately represent natural acquisition of DENV by mosquitoes from humans. In this study, we assessed artificial and natural blood feeding methods based on rates of DENV infection and dissemination within mosquitoes for use in a field-based epidemiological cohort study in Iquitos, Peru.

METHODOLOGY/PRINCIPAL FINDINGS

Our study was implemented, stepwise, between 2011 and 2015. Participants who were 5 years and older with 5 or fewer days of fever were enrolled from ongoing clinic- and neighborhood-based studies on dengue in Iquitos. Wild type, laboratory-reared Aedes aegypti were fed directly on febrile individuals or on blood collected from participants that was either untreated or treated with EDTA. Mosquitoes were tested after approximately 14 days of extrinsic incubation for DENV infection and dissemination. A total of 58 participants, with viremias ranging from 1.3 × 10(2) to 2.9 × 10(6) focus-forming units per mL of serum, participated in one or more feeding methods. DENV infection and dissemination rates were not significantly different following direct and indirect-EDTA feeding; however, they were significantly lower for mosquitoes that fed indirectly on blood with no additive. Relative to direct feeding, infection rates showed greater variation following indirect-EDTA than indirect-no additive feeding. Dissemination rates were similar across all feeding methods. No differences were detected in DENV infection or dissemination rates in mosquitoes fed directly on participants with different dengue illness severity.

CONCLUSIONS/SIGNIFICANCE

Our study demonstrates the feasibility of using direct and indirect feeding methods for field-based studies on vector competence. Direct mosquito feeding is preferable in terms of logistical ease, biosecurity, and reliability.

Climate-driven variation in mosquito density predicts the spatiotemporal dynamics of dengue

Using extensive data on dengue fever and mosquito density, we demonstrate that local weather conditions, through their impact on the variation of mosquito abundance, are a driver of dengue dynamics in China. We believe that this mechanism can be applied to explain dengue dynamics in other places as well. We furthermore conjecture that our integrative approach would be applicable to other vector-borne diseases, such as Zika, malaria, and chikungunya.

Dengue is a climate-sensitive mosquito-borne disease with increasing geographic extent and human incidence. Although the climate–epidemic association and outbreak risks have been assessed using both statistical and mathematical models, local mosquito population dynamics have not been incorporated in a unified predictive framework. Here, we use mosquito surveillance data from 2005 to 2015 in China to integrate a generalized additive model of mosquito dynamics with a susceptible–infected–recovered (SIR) compartmental model of viral transmission to establish a predictive model linking climate and seasonal dengue risk. The findings illustrate that spatiotemporal dynamics of dengue are predictable from the local vector dynamics, which in turn, can be predicted by climate conditions. On the basis of the similar epidemiology and transmission cycles, we believe that this integrated approach and the finer mosquito surveillance data provide a framework that can be extended to predict outbreak risk of other mosquito-borne diseases as well as project dengue risk maps for future climate scenarios.

Comparison of stochastic and deterministic frameworks in dengue modelling

We perform estimations of compartment models for dengue transmission in rural Cambodia with increasing complexity regarding both model structure and the account for stochasticity. On the one hand, we successively account for three embedded sources of stochasticity: observation noise, demographic variability and environmental hazard. On the other hand, complexity in the model structure is increased by introducing vector-borne transmission, explicit asymptomatic infections and interacting virus serotypes. Using two sources of case data from dengue epidemics in Kampong Cham (Cambodia), models are estimated in the bayesian framework, with Markov Chain Monte Carlo and Particle Markov Chain Monte Carlo. We highlight the advantages and drawbacks of the different formulations in a practical setting. Although in this case the deterministic models provide a good approximation of the mean trajectory for a low computational cost, the stochastic frameworks better reflect and account for parameter and simulation uncertainty.

Acceptability of Aedes aegypti blood feeding on dengue virus-infected human volunteers for vector competence studies in Iquitos, Peru

As part of a study to investigate drivers of dengue virus (DENV) transmission dynamics, this qualitative study explored whether DENV-infected residents of Iquitos, Peru, considered it acceptable (1) to participate in direct mosquito feeding experiments (lab-reared Aedes aegypti mosquitoes fed directly on human volunteers) and (2) to provide blood meals indirectly (Ae. aegypti fed on blood drawn from participants by venipuncture). Twelve focus group discussions (FGDs; 94 participants: 82 females and 12 males) were conducted in January 2014 to explore six themes: (1) concerns and preferences regarding direct mosquito feeds and blood draws, (2) comprehension of and misconceptions about study procedures, (3) motivating factors for participation, (4) acceptability of children's participation, (5) willingness to provide multiple samples over several days, and (6) preference for direct feedings in homes versus the study laboratory. Results of FGDs, including one with 5 of 53 past direct mosquito feed participants, indicated that mosquito feeding procedures are acceptable to Iquitos residents when they are provided with information and a few key messages are properly reinforced. FGD participants' concerns focused primarily on safety issues rather than discomfort associated with mosquito bites. A video explaining the study dramatically increased comprehension of the study procedures. The majority of participants expressed a preference for mosquito feeding over venipuncture. Adults supported child participation if the children themselves assented. For most participants, home feedings were preferred over those in a laboratory. A major impetus for participation was the idea that results would contribute to an improved understanding of DENV transmission in Iquitos. Findings from our study will support future large-scale studies that employ direct mosquito feeding, a low-risk, non-invasive procedure that is experimentally superior to artificial mosquito feeding methods.

Probability of dengue transmission and propagation in a non-endemic temperate area: conceptual model and decision risk levels for early alert, prevention and control

BACKGROUND

Dengue viruses have spread rapidly across tropical regions of the world in recent decades. Today, dengue transmission is observed in the Americas, Southeast Asia, Western Pacific, Africa and in non-endemic areas of the USA and Europe. Dengue is responsible for 16% of travel-related febrile illnesses. Although most prevalent in tropical areas, risk maps indicate that subtropical regions are suitable for transmission. Dengue-control programs in these regions should focus on minimizing virus importation, community engagement, improved vector surveillance and control.

RESULTS

We developed a conceptual model for the probability of local introduction and propagation of dengue, comprising disease vulnerability and receptivity, in a temperate area, considering risk factors and social media indicators. Using a rich data set from a temperate area in the south of Brazil (where there is active surveillance of mosquitoes, viruses and human cases), we used a conceptual model as a framework to build two probabilistic models to estimate the probability of initiation and propagation of local dengue transmission. The final models estimated with good accuracy the probabilities of local transmission and propagation, with three and four weeks in advance, respectively. Vulnerability indicators (number of imported cases and dengue virus circulation in mosquitoes) and a receptivity indicator (vector abundance) could be optimally integrated with tweets and temperature data to estimate probability of early local dengue transmission.

CONCLUSIONS

We demonstrated how vulnerability and receptivity indicators can be integrated into probabilistic models to estimate initiation and propagation of dengue transmission. The models successfully estimate disease risk in different scenarios and periods of the year. We propose a decision model with three different risk levels to assist in the planning of prevention and control measures in temperate regions at risk of dengue introduction.

Where did I get dengue? Detecting spatial clusters of infection risk with social network data

Typical spatial disease surveillance systems associate a single address to each disease case reported, usually the residence address. Social network data offers a unique opportunity to obtain information on the spatial movements of individuals as well as their disease status as cases or controls. This provides information to identify visit locations with high risk of infection, even in regions where no one lives such as parks and entertainment zones. We develop two probability models to characterize the high-risk regions. We use a large Twitter dataset from Brazilian users to search for spatial clusters through analysis of the tweets' locations and textual content. We apply our models to both real-world and simulated data, demonstrating the advantage of our models as compared to the usual spatial scan statistic for this type of data.

Epidemiological characteristics and determinants of dengue transmission during epidemic and non-epidemic years in Fortaleza, Brazil: 2011-2015

BACKGROUND

After being eliminated during the 1950s, dengue reemerged in Brazil in the 1980s. Since then, incidence of the disease has increased, as serotypes move within and between cities. The co-circulation of multiple serotypes contributes to cycles of epidemic and interepidemic years, and a seasonal pattern of transmission is observed annually. Little is known regarding possible differences in the epidemiology of dengue under epidemic and interepidemic scenarios. This study addresses this gap and aims to assess the epidemiological characteristics and determinants of epidemic and interepidemic dengue transmission, utilizing data from the 5th largest city in Brazil (Fortaleza), at fine spatial and temporal scales.

METHODS/PRINCIPAL FINDINGS

Longitudinal models of monthly rates of confirmed dengue cases were used to estimate the differential contribution of contextual factors to dengue transmission in Fortaleza between 2011 and 2015. Models were stratified by annual climatological schedules and periods of interepidemic and epidemic transmission, controlling for social, economic, structural, entomological, and environmental factors. Results revealed distinct seasonal patterns between interepidemic and epidemic years, with persistent transmission after June in interepidemic years. Dengue was strongly associated with violence across strata, and with poverty and irregular garbage collection during periods of low transmission, but not with other indicators of public service provision or structural deprivation. Scrapyards and sites associated with tire storage were linked to incidence differentially between seasons, with the strongest associations during transitional precipitation periods. Hierarchical clustering analysis suggests that the dengue burden concentrates in the southern periphery of the city, particularly during periods of minimal transmission.

CONCLUSIONS/SIGNIFICANCE

Our findings have direct programmatic implications. Vector control operations must be sustained after June even in non-epidemic years. More specifically, scrapyards and sites associated with tires (strongly associated with incidence during periods of minimal transmission), require sustained entomological surveillance, particularly during interepidemic intervals and in the urban periphery. Intersectoral collaborations that address urban violence are critical for facilitating the regular activities of vector control agents.

Integrated Aedes management for the control of Aedes-borne diseases

BACKGROUND

Diseases caused by Aedes-borne viruses, such as dengue, Zika, chikungunya, and yellow fever, are emerging and reemerging globally. The causes are multifactorial and include global trade, international travel, urbanisation, water storage practices, lack of resources for intervention, and an inadequate evidence base for the public health impact of Aedes control tools. National authorities need comprehensive evidence-based guidance on how and when to implement Aedes control measures tailored to local entomological and epidemiological conditions.

METHODS AND FINDINGS

This review is one of a series being conducted by the Worldwide Insecticide resistance Network (WIN). It describes a framework for implementing Integrated Aedes Management (IAM) to improve control of diseases caused by Aedes-borne viruses based on available evidence. IAM consists of a portfolio of operational actions and priorities for the control of Aedes-borne viruses that are tailored to different epidemiological and entomological risk scenarios. The framework has 4 activity pillars: (i) integrated vector and disease surveillance, (ii) vector control, (iii) community mobilisation, and (iv) intra- and intersectoral collaboration as well as 4 supporting activities: (i) capacity building, (ii) research, (iii) advocacy, and (iv) policies and laws.

CONCLUSIONS

IAM supports implementation of the World Health Organisation Global Vector Control Response (WHO GVCR) and provides a comprehensive framework for health authorities to devise and deliver sustainable, effective, integrated, community-based, locally adapted vector control strategies in order to reduce the burden of Aedes-transmitted arboviruses. The success of IAM requires strong commitment and leadership from governments to maintain proactive disease prevention programs and preparedness for rapid responses to outbreaks.

Causes and Consequences of Spatial Within-Host Viral Spread

The spread of viral pathogens both between and within hosts is inherently a spatial process. While the spatial aspects of viral spread at the epidemiological level have been increasingly well characterized, the spatial aspects of viral spread within infected hosts are still understudied. Here, with a focus on influenza A viruses (IAVs), we first review experimental studies that have shed light on the mechanisms and spatial dynamics of viral spread within hosts. These studies provide strong empirical evidence for highly localized IAV spread within hosts. Since mathematical and computational within-host models have been increasingly used to gain a quantitative understanding of observed viral dynamic patterns, we then review the (relatively few) computational modeling studies that have shed light on possible factors that structure the dynamics of spatial within-host IAV spread. These factors include the dispersal distance of virions, the localization of the immune response, and heterogeneity in host cell phenotypes across the respiratory tract. While informative, we find in these studies a striking absence of theoretical expectations of how spatial dynamics may impact the dynamics of viral populations. To mitigate this, we turn to the extensive ecological and evolutionary literature on range expansions to provide informed theoretical expectations. We find that factors such as the type of density dependence, the frequency of long-distance dispersal, specific life history characteristics, and the extent of spatial heterogeneity are critical factors affecting the speed of population spread and the genetic composition of spatially expanding populations. For each factor that we identified in the theoretical literature, we draw parallels to its analog in viral populations. We end by discussing current knowledge gaps related to the spatial component of within-host IAV spread and the potential for within-host spatial considerations to inform the development of disease control strategies.

A comparison of Zika and dengue outbreaks using national surveillance data in the Dominican Republic

BACKGROUND

Aedes-borne arboviruses continue to precipitate epidemics worldwide. In Dominican Republic, the appearance of Zika virus cases that closely followed a large dengue epidemic provided an opportunity to study the different transmission drivers behind these two flaviviruses. Retrospective datasets were used to collect information on the populations at risk and descriptive statistics were used to describe the outbreaks on a national scale.

METHODOLOGY/ PRINCIPAL FINDINGS

Expectedly, box plots showed that 75% of dengue was reported in those aged <20 years while Zika infections were more widely dispersed among the population. Dengue attack rates were marginally higher among males at 25.9 per 10,000 population vs. 21.5 per 10,000 population for females. Zika infections appeared to be highly clustered among females (73.8% (95% CI 72.6%, 75.0%; p<0.05)); age-adjusted Zika attack rates among females were 7.64 per 10,000 population compared with 2.72 per 10,000 population among males. R0 calculations stratified by sex also showed a significantly higher metric among females: 1.84 (1.82, 1.87; p<0.05) when compared to males at 1.72 (1.69, 1.75; p<0.05). However, GBS attack rates stratified by sex revealed slightly higher risk in males vs. females, at 0.62 and 0.57 per 10,000 population respectively.

CONCLUSIONS/ SIGNIFICANCE

Evidence suggests little impact of existing dengue immunity on reported attack rates of Zika at the population level. Confounding of R0 and incident risk calculations by sex-specific over-reporting can alter the reliability of epidemiological metrics, which could be addressed using associated proxy syndromes or conditions to explore seemingly sex-skewed incidence. The findings indicate that community awareness campaigns, through influencing short-term health seeking behaviour, remain the most plausible mechanism behind increased reporting among women of reproductive age, although biological susceptibility cannot yet be ruled out. Media campaigns and screening are therefore recommended for women of reproductive age during Zika outbreaks. Future research should focus on clinical Zika outcomes among dengue seropositive individuals.

Epidemiology of dengue and other arboviruses in a cohort of school children and their families in Yucatan, Mexico: Baseline and first year follow-up

Dengue is the most prevalent mosquito-borne viral disease of humans and is caused by the four serotypes of dengue virus. To estimate the incidence of dengue and other arboviruses, we analyzed the baseline and first year follow-up of a prospective school-based cohort study and their families in three cities in the state of Yucatan, Mexico. Through enhanced surveillance activities, acute febrile illnesses in the participants were detected and yearly blood samples were collected to evaluate dengue infection incidence. A Cox model was fitted to identify hazard ratios of arboviral infections in the first year of follow-up of the cohort. The incidence of dengue symptomatic infections observed during the first year of follow-up (2015-2016) was 3.5 cases per 1,000 person-years (95% CI: 1.9, 5.9). The incidence of dengue infections was 33.9 infections per 1,000 person-years (95% CI: 31.7, 48.0). The majority of dengue infections and seroconversions were observed in the younger age groups (≤ 14 years old). Other arboviruses were circulating in the state of Yucatan during the study period. The incidence of symptomatic chikungunya infections was 8.6 per 1,000 person-years (95% CI: 5.8, 12.3) and the incidence of symptomatic Zika infections was 2.3 per 1,000 person-years (95% CI: 0.9, 4.5). Our model shows that having a dengue infection during the first year of follow-up was significantly associated with being female, living in Ticul or Progreso, and being dengue naïve at baseline. Age was not significantly associated with the outcome, it was confounded by prior immunity to dengue that increases with age. This is the first report of a cohort in Latin America that provides incidence estimates of the three arboviruses co-circulating in all age groups. This study provides important information for understanding the epidemiology of dengue and other arboviruses and better informing public health policies.

Bayesian spatial and spatio-temporal approaches to modelling dengue fever: a systematic review

Dengue fever (DF) is one of the world's most disabling mosquito-borne diseases, with a variety of approaches available to model its spatial and temporal dynamics. This paper aims to identify and compare the different spatial and spatio-temporal Bayesian modelling methods that have been applied to DF and examine influential covariates that have been reportedly associated with the risk of DF. A systematic search was performed in December 2017, using Web of Science, Scopus, ScienceDirect, PubMed, ProQuest and Medline (via Ebscohost) electronic databases. The search was restricted to refereed journal articles published in English from January 2000 to November 2017. Thirty-one articles met the inclusion criteria. Using a modified quality assessment tool, the median quality score across studies was 14/16. The most popular Bayesian statistical approach to dengue modelling was a generalised linear mixed model with spatial random effects described by a conditional autoregressive prior. A limited number of studies included spatio-temporal random effects. Temperature and precipitation were shown to often influence the risk of dengue. Developing spatio-temporal random-effect models, considering other priors, using a dataset that covers an extended time period, and investigating other covariates would help to better understand and control DF transmission.

Using Multiple Scale Spatio-Temporal Patterns for Validating Spatially Explicit Agent-Based Models

Spatially explicit agent-based models (ABMs) have been widely utilized to simulate the dynamics of spatial processes that involve the interactions of individual agents. The assumptions embedded in the ABMs may be responsible for uncertainty in the model outcomes. To ensure the reliability of the outcomes in terms of their space-time patterns, model validation should be performed. In this paper, we propose the use of multiple scale spatio-temporal patterns for validating spatially explicit ABMs. We evaluated several specifications of vector-borne disease transmission models by comparing space-time patterns of model outcomes to observations at multiple scales via the sum of root mean square error (RMSE) measurement. The results indicate that specifications of the spatial configurations of residential area and immunity status of individual humans are of importance to reproduce observed patterns of dengue outbreaks at multiple space-time scales. Our approach to using multiple scale spatio-temporal patterns can help not only to understand the dynamic associations between model specifications and model outcomes, but also to validate spatially explicit ABMs.

Projecting the end of the Zika virus epidemic in Latin America: a modelling analysis

BACKGROUND

Zika virus (ZIKV) emerged in Latin America and the Caribbean (LAC) region in 2013, with serious implications for population health in the region. In 2016, the World Health Organization declared the ZIKV outbreak a Public Health Emergency of International Concern following a cluster of associated neurological disorders and neonatal malformations. In 2017, Zika cases declined, but future incidence in LAC remains uncertain due to gaps in our understanding, considerable variation in surveillance and the lack of a comprehensive collation of data from affected countries.

METHODS

Our analysis combines information on confirmed and suspected Zika cases across LAC countries and a spatio-temporal dynamic transmission model for ZIKV infection to determine key transmission parameters and projected incidence in 90 major cities within 35 countries. Seasonality was determined by spatio-temporal estimates of Aedes aegypti vectorial capacity. We used country and state-level data from 2015 to mid-2017 to infer key model parameters, country-specific disease reporting rates, and the 2018 projected incidence. A 10-fold cross-validation approach was used to validate parameter estimates to out-of-sample epidemic trajectories.

RESULTS

There was limited transmission in 2015, but in 2016 and 2017 there was sufficient opportunity for wide-spread ZIKV transmission in most cities, resulting in the depletion of susceptible individuals. We predict that the highest number of cases in 2018 would present within some Brazilian States (Sao Paulo and Rio de Janeiro), Colombia and French Guiana, but the estimated number of cases were no more than a few hundred. Model estimates of the timing of the peak in incidence were correlated (p < 0.05) with the reported peak in incidence. The reporting rate varied across countries, with lower reporting rates for those with only confirmed cases compared to those who reported both confirmed and suspected cases.

CONCLUSIONS

The findings suggest that the ZIKV epidemic is by and large over within LAC, with incidence projected to be low in most cities in 2018. Local low levels of transmission are probable, but the estimated rate of infection suggests that most cities have a population with high levels of herd immunity.

Optimal multi-source forecasting of seasonal influenza

Forecasting the emergence and spread of influenza viruses is an important public health challenge. Timely and accurate estimates of influenza prevalence, particularly of severe cases requiring hospitalization, can improve control measures to reduce transmission and mortality. Here, we extend a previously published machine learning method for influenza forecasting to integrate multiple diverse data sources, including traditional surveillance data, electronic health records, internet search traffic, and social media activity. Our hierarchical framework uses multi-linear regression to combine forecasts from multiple data sources and greedy optimization with forward selection to sequentially choose the most predictive combinations of data sources. We show that the systematic integration of complementary data sources can substantially improve forecast accuracy over single data sources. When forecasting the Center for Disease Control and Prevention (CDC) influenza-like-illness reports (ILINet) from week 48 through week 20, the optimal combination of predictors includes public health surveillance data and commercially available electronic medical records, but neither search engine nor social media data.

Spatiotemporal diffusion of influenza A (H1N1): Starting point and risk factors

Influenza constitutes a major challenge to world health authorities due to high transmissibility and the capacity to generate large epidemics. This study aimed to characterize the diffusion process of influenza A (H1N1) by identifying the starting point of the epidemic as well as climatic and sociodemographic factors associated with the occurrence and intensity of transmission of the disease. The study was carried out in the Brazilian state of Paraná, where H1N1 caused the largest impact. The units of spatial and temporal analysis were the municipality of residence of the cases and the epidemiological weeks of the year 2009, respectively. Under the Bayesian paradigm, parametric inference was performed through a two-part spatiotemporal model and the integrated nested Laplace approximation (INLA) algorithm. We identified the most likely starting points through the effective distance measure based on mobility networks. The proposed estimation methodology allowed for rapid and efficient implementation of the spatiotemporal model, and provided evidence of different patterns for chance of occurrence and risk of influenza throughout the epidemiological weeks. The results indicate the capital city of Curitiba as the probable starting point, and showed that the interventions that focus on municipalities with greater migration and density of people, especially those with higher Human Development Indexes (HDIs) and the presence of municipal air and road transport, could play an important role in mitigation of effects of future influenza pandemics on public health. These results provide important information on the process of introduction and spread of influenza, and could contribute to the identification of priority areas for surveillance as well as establishment of strategic measures for disease prevention and control. The proposed model also allows identification of epidemiological weeks with high chance of influenza occurrence, which can be used as a reference criterion for creating an immunization campaign schedule.

Fine-scale GPS tracking to quantify human movement patterns and exposure to leptospires in the urban slum environment

Background

Human movement is likely an important risk factor for environmentally-transmitted pathogens. While epidemiologic studies have traditionally focused on household risk factors, individual movement data could provide critical additional information about risk of exposure to such pathogens. We conducted global positioning system (GPS) tracking of urban slum residents to quantify their fine-scale movement patterns and evaluate their exposures to environmental sources of leptospirosis transmission.

Methodology/Principal findings

We recruited participants from an ongoing cohort study in an urban slum in Brazil and tracked them for 24 hours at 30-second intervals. Among 172 subjects asked to participate in this cross-sectional study, 130 agreed to participate and 109 had good quality data and were included in analyses. The majority of recorded locations were near participant residences (87.7% within 50 meters of the house), regardless of age or gender. Similarly, exposure to environmental sources of leptospirosis transmission did not vary by age or gender. However, males, who have higher infection rates, visited a significantly larger area during the 24-hour period than did females (34,549m² versus 22,733m², p = 0.005). Four male participants had serologic evidence of Leptospira infection during the study period. These individuals had significantly larger activity spaces than uninfected males (61,310m² vs 31,575m², p = 0.006) and elevated exposure to rodent activity (p = 0.046) and trash deposits (p = 0.031).

Conclusions/Significance

GPS tracking was an effective tool for quantifying individual mobility in the complex urban slum environment and identifying risk exposures associated with that movement. This study suggests that in addition to source reduction, barrier interventions that reduce contact with transmission sources as slum residents move within their communities may be a useful prevention strategy for leptospirosis.

Environmental features of urban slums including inadequate sanitation, substandard housing, and population crowding predispose residents to numerous infections. Despite this shared environment, not all slum residents, even within households, have equal risk of infection with specific pathogens and we do not know why. Individual movement data will help us better understand how slum residents interact with their environment. We conducted GPS tracking of 109 urban slum residents in Brazil to quantify their movement patterns and how these influence exposure to leptospirosis, an environmentally transmitted infection common in urban slums. Slum inhabitants, regardless of age and gender, spent most of their time close to home and had similar exposures to environmental features associated with leptospirosis infection. However, males visited a larger area on a daily basis, which may explain their higher leptospirosis risk. Based on screening of the slum population conducted at six-month intervals, four individuals (all male) became infected with Leptospira during our study. These individuals visited a significantly larger area than other males and had higher exposure to rodents and trash deposits than did other participants. GPS tracking allowed us to identify movement and movement-induced exposure as risk factors for leptospirosis infection and could provide similarly important information for other environmentally-transmitted pathogens.

Mosquito-borne transmission in urban landscapes: the missing link between vector abundance and human density

With escalating urbanization, the environmental, demographic, and socio-economic heterogeneity of urban landscapes poses a challenge to mathematical models for the transmission of vector-borne infections. Classical coupled vector–human models typically assume that mosquito abundance is either independent from, or proportional to, human population density, implying a decreasing force of infection, or per capita infection rate with host number. We question these assumptions by introducing an explicit dependence between host and vector densities through different recruitment functions, whose dynamical consequences we examine in a modified model formulation. Contrasting patterns in the force of infection are demonstrated, including in particular increasing trends when recruitment grows sufficiently fast with human density. Interaction of these patterns with seasonality in temperature can give rise to pronounced differences in timing, relative peak sizes, and duration of epidemics. These proposed dependencies explain empirical dengue risk patterns observed in the city of Delhi where socio-economic status has an impact on both human and mosquito densities. These observed risk trends with host density are inconsistent with current standard models. A better understanding of the connection between vector recruitment and host density is needed to address the population dynamics of mosquito-transmitted infections in urban landscapes.

Stratified sero-prevalence revealed overall high disease burden of dengue but suboptimal immunity in younger age groups in Pune, India

BACKGROUND

In India, dengue disease is emerging as the most important vector borne public health problem due to rapid and unplanned urbanization, high human density and week management of the disease. Clinical cases are grossly underreported and not much information is available on prevalence and incidence of the disease.

METHODOLOGY

A cross sectional, stratified, facility based, multistage cluster sampling was conducted between May 4 and June 27, 2017 in Pune city. A total of 1,434 participants were enrolled. The serum samples were tested for detection of historical dengue IgG antibodies by ELISA using the commercial Panbio Dengue IgG Indirect ELISA kit. Anti-dengue IgG-capture Panbio ELISA was used for detection of high titered antibodies to detect recent secondary infection. We used this data to estimate key transmission parameters like force of infection and basic reproductive number. A subset of 120 indirect ELISA positive samples was also tested for Plaque Reduction Neutralizing Antibodies for determining serotype-specific prevalence.

FINDINGS

Overall, 81% participants were infected with dengue virus (DENV) at least once if not more. The positivity was significantly different in different age groups. All the adults above 70 years were positive for DENV antibodies. Over 69% participants were positive for neutralizing antibodies against all 4 serotypes suggesting intense transmission of all DENV serotypes in Pune. Age-specific seroprevalence was consistent with long-term, endemic circulation of DENV. There was an increasing trend with age, from 21.6% among <36 months to 59.4% in age group 10-12 years. We estimate that 8.68% of the susceptible population gets infected by DENV each year resulting into more than 3,00,000 infections and about 47,000 to 59,000 cases per year. This transmission intensity is similar to that reported from other known hyper-endemic settings in Southeast Asia and the Americas but significantly lower than report from Chennai.

CONCLUSIONS

Our study suggests that Pune city has high disease burden, all 4 serotypes are circulating, significant spatial heterogeneity in seroprevalence and suboptimal immunity in younger age groups. This would allow informed decisions to be made on management of dengue and introduction of upcoming dengue vaccines in the city.

Using Google Location History data to quantify fine-scale human mobility

Background

Human mobility is fundamental to understanding global issues in the health and social sciences such as disease spread and displacements from disasters and conflicts. Detailed mobility data across spatial and temporal scales are difficult to collect, however, with movements varying from short, repeated movements to work or school, to rare migratory movements across national borders. While typical sources of mobility data such as travel history surveys and GPS tracker data can inform different typologies of movement, almost no source of readily obtainable data can address all types of movement at once.

Methods

Here, we collect Google Location History (GLH) data and examine it as a novel source of information that could link fine scale mobility with rare, long distance and international trips, as it uniquely spans large temporal scales with high spatial granularity. These data are passively collected by Android smartphones, which reach increasingly broad audiences, becoming the most common operating system for accessing the Internet worldwide in 2017. We validate GLH data against GPS tracker data collected from Android users in the United Kingdom to assess the feasibility of using GLH data to inform human movement.

Results

We find that GLH data span very long temporal periods (over a year on average in our sample), are spatially equivalent to GPS tracker data within 100 m, and capture more international movement than survey data. We also find GLH data avoid compliance concerns seen with GPS trackers and bias in self-reported travel, as GLH is passively collected. We discuss some settings where GLH data could provide novel insights, including infrastructure planning, infectious disease control, and response to catastrophic events, and discuss advantages and disadvantages of using GLH data to inform human mobility patterns.

Conclusions

GLH data are a greatly underutilized and novel dataset for understanding human movement. While biases exist in populations with GLH data, Android phones are becoming the first and only device purchased to access the Internet and various web services in many middle and lower income settings, making these data increasingly appropriate for a wide range of scientific questions.

Electronic supplementary material

The online version of this article (10.1186/s12942-018-0150-z) contains supplementary material, which is available to authorized users.

Determinants of Short-term Movement in a Developing Region and Implications for Disease Transmission

BACKGROUND

Human mobility is important for infectious disease spread. However, little is known about how travel varies by demographic groups and how this heterogeneity influences infectious disease risk.

METHODS

We analyzed 10 years of survey data from 15 communities in a remote but rapidly changing region in rural Ecuador where road development in the past 15-20 years has dramatically changed travel. We identify determinants of travel and incorporate them into an infection transmission model.

RESULTS

Individuals living in communities more remote at baseline had lower travel rates compared with less remote villages (adjusted odds ratio [OR] = 0.51; 95% confidence interval [CI] = 0.38, 0.67). Our model predicts that less remote villages are, therefore, at increased disease risk. Though road building and travel increased for all communities, this risk differential remained over 10 years of observation. Our transmission model also suggests that travelers and nontravelers have different roles in disease transmission. Adults travel more than children (adjusted OR = 1.73; 95% CI = 1.30, 2.31) and therefore disseminate infection from population centers to rural communities. Children are more likely than adults to be infected locally (attributable fraction = 0.24 and 0.09, respectively) and were indirectly affected by adult travel patterns.

CONCLUSIONS

These results reinforce the importance of large population centers for regional transmission and show that children and adults may play different roles in disease spread. Changing transportation infrastructure and subsequent economic and social transitions are occurring worldwide, potentially causing increased regional risk of disease.

Quantifying the spatial spread of dengue in a non-endemic Brazilian metropolis via transmission chain reconstruction

The ongoing geographical expansion of dengue is inducing an epidemiological transition in many previously transmission-free urban areas, which are now prone to annual epidemics. To analyze the spatiotemporal dynamics of dengue in these settings, we reconstruct transmission chains in Porto Alegre, Brazil, by applying a Bayesian inference model to geo-located dengue cases from 2013 to 2016. We found that transmission clusters expand by linearly increasing their diameter with time, at an average rate of about 600 m month⁻¹. The majority (70.4%, 95% CI: 58.2–79.8%) of individual transmission events occur within a distance of 500 m. Cluster diameter, duration, and epidemic size are proportionally smaller when control interventions were more timely and intense. The results suggest that a large proportion of cases are transmitted via short-distance human movement (<1 km) and a limited contribution of long distance commuting within the city. These results can assist the design of control policies, including insecticide spraying and strategies for active case finding.

There is increasing urgency to understand the spatiotemporal dynamics of dengue in non-endemic regions. Here, the authors reconstruct likely dengue transmission chains in the city of Porto Alegre based on geo-located cases only, and find that most transmission events occur over short-distances.

Quantifying the spatial spread of dengue in a non-endemic Brazilian metropolis via transmission chain reconstruction

The ongoing geographical expansion of dengue is inducing an epidemiological transition in many previously transmission-free urban areas, which are now prone to annual epidemics. To analyze the spatiotemporal dynamics of dengue in these settings, we reconstruct transmission chains in Porto Alegre, Brazil, by applying a Bayesian inference model to geo-located dengue cases from 2013 to 2016. We found that transmission clusters expand by linearly increasing their diameter with time, at an average rate of about 600 m month^-1. The majority (70.4%, 95% CI: 58.2-79.8%) of individual transmission events occur within a distance of 500 m. Cluster diameter, duration, and epidemic size are proportionally smaller when control interventions were more timely and intense. The results suggest that a large proportion of cases are transmitted via short-distance human movement (<1 km) and a limited contribution of long distance commuting within the city. These results can assist the design of control policies, including insecticide spraying and strategies for active case finding.