Empirical model for estimating dengue incidence using temperature, rainfall, and relative humidity: a 19-year retrospective analysis in East Delhi

Improving dengue fever predictions in Taiwan based on feature selection and random forests

BACKGROUND

Dengue fever is a well-studied vector-borne disease in tropical and subtropical areas of the world. Several methods for predicting the occurrence of dengue fever in Taiwan have been proposed. However, to the best of our knowledge, no study has investigated the relationship between air quality indices (AQIs) and dengue fever in Taiwan.

RESULTS

This study aimed to develop a dengue fever prediction model in which meteorological factors, a vector index, and AQIs were incorporated into different machine learning algorithms. A total of 805 meteorological records from 2013 to 2015 were collected from government open-source data after preprocessing. In addition to well-known dengue-related factors, we investigated the effects of novel variables, including particulate matter with an aerodynamic diameter < 10 µm (PM10), PM2.5, and an ultraviolet index, for predicting dengue fever occurrence. The collected dataset was randomly divided into an 80% training set and a 20% test set. The experimental results showed that the random forests achieved an area under the receiver operating characteristic curve of 0.9547 for the test set, which was the best compared with the other machine learning algorithms. In addition, the temperature was the most important factor in our variable importance analysis, and it showed a positive effect on dengue fever at < 30 °C but had less of an effect at > 30 °C. The AQIs were not as important as temperature, but one was selected in the process of filtering the variables and showed a certain influence on the final results.

CONCLUSIONS

Our study is the first to demonstrate that AQI negatively affects dengue fever occurrence in Taiwan. The proposed prediction model can be used as an early warning system for public health to prevent dengue fever outbreaks.

The effect of temperature on dengue virus transmission by Aedes mosquitoes

Dengue is prevalent in tropical and subtropical regions. As an arbovirus disease, it is mainly transmitted by Aedes aegypti and Aedes albopictus. According to the previous studies, temperature is closely related to the survival of Aedes mosquitoes, the proliferation of dengue virus (DENV) and the vector competence of Aedes to transmit DENV. This review describes the correlations between temperature and dengue epidemics, and explores the potential reasons including the distribution and development of Aedes mosquitoes, the structure of DENV, and the vector competence of Aedes mosquitoes. In addition, the immune and metabolic mechanism are discussed on how temperature affects the vector competence of Aedes mosquitoes to transmit DENV.

The application of geographic information system for dengue epidemic in Southeast Asia: A review on trends and opportunity

The infectious disease dengue hemorrhagic fever remains an unresolved global problem, with climatic conditions and the location of areas located at the equator more often infected with dengue fever. Various modeling approaches have been employed for the development of a dengue risk map. The geographic information system approach was used as an instrument in applying mathematical algorithms to process field vector data into a preventive objective which is studied, then the application of remote sensing provides spatial-temporal data related to land use/land cover data sources as other variable categories. Map of hotspots for dengue fever cases is used to identify the risk of dengue fever areas by applying various complex methodologies, analysis, and visualization of advanced data are needed for its application in public health. In the last 10 years, the increase in the publication of dengue hemorrhagic fever in Southeast Asia in reputable international journals has increased significantly.

A retrospective study of environmental predictors of dengue in Delhi from 2015 to 2018 using the generalized linear model

Dengue fever is a mosquito-borne infection with a rising trend, expected to increase further with the rise in global temperature. The study aimed to use the environmental and dengue data 2015–2018 to examine the seasonal variation and establish a probabilistic model of environmental predictors of dengue using the generalized linear model (GLM). In Delhi, dengue cases started emerging in the monsoon season, peaked in the post-monsoon, and thereafter, declined in early winter. The annual trend of dengue cases declined, but the seasonal pattern remained alike (2015–18). The Spearman correlation coefficient of dengue was significantly high with the maximum and minimum temperature at 2 months lag, but it was negatively correlated with the difference of average minimum and maximum temperature at lag 1 and 2. The GLM estimated β coefficients of environmental predictors such as temperature difference, cumulative rainfall, relative humidity and maximum temperature were significant (p < 0.01) at different lag (0 to 2), and maximum temperature at lag 2 was having the highest effect (IRR 1.198). The increasing temperature of two previous months and cumulative rainfall are the best predictors of dengue incidence. The vector control should be implemented at least 2 months ahead of disease transmission (August–November).

Development and Comparison of Dengue Vulnerability Indices Using GIS-Based Multi-Criteria Decision Analysis in Lao PDR and Thailand

Dengue is a continuous health burden in Laos and Thailand. We assessed and mapped dengue vulnerability in selected provinces of Laos and Thailand using multi-criteria decision approaches. An ecohealth framework was used to develop dengue vulnerability indices (DVIs) that explain links between population, social and physical environments, and health to identify exposure, susceptibility, and adaptive capacity indicators. Three DVIs were constructed using two objective approaches, Shannon's Entropy (SE) and the Water-Associated Disease Index (WADI), and one subjective approach, the Best-Worst Method (BWM). Each DVI was validated by correlating the index score with dengue incidence for each spatial unit (district and subdistrict) over time. A Pearson's correlation coefficient (r) larger than 0.5 and a p-value less than 0.05 implied a good spatial and temporal performance. Spatially, DVIWADI was significantly correlated on average in 19% (4-40%) of districts in Laos (mean r = 0.5) and 27% (15-53%) of subdistricts in Thailand (mean r = 0.85). The DVISE was validated in 22% (12-40%) of districts in Laos and in 13% (3-38%) of subdistricts in Thailand. The DVIBWM was only developed for Laos because of lack of data in Thailand and was significantly associated with dengue incidence on average in 14% (0-28%) of Lao districts. The DVIWADI indicated high vulnerability in urban centers and in areas with plantations and forests. In 2019, high DVIWADI values were observed in sparsely populated areas due to elevated exposure, possibly from changes in climate and land cover, including urbanization, plantations, and dam construction. Of the three indices, DVIWADI was the most suitable vulnerability index for the study area. The DVIWADI can also be applied to other water-associated diseases, such as Zika and chikungunya, to highlight priority areas for further investigation and as a tool for prevention and interventions.

Co-Circulation of All Four Dengue Viruses and Zika Virus in Guerrero, Mexico, 2019

A clinical and entomological investigation was performed to identify flavivirus infections in humans and mosquitoes in impoverished areas of Guerrero, a coastal state in southwestern Mexico. A total of 639 patients with acute febrile illness and 830 resting female mosquitoes in low-income communities of Guerrero in 2019 were tested for evidence of flavivirus infection. Sera were collected from all patients and screened at a dilution of 1:20 by plaque reduction neutralization test (PRNT) using dengue virus (DENV)2. A total of 431 (67.4%) patients were seropositive. Sera from a subset of seropositive patients (n = 263) were tested for flavivirus NS1 by enzyme-linked immunosorbent assay. Forty-eight (18.3%) sera contained viral antigen. All NS1-positive sera were titrated and further tested by PRNT using DENV-1 to -4, St. Louis encephalitis virus, West Nile virus, and Zika virus (ZIKV). Seven patients were seropositive for DENV-1, five patients were seropositive for DENV-2, one patient was seropositive for DENV-3, and two patients each were seropositive for DENV-4 and ZIKV. The remainder had secondary flavivirus infections or antibodies to an undetermined flavivirus. Comparative PRNTs were also performed on 60 randomly selected NS1-negative sera, identifying patients seropositive for DENV-2, DENV-3, and ZIKV. The entomological investigation yielded 736 Aedes aegypti and 94 Culex quinquefasciatus that were sorted into 183 pools and 20 pools, respectively. Mosquitoes were assayed for flavivirus RNA by RT-PCR and Sanger sequencing. DENV-2 RNA was detected in three pools of A. aegypti. In summary, we provide evidence for the concurrent circulation of all four DENVs and ZIKV in Guerrero, Mexico. The public health authorities reported no cases of DENV-3, DENV-4, and ZIKV in Guerrero in 2019 and thus, we provide evidence of under-reporting in the region.

Dengue risk assessment using multicriteria decision analysis: A case study of Bhutan

Background

Dengue is the most rapidly spreading vector-borne disease globally, with a 30-fold increase in global incidence over the last 50 years. In Bhutan, dengue incidence has been on the rise since 2004, with numerous outbreaks reported across the country. The aim of this study was to identify and map areas that are vulnerable to dengue in Bhutan.

Methodology/Principal findings

We conducted a multicriteria decision analysis (MCDA) using a weighted linear combination (WLC) to obtain a vulnerability map of dengue. Risk factors (criteria) were identified and assigned with membership values for vulnerability according to the available literature. Sensitivity analysis and validation of the model was conducted to improve the robustness and predictive ability of the map. Our study revealed marked differences in geographical vulnerability to dengue by location and season. Low-lying areas and those located along the southern border were consistently found to be at higher risk of dengue. The vulnerability extended to higher elevation areas including some areas in the Capital city Thimphu during the summer season. The higher risk was mostly associated with relatively high population density, agricultural and built-up landscapes and relatively good road connectivity.

Conclusions

Using MCDA, our study identified vulnerable areas in Bhutan during specific seasons when and where the transmission of dengue is most likely to occur. This study provides evidence for the National Vector-borne Disease Control programme to optimize the use of limited public health resources for surveillance and vector control, to mitigate the public health threat of dengue.

Dengue is an important vector-borne viral disease affecting humans. In Bhutan, dengue incidence is on the rise with increased frequency of outbreaks and spread to new areas. Outbreaks were reported from places as high as above 900m above sea level in recent years. However, dengue control activities in Bhutan are usually initiated at the time of outbreaks. This often leads to a large number of cases and overburden the health system. To address these issues, we developed dengue risk maps at a fine spatial resolution by combining risk factors that mediate the transmission of dengue using a weighted linear combination. Vulnerability to dengue was spatially heterogeneous and varied by season. Dengue is highly vulnerable in low-lying areas throughout the season. However, the vulnerability extended to higher geographical elevations including the nation’s capital during the summer season. The study provides a firm evidence-base to prioritize areas and seasons for dengue control strategies in Bhutan.

Temporal variation and geospatial clustering of dengue in Delhi, India 2015-2018

OBJECTIVES

The study was focused on geographical mapping of dengue cases and also to identify the hotspots or high-risk areas of dengue in Delhi.

DESIGN

A retrospective spatial-temporal (ecological) study. Descriptive analysis was used to know the distribution of dengue cases by age, sex, seasons and districts of Delhi. The spatiotemporal analysis was performed using inverse distance weighting and Getis-Ord Gi* statistic to know the geographical distribution and identify the hotspot areas.

SETTINGS

All the confirmed and diagnosed dengue cases (IgM +ve or NS1 Antigen +ve ELISA) recorded by the Municipal Corporation of Delhi for the last 4 years (2015-2018) were collected with their local address. The location of all the dengue cases was geocoded using their address to prepare the spatiotemporal dengue database.

PARTICIPANTS

Record of all the dengue cases (4179) reported for treatment in the hospitals during the past 4 years were extracted and included in the study. Data were not collected directly from dengue patients.

RESULTS

Seasonal occurrence of dengue cases (4179) shows that the cases start emerging in July, peaked in September-October and declined in December. The proportions of dengue cases were recorded high among the males 57.3% compared with females 42.6%, and differences were also recorded in all the age groups with more cases in age groups <15 and 16-30 years. Mapping of the cases reflects the spatial heterogeneity in the geographical distribution. The geomapping of cases indicates the presence of a significantly high number of cases in West, Southwest, South and Southeast districts of Delhi. High-risk areas or hotspots were also identified in this region.

CONCLUSION

Dengue occurrence shows significant association with age, sex and seasons. The spatial analysis identified the high-risk areas, which can aid health administrators to take necessary action for prevention and better disease management.

Spatial and temporal patterns of dengue incidence in Bhutan: a Bayesian analysis

Dengue is an important emerging vector-borne disease in Bhutan. This study aimed to quantify the spatial and temporal patterns of dengue and their relationship to environmental factors in dengue-affected areas at the sub-district level. A multivariate zero-inflated Poisson regression model was developed using a Bayesian framework with spatial and spatiotemporal random effects modelled using a conditional autoregressive prior structure. The posterior parameters were estimated using Bayesian Markov Chain Monte Carlo simulation with Gibbs sampling. A total of 708 dengue cases were notified through national surveillance between January 2016 and June 2019. Individuals aged ≤14 years were found to be 53% (95% CrI: 42%, 62%) less likely to have dengue infection than those aged >14 years. Dengue cases increased by 63% (95% CrI: 49%, 77%) for a 1°C increase in maximum temperature, and decreased by 48% (95% CrI: 25%, 64%) for a one-unit increase in normalized difference vegetation index (NDVI). There was significant residual spatial clustering after accounting for climate and environmental variables. The temporal trend was significantly higher than the national average in eastern sub-districts. The findings highlight the impact of climate and environmental variables on dengue transmission and suggests prioritizing high-risk areas for control strategies.

Real-time dengue forecast for outbreak alerts in Southern Taiwan

Dengue fever is a viral disease transmitted by mosquitoes. In recent decades, dengue fever has spread throughout the world. In 2014 and 2015, southern Taiwan experienced its most serious dengue outbreak in recent years. Some statistical models have been established in the past, however, these models may not be suitable for predicting huge outbreaks in 2014 and 2015. The control of dengue fever has become the primary task of local health agencies. This study attempts to predict the occurrence of dengue fever in order to achieve the purpose of timely warning. We applied a newly developed autoregressive model (AR model) to assess the association between daily weather variability and daily dengue case number in 2014 and 2015 in Kaohsiung, the largest city in southern Taiwan. This model also contained additional lagged weather predictors, and developed 5-day-ahead and 15-day-ahead predictive models. Our results indicate that numbers of dengue cases in Kaohsiung are associated with humidity and the biting rate (BR). Our model is simple, intuitive and easy to use. The developed model can be embedded in a "real-time" schedule, and the data (at present) can be updated daily or weekly based on the needs of public health workers. In this study, a simple model using only meteorological factors performed well. The proposed real-time forecast model can help health agencies take public health actions to mitigate the influences of the epidemic.

Meteorological conditions are the most frequently mentioned factors in the study of dengue fever. Some of the main factors other than the purely meteorological about which the public-health authorities might have data, such as numbers of cases or other current measurements of dengue outbreaks in neighboring cities, had been used in some of the past dengue studies. In this study, we developed models for predicting dengue case number based on past dengue case data and meteorological data. The goal of the models is to provide early warning of the occurrence of dengue fever to assist public health agencies in preparing an epidemic response plan.

Spatiotemporal analysis of historical records (2001-2012) on dengue fever in Vietnam and development of a statistical model for forecasting risk

BACKGROUND

Dengue fever is the most widespread infectious disease of humans transmitted by Aedes mosquitoes. It is the leading cause of hospitalization and death in children in the Southeast Asia and western Pacific regions. We analyzed surveillance records from health centers in Vietnam collected between 2001-2012 to determine seasonal trends, develop risk maps and an incidence forecasting model.

METHODS

The data were analyzed using a hierarchical spatial Bayesian model that approximates its posterior parameter distributions using the integrated Laplace approximation algorithm (INLA). Meteorological, altitude and land cover (LC) data were used as predictors. The data were grouped by province (n = 63) and month (n = 144) and divided into training (2001-2009) and validation (2010-2012) sets. Thirteen meteorological variables, 7 land cover data and altitude were considered as predictors. Only significant predictors were kept in the final multivariable model. Eleven dummy variables representing month were also fitted to account for seasonal effects. Spatial and temporal effects were accounted for using Besag-York-Mollie (BYM) and autoregressive (1) models. Their levels of significance were analyzed using deviance information criterion (DIC). The model was validated based on the Theil's coefficient which compared predicted and observed incidence estimated using the validation data. Dengue incidence predictions for 2010-2012 were also used to generate risk maps.

RESULTS

The mean monthly dengue incidence during the period was 6.94 cases (SD 14.49) per 100,000 people. Analyses on the temporal trends of the disease showed regular seasonal epidemics that were interrupted every 3 years (specifically in July 2004, July 2007 and September 2010) by major fluctuations in incidence. Monthly mean minimum temperature, rainfall, area under urban settlement/build-up areas and altitude were significant in the final model. Minimum temperature and rainfall had non-linear effects and lagging them by two months provided a better fitting model compared to using unlagged variables. Forecasts for the validation period closely mirrored the observed data and accurately captured the troughs and peaks of dengue incidence trajectories. A favorable Theil's coefficient of inequality of 0.22 was generated.

CONCLUSIONS

The study identified temperature, rainfall, altitude and area under urban settlement as being significant predictors of dengue incidence. The statistical model fitted the data well based on Theil's coefficient of inequality, and risk maps generated from its predictions identified most of the high-risk provinces throughout the country.

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.

Forecasting Zoonotic Infectious Disease Response to Climate Change: Mosquito Vectors and a Changing Environment

Infectious diseases are changing due to the environment and altered interactions among hosts, reservoirs, vectors, and pathogens. This is particularly true for zoonotic diseases that infect humans, agricultural animals, and wildlife. Within the subset of zoonoses, vector-borne pathogens are changing more rapidly with climate change, and have a complex epidemiology, which may allow them to take advantage of a changing environment. Most mosquito-borne infectious diseases are transmitted by mosquitoes in three genera: Aedes, Anopheles, and Culex, and the expansion of these genera is well documented. There is an urgent need to study vector-borne diseases in response to climate change and to produce a generalizable approach capable of generating risk maps and forecasting outbreaks. Here, we provide a strategy for coupling climate and epidemiological models for zoonotic infectious diseases. We discuss the complexity and challenges of data and model fusion, baseline requirements for data, and animal and human population movement. Disease forecasting needs significant investment to build the infrastructure necessary to collect data about the environment, vectors, and hosts at all spatial and temporal resolutions. These investments can contribute to building a modeling community around the globe to support public health officials so as to reduce disease burden through forecasts with quantified uncertainty.

Factors determining dengue outbreak in Malaysia

A large scale study was conducted to elucidate the true relationship among entomological, epidemiological and environmental factors that contributed to dengue outbreak in Malaysia. Two large areas (Selayang and Bandar Baru Bangi) were selected in this study based on five consecutive years of high dengue cases. Entomological data were collected using ovitraps where the number of larvae was used to reflect Aedes mosquito population size; followed by RT-PCR screening to detect and serotype dengue virus in mosquitoes. Notified cases, date of disease onset, and number and type of the interventions were used as epidemiological endpoint, while rainfall, temperature, relative humidity and air pollution index (API) were indicators for environmental data. The field study was conducted during 81 weeks of data collection. Correlation and Autoregressive Distributed Lag Model were used to determine the relationship. The study showed that, notified cases were indirectly related with the environmental data, but shifted one week, i.e. last 3 weeks positive PCR; last 4 weeks rainfall; last 3 weeks maximum relative humidity; last 3 weeks minimum and maximum temperature; and last 4 weeks air pollution index (API), respectively. Notified cases were also related with next week intervention, while conventional intervention only happened 4 weeks after larvae were found, indicating ample time for dengue transmission. Based on a significant relationship among the three factors (epidemiological, entomological and environmental), estimated Autoregressive Distributed Lag (ADL) model for both locations produced high accuracy 84.9% for Selayang and 84.1% for Bandar Baru Bangi in predicting the actual notified cases. Hence, such model can be used in forestalling dengue outbreak and acts as an early warning system. The existence of relationships among the entomological, epidemiological and environmental factors can be used to build an early warning system for the prediction of dengue outbreak so that preventive interventions can be taken early to avert the outbreaks.

Developing a dengue forecast model using machine learning: A case study in China

BACKGROUND

In China, dengue remains an important public health issue with expanded areas and increased incidence recently. Accurate and timely forecasts of dengue incidence in China are still lacking. We aimed to use the state-of-the-art machine learning algorithms to develop an accurate predictive model of dengue.

METHODOLOGY/PRINCIPAL FINDINGS

Weekly dengue cases, Baidu search queries and climate factors (mean temperature, relative humidity and rainfall) during 2011-2014 in Guangdong were gathered. A dengue search index was constructed for developing the predictive models in combination with climate factors. The observed year and week were also included in the models to control for the long-term trend and seasonality. Several machine learning algorithms, including the support vector regression (SVR) algorithm, step-down linear regression model, gradient boosted regression tree algorithm (GBM), negative binomial regression model (NBM), least absolute shrinkage and selection operator (LASSO) linear regression model and generalized additive model (GAM), were used as candidate models to predict dengue incidence. Performance and goodness of fit of the models were assessed using the root-mean-square error (RMSE) and R-squared measures. The residuals of the models were examined using the autocorrelation and partial autocorrelation function analyses to check the validity of the models. The models were further validated using dengue surveillance data from five other provinces. The epidemics during the last 12 weeks and the peak of the 2014 large outbreak were accurately forecasted by the SVR model selected by a cross-validation technique. Moreover, the SVR model had the consistently smallest prediction error rates for tracking the dynamics of dengue and forecasting the outbreaks in other areas in China.

CONCLUSION AND SIGNIFICANCE

The proposed SVR model achieved a superior performance in comparison with other forecasting techniques assessed in this study. The findings can help the government and community respond early to dengue epidemics.

Our Warming Planet: Is the HIV-1-Infected Population in the Crosshairs

Global climate change exacerbated by human energy use threatens to have a profound impact on human health, including from infectious diseases. Particularly vulnerable populations include the immunocompromised, including persons with HIV infection. Global warming can be expected to increase the geographic range of pathogens such as Vibrio cholerae as well as vectors that transmit disease, including ticks and mosquitoes. Higher temperatures also contribute to increased pathogen and vector efficiency in spreading disease. Natural disasters due to climate change result in population displacement, increased population density, and living conditions conducive to the spread of infectious diseases. Political mobilization is crucial to implementing and enforcing policies for prudent energy use, reversing the drivers of global warming, and ensuring that we are prepared for the adverse health consequences of climate change. This article summarizes a presentation by Robert T. Schooley, MD, at the IAS-USA continuing education program held in Berkeley in May 2017.