Exploring fine-scale urban landscapes using satellite data to predict the distribution of Aedes mosquito breeding sites.

BACKGROUND: The spread of mosquito-transmitted diseases such as dengue is a major public health issue worldwide. The Aedes aegypti mosquito, a primary vector for dengue, thrives in urban environments and breeds mainly in artificial or natural water containers. While the relationship between urban landscapes and potential breeding sites remains poorly understood, such a knowledge could help mitigate the risks associated with these diseases. This study aimed to analyze the relationships between urban landscape characteristics and potential breeding site abundance and type in cities of French Guiana (South America), and to evaluate the potential of such variables to be used in predictive models. METHODS: We use Multifactorial Analysis to explore the relationship between urban landscape characteristics derived from very high resolution satellite imagery, and potential breeding sites recorded from in-situ surveys. We then applied Random Forest models with different sets of urban variables to predict the number of potential breeding sites where entomological data are not available. RESULTS: Landscape analyses applied to satellite images showed that urban types can be clearly identified using texture indices. The Multiple Factor Analysis helped identify variables related to the distribution of potential breeding sites, such as buildings class area, landscape shape index, building number, and the first component of texture indices. Models predicting the number of potential breeding sites using the entire dataset provided an R² of 0.90, possibly influenced by overfitting, but allowing the prediction over all the study sites. Predictions of potential breeding sites varied highly depending on their type, with better results on breeding sites types commonly found in urban landscapes, such as containers of less than 200 L, large volumes and barrels. The study also outlined the limitation offered by the entomological data, whose sampling was not specifically designed for this study. Model outputs could be used as input to a mosquito dynamics model when no accurate field data are available. CONCLUSION: This study offers a first use of routinely collected data on potential breeding sites in a research study. It highlights the potential benefits of including satellite-based characterizations of the urban environment to improve vector control strategies.

Temporal dynamic of malaria in a suburban area along the Niger River.

BACKGROUND: Even if rainfall and temperature are factors classically associated to malaria, little is known about other meteorological factors, their variability and combinations related to malaria, in association with river height variations. Furthermore, in suburban area, urbanization and growing population density should be assessed in relation to these environmental factors. The aim of this study was to assess the impact of combined environmental, meteorological and hydrological factors on malaria incidence through time in the context of urbanization. METHODS: Population observational data were prospectively collected. Clinical malaria was defined as the presence of parasites in addition to clinical symptoms. Meteorological and hydrological factors were measured daily. For each factors variation indices were estimated. Urbanization was yearly estimated assessing satellite imaging and field investigations. Principal component analysis was used for dimension reduction and factors combination. Lags between malaria incidences and the main components were assessed by cross-correlation functions. Generalized additive model was used to assess relative impact of different environmental components, taking into account lags, and modelling non-linear relationships. Change-point analysis was used to determine transmission periods within years. RESULTS: Malaria incidences were dominated by annual periodicity and varied through time without modification of the dynamic, with no impact of the urbanization. The main meteorological factor associated with malaria was a combination of evaporation, humidity and rainfall, with a lag of 3 months. The relationship between combined temperature factors showed a linear impact until reaching high temperatures limiting malaria incidence, with a lag 3.25 months. Height and variation of the river were related to malaria incidence (respectively 6 week lag and no lag). CONCLUSIONS: The study emphasizes no decreasing trend of malaria incidence despite accurate access to care and control strategies in accordance to international recommendations. Furthermore, no decreasing trend was showed despite the urbanization of the area. Malaria transmission remain increase 3 months after the beginning of the dry season. Addition to evaporation versus humidity/rainfall, nonlinear relationship for temperature and river height and variations have to be taken into account when implementing malaria control programmes.

Development of a spatial sampling protocol using GIS to measure health disparities in Bobo-Dioulasso, Burkina Faso, a medium-sized African city.

BACKGROUND: Many cities in developing countries experience an unplanned and rapid growth. Several studies have shown that the irregular urbanization and equipment of cities produce different health risks and uneven exposure to specific diseases. Consequently, health surveys within cities should be carried out at the micro-local scale and sampling methods should try to capture this urban diversity. METHODS: This article describes the methodology used to develop a multi-stage sampling protocol to select a population for a demographic survey that investigates health disparities in the medium-sized city of Bobo-Dioulasso, Burkina Faso. It is based on the characterization of Bobo-Dioulasso city typology by taking into account the city heterogeneity, as determined by analysis of the built environment and of the distribution of urban infrastructures, such as healthcare structures or even water fountains, by photo-interpretation of aerial photographs and satellite images. Principal component analysis and hierarchical ascendant classification were then used to generate the city typology. RESULTS: Five groups of spaces with specific profiles were identified according to a set of variables which could be considered as proxy indicators of health status. Within these five groups, four sub-spaces were randomly selected for the study. We were then able to survey 1045 households in all the selected sub-spaces. The pertinence of this approach is discussed regarding to classical sampling as random walk method for example. CONCLUSION: This urban space typology allowed to select a population living in areas representative of the uneven urbanization process, and to characterize its health status in regards to several indicators (nutritional status, communicable and non-communicable diseases, and anaemia). Although this method should be validated and compared with more established methods, it appears as an alternative in developing countries where geographic and population data are scarce.

Spatio-temporal analysis of malaria within a transmission season in Bandiagara, Mali.

BACKGROUND: Heterogeneous patterns of malaria transmission are thought to be driven by factors including host genetics, distance to mosquito breeding sites, housing construction, and socio-behavioural characteristics. Evaluation of local transmission epidemiology to characterize malaria risk is essential for planning malaria control and elimination programmes. The use of geographical information systems (GIS) techniques has been a major asset to this approach. To assess time and space distribution of malaria disease in Bandiagara, Mali, within a transmission season, data were used from an ongoing malaria incidence study that enrolled 300 participants aged under six years old". METHODS: Children's households were georeferenced using a handheld global position system. Clinical malaria was defined as a positive blood slide for Plasmodium falciparum asexual stages associated with at least one of the following signs: headache, body aches, fever, chills and weakness. Daily rainfall was measured at the local weather station.Landscape features of Bandiagara were obtained from satellite images and field survey. QGIS™ software was used to map malaria cases, affected and non-affected children, and the number of malaria episodes per child in each block of Bandiagara. Clusters of high or low risk were identified under SaTScan(®) software according to a Bernoulli model. RESULTS: From June 2009 to May 2010, 296 clinical malaria cases were recorded. Though clearly temporally related to the rains, Plasmodium falciparum occurrence persisted late in the dry season. Two "hot spots" of malaria transmission also found, notably along the Yamé River, characterized by higher than expected numbers of malaria cases, and high numbers of clinical episodes per child. Conversely, the north-eastern sector of the town had fewer cases despite its proximity to a large body of standing water which was mosquito habitat. CONCLUSION: These results confirm the existence of a marked spatial heterogeneity of malaria transmission in Bandiagara, providing support for implementation of targeted interventions.

Land cover, land use and malaria in the Amazon: a systematic literature review of studies using remotely sensed data.

The nine countries sharing the Amazon forest accounted for 89% of all malaria cases reported in the Americas in 2008. Remote sensing can help identify the environmental determinants of malaria transmission and their temporo-spatial evolution. Seventeen studies characterizing land cover or land use features, and relating them to malaria in the Amazon subregion, were identified. These were reviewed in order to improve the understanding of the land cover/use class roles in malaria transmission. The indicators affecting the transmission risk were summarized in terms of temporal components, landscape fragmentation and anthropic pressure. This review helps to define a framework for future studies aiming to characterize and monitor malaria.

Objective sampling design in a highly heterogeneous landscape - characterizing environmental determinants of malaria vector distribution in French Guiana, in the Amazonian region.

BACKGROUND: Sampling design is a key issue when establishing species inventories and characterizing habitats within highly heterogeneous landscapes. Sampling efforts in such environments may be constrained and many field studies only rely on subjective and/or qualitative approaches to design collection strategy. The region of Cacao, in French Guiana, provides an excellent study site to understand the presence and abundance of Anopheles mosquitoes, their species dynamics and the transmission risk of malaria across various environments. We propose an objective methodology to define a stratified sampling design. Following thorough environmental characterization, a factorial analysis of mixed groups allows the data to be reduced and non-collinear principal components to be identified while balancing the influences of the different environmental factors. Such components defined new variables which could then be used in a robust k-means clustering procedure. Then, we identified five clusters that corresponded to our sampling strata and selected sampling sites in each stratum. RESULTS: We validated our method by comparing the species overlap of entomological collections from selected sites and the environmental similarities of the same sites. The Morisita index was significantly correlated (Pearson linear correlation) with environmental similarity based on i) the balanced environmental variable groups considered jointly (p = 0.001) and ii) land cover/use (p-value << 0.001). The Jaccard index was significantly correlated with land cover/use-based environmental similarity (p-value = 0.001). CONCLUSIONS: The results validate our sampling approach. Land cover/use maps (based on high spatial resolution satellite images) were shown to be particularly useful when studying the presence, density and diversity of Anopheles mosquitoes at local scales and in very heterogeneous landscapes.

Urban Land Expansion from Scratch to Urban Agglomeration in the Federal District of Brazil in the Past 60 Years.

Empirical studies of urban expansion have increased rapidly in recent decades worldwide. Previous studies mainly focused on cities in China, the United States or African countries, with Brazilian cities receiving less attention. Moreover, such studies are rare in purpose-built cities. Taking the urban expansion from scratch (1960) to urban agglomeration (2015) in the Federal District of Brazil (FDB) as an example, this study aims to quantify the magnitude, patterns, modes, types and efficiency of urban land expansion and attempts to reveal some implications within sustainable urban expansion thinking. Annual expansion, landscape metrics, local Moran's I index, area weighted mean expansion index, and land-use efficiency were computed. The suitability of diffusion-coalescence theory and the impact of population growth and urban development policies on urban expansion were discussed. Urban land continuously expanded and became more fragmented during 1960-2015, which mainly occurred in SSW and WSW directions. Urban land evolved in a polycentric way. Edge expansion was identified as the stable contributor, and the importance of infilling and spontaneous growth alternated. Urban expansion in this region supported the diffusion-coalescence theory. Population growth promoted urban expansion, and the creation of peripheral urban nuclei and their development were associated with the urban expansion and the changes in urban land structure. This study adds new empirical evidence of urban expansion to Brazil urbanization, and compact urbanization, population control, and efficient urban land use should be considered in the future.

Sub-national tailoring of seasonal malaria chemoprevention in Mali based on malaria surveillance and rainfall data.

BACKGROUND: In malaria endemic countries, seasonal malaria chemoprevention (SMC) interventions are performed during the high malaria transmission in accordance with epidemiological surveillance data. In this study we propose a predictive approach for tailoring the timing and number of cycles of SMC in all health districts of Mali based on sub-national epidemiological surveillance and rainfall data. Our primary objective was to select the best of two approaches for predicting the onset of the high transmission season at the operational scale. Our secondary objective was to evaluate the number of malaria cases, hospitalisations and deaths in children under 5 years of age that would be prevented annually and the additional cost that would be incurred using the best approach. METHODS: For each of the 75 health districts of Mali over the study period (2014-2019), we determined (1) the onset of the rainy season period based on weekly rainfall data; (ii) the onset and duration of the high transmission season using change point analysis of weekly incidence data; and (iii) the lag between the onset of the rainy season and the onset of the high transmission. Two approaches for predicting the onset of the high transmission season in 2019 were evaluated. RESULTS: In the study period (2014-2019), the onset of the rainy season ranged from week (W) 17 (W17; April) to W34 (August). The onset of the high transmission season ranged from W25 (June) to W40 (September). The lag between these two events ranged from 5 to 12 weeks. The duration of the high transmission season ranged from 3 to 6 months. The best of the two approaches predicted the onset of the high transmission season in 2019 to be in June in two districts, in July in 46 districts, in August in 21 districts and in September in six districts. Using our proposed approach would prevent 43,819 cases, 1943 hospitalisations and 70 deaths in children under 5 years of age annually for a minimal additional cost. Our analysis shows that the number of cycles of SMC should be changed in 36 health districts. CONCLUSION: Adapting the timing of SMC interventions using our proposed approach could improve the prevention of malaria cases and decrease hospitalisations and deaths. Future studies should be conducted to validate this approach.

Semi-Supervised Text Classification Framework: An Overview of Dengue Landscape Factors and Satellite Earth Observation.

In recent years there has been an increasing use of satellite Earth observation (EO) data in dengue research, in particular the identification of landscape factors affecting dengue transmission. Summarizing landscape factors and satellite EO data sources, and making the information public are helpful for guiding future research and improving health decision-making. In this case, a review of the literature would appear to be an appropriate tool. However, this is not an easy-to-use tool. The review process mainly includes defining the topic, searching, screening at both title/abstract and full-text levels and data extraction that needs consistent knowledge from experts and is time-consuming and labor intensive. In this context, this study integrates the review process, text scoring, active learning (AL) mechanism, and bidirectional long short-term memory (BiLSTM) networks, and proposes a semi-supervised text classification framework that enables the efficient and accurate selection of the relevant articles. Specifically, text scoring and BiLSTM-based active learning were used to replace the title/abstract screening and full-text screening, respectively, which greatly reduces the human workload. In this study, 101 relevant articles were selected from 4 bibliographic databases, and a catalogue of essential dengue landscape factors was identified and divided into four categories: land use (LU), land cover (LC), topography and continuous land surface features. Moreover, various satellite EO sensors and products used for identifying landscape factors were tabulated. Finally, possible future directions of applying satellite EO data in dengue research in terms of landscape patterns, satellite sensors and deep learning were proposed. The proposed semi-supervised text classification framework was successfully applied in research evidence synthesis that could be easily applied to other topics, particularly in an interdisciplinary context.

Geo-Epidemiology of Malaria at the Health Area Level, Dire Health District, Mali, 2013-2017.

Background: According to the World Health Organization, there were more than 228 million cases of malaria globally in 2018, with 93% of cases occurring in Africa; in Mali, a 13% increase in the number of cases was observed between 2015 and 2018; this study aimed to evaluate the impact of meteorological and environmental factors on the geo-epidemiology of malaria in the health district of Dire, Mali. Methods: Meteorological and environmental variables were synthesized using principal component analysis and multiple correspondence analysis, the relationship between malaria incidence and synthetic indicators was determined using a multivariate general additive model; hotspots were detected by SaTScan. Results: Malaria incidence showed high inter and intra-annual variability; the period of high transmission lasted from September to February; health areas characterized by proximity to the river, propensity for flooding and high agricultural yield were the most at risk, with an incidence rate ratio of 2.21 with confidence intervals (95% CI: 1.85-2.58); malaria incidence in Dire declined from 120 to 20 cases per 10,000 person-weeks between 2013 and 2017. Conclusion: The identification of areas and periods of high transmission can help improve malaria control strategies.

Modelling malaria incidence with environmental dependency in a locality of Sudanese savannah area, Mali.

BACKGROUND: The risk of Plasmodium falciparum infection is variable over space and time and this variability is related to environmental variability. Environmental factors affect the biological cycle of both vector and parasite. Despite this strong relationship, environmental effects have rarely been included in malaria transmission models.Remote sensing data on environment were incorporated into a temporal model of the transmission, to forecast the evolution of malaria epidemiology, in a locality of Sudanese savannah area. METHODS: A dynamic cohort was constituted in June 1996 and followed up until June 2001 in the locality of Bancoumana, Mali. The 15-day composite vegetation index (NDVI), issued from satellite imagery series (NOAA) from July 1981 to December 2006, was used as remote sensing data.The statistical relationship between NDVI and incidence of P. falciparum infection was assessed by ARIMA analysis. ROC analysis provided an NDVI value for the prediction of an increase in incidence of parasitaemia.Malaria transmission was modelled using an SIRS-type model, adapted to Bancoumana's data. Environmental factors influenced vector mortality and aggressiveness, as well as length of the gonotrophic cycle. NDVI observations from 1981 to 2001 were used for the simulation of the extrinsic variable of a hidden Markov chain model. Observations from 2002 to 2006 served as external validation. RESULTS: The seasonal pattern of P. falciparum incidence was significantly explained by NDVI, with a delay of 15 days (p = 0.001). An NDVI threshold of 0.361 (p = 0.007) provided a Diagnostic Odd Ratio (DOR) of 2.64 (CI95% [1.26;5.52]).The deterministic transmission model, with stochastic environmental factor, predicted an endemo-epidemic pattern of malaria infection. The incidences of parasitaemia were adequately modelled, using the observed NDVI as well as the NDVI simulations. Transmission pattern have been modelled and observed values were adequately predicted. The error parameters have shown the smallest values for a monthly model of environmental changes. CONCLUSION: Remote-sensed data were coupled with field study data in order to drive a malaria transmission model. Several studies have shown that the NDVI presents significant correlations with climate variables, such as precipitations particularly in Sudanese savannah environments. Non-linear model combining environmental variables, predisposition factors and transmission pattern can be used for community level risk evaluation.

Wetlands and Malaria in the Amazon: Guidelines for the Use of Synthetic Aperture Radar Remote-Sensing.

The prevention and control of mosquito-borne diseases, such as malaria, are important health issues in tropical areas. Malaria transmission is a multi-scale process strongly controlled by environmental factors, and the use of remote-sensing data is suitable for the characterization of its spatial and temporal dynamics. Synthetic aperture radar (SAR) is well-adapted to tropical areas, since it is capable of imaging independent of light and weather conditions. In this study, we highlight the contribution of SAR sensors in the assessment of the relationship between vectors, malaria and the environment in the Amazon region. More specifically, we focus on the SAR-based characterization of potential breeding sites of mosquito larvae, such as man-made water collections and natural wetlands, providing guidelines for the use of SAR capabilities and techniques in order to optimize vector control and malaria surveillance. In light of these guidelines, we propose a framework for the production of spatialized indicators and malaria risk maps based on the combination of SAR, entomological and epidemiological data to support malaria risk prevention and control actions in the field.

Spatio-Temporal Dynamics of Asymptomatic Malaria: Bridging the Gap Between Annual Malaria Resurgences in a Sahelian Environment.

In areas of seasonal malaria transmission, the incidence rate of malaria infection is presumed to be near zero at the end of the dry season. Asymptomatic individuals may constitute a major parasite reservoir during this time. We conducted a longitudinal analysis of the spatio-temporal distribution of clinical malaria and asymptomatic parasitemia over time in a Malian town to highlight these malaria transmission dynamics. For a cohort of 300 rural children followed over 2009-2014, periodicity and phase shift between malaria and rainfall were determined by spectral analysis. Spatial risk clusters of clinical episodes or carriage were identified. A nested-case-control study was conducted to assess the parasite carriage factors. Malaria infection persisted over the entire year with seasonal peaks. High transmission periods began 2-3 months after the rains began. A cluster with a low risk of clinical malaria in the town center persisted in high and low transmission periods. Throughout 2009-2014, cluster locations did not vary from year to year. Asymptomatic and gametocyte carriage were persistent, even during low transmission periods. For high transmission periods, the ratio of asymptomatic to clinical cases was approximately 0.5, but was five times higher during low transmission periods. Clinical episodes at previous high transmission periods were a protective factor for asymptomatic carriage, but carrying parasites without symptoms at a previous high transmission period was a risk factor for asymptomatic carriage. Stable malaria transmission was associated with sustained asymptomatic carriage during dry seasons. Control strategies should target persistent low-level parasitemia clusters to interrupt transmission.

Dynamical Mapping of Anopheles darlingi Densities in a Residual Malaria Transmission Area of French Guiana by Using Remote Sensing and Meteorological Data.

Local variation in the density of Anopheles mosquitoes and the risk of exposure to bites are essential to explain the spatial and temporal heterogeneities in the transmission of malaria. Vector distribution is driven by environmental factors. Based on variables derived from satellite imagery and meteorological observations, this study aimed to dynamically model and map the densities of Anopheles darlingi in the municipality of Saint-Georges de l'Oyapock (French Guiana). Longitudinal sampling sessions of An. darlingi densities were conducted between September 2012 and October 2014. Landscape and meteorological data were collected and processed to extract a panel of variables that were potentially related to An. darlingi ecology. Based on these data, a robust methodology was formed to estimate a statistical predictive model of the spatial-temporal variations in the densities of An. darlingi in Saint-Georges de l'Oyapock. The final cross-validated model integrated two landscape variables-dense forest surface and built surface-together with four meteorological variables related to rainfall, evapotranspiration, and the minimal and maximal temperatures. Extrapolation of the model allowed the generation of predictive weekly maps of An. darlingi densities at a resolution of 10-m. Our results supported the use of satellite imagery and meteorological data to predict malaria vector densities. Such fine-scale modeling approach might be a useful tool for health authorities to plan control strategies and social communication in a cost-effective, targeted, and timely manner.