Spatial and spatio-temporal methods for mapping malaria risk: a systematic review

Malaria risk mapping among children under five in Togo

Malaria is a major health threat in sub-Sahara Africa, especially for children under five. However, there is considerable heterogeneity between areas in malaria risk reported, associated with environmental and climatic. We used data from Togo to explore spatial patterns of malaria incidence. Geospatial covariate datasets, including climatic and environmental variables from the 2017 Malaria Indicator Survey in Togo, were used for this study. The association between malaria incidence and ecological predictors was assessed using three regression techniques, namely the Ordinary Least Squares (OLS), spatial lag model (SLM), and spatial error model (SEM). A total of 171 clusters were included in the survey and provided data on environmental and climate variables. Spatial autocorrelation showed that the distribution of malaria incidence was not random and revealed significant spatial clustering. Mean temperature, precipitation, aridity and proximity to water bodies showed a significant and direct association with malaria incidence rate in the SLM model, which best fitted the data according to AIC. Five malaria incidence hotspots were identified. Malaria incidence is spatially clustered in Togo associated with climatic and environmental factors. The results can contribute to the development of specific malaria control plans taking geographical variation into consideration and targeting transmission hotspots.

Application of advanced very high-resolution radiometer (AVHRR)-based vegetation health indices for modelling and predicting malaria in Northern Benin, West Africa

BACKGROUND

Vegetation health (VH) is a powerful characteristic for forecasting malaria incidence in regions where the disease is prevalent. This study aims to determine how vegetation health affects the prevalence of malaria and create seasonal weather forecasts using NOAA/AVHRR environmental satellite data that can be substituted for malaria epidemic forecasts.

METHODS

Weekly advanced very high-resolution radiometer (AVHRR) data were retrieved from the NOAA satellite website from 2009 to 2021. The monthly number of malaria cases was collected from the Ministry of Health of Benin from 2009 to 2021 and matched with AVHRR data. Pearson correlation was calculated to investigate the impact of vegetation health on malaria transmission. Ordinary least squares (OLS), support vector machine (SVM) and principal component regression (PCR) were applied to forecast the monthly number of cases of malaria in Northern Benin. A random sample of proposed models was used to assess accuracy and bias.

RESULTS

Estimates place the annual percentage rise in malaria cases at 9.07% over 2009-2021 period. Moisture (VCI) for weeks 19-21 predicts 75% of the number of malaria cases in the month of the start of high mosquito activities. Soil temperature (TCI) and vegetation health index (VHI) predicted one month earlier than the start of mosquito activities through transmission, 78% of monthly malaria incidence.

CONCLUSIONS

SVM model D is more effective than OLS model A in the prediction of malaria incidence in Northern Benin. These models are a very useful tool for stakeholders looking to lessen the impact of malaria in Benin.

Comparison of new computational methods for spatial modelling of malaria

BACKGROUND

Geostatistical analysis of health data is increasingly used to model spatial variation in malaria prevalence, burden, and other metrics. Traditional inference methods for geostatistical modelling are notoriously computationally intensive, motivating the development of newer, approximate methods for geostatistical analysis or, more broadly, computational modelling of spatial processes. The appeal of faster methods is particularly great as the size of the region and number of spatial locations being modelled increases.

METHODS

This work presents an applied comparison of four proposed 'fast' computational methods for spatial modelling and the software provided to implement them-Integrated Nested Laplace Approximation (INLA), tree boosting with Gaussian processes and mixed effect models (GPBoost), Fixed Rank Kriging (FRK) and Spatial Random Forests (SpRF). The four methods are illustrated by estimating malaria prevalence on two different spatial scales-country and continent. The performance of the four methods is compared on these data in terms of accuracy, computation time, and ease of implementation.

RESULTS

Two of these methods-SpRF and GPBoost-do not scale well as the data size increases, and so are likely to be infeasible for larger-scale analysis problems. The two remaining methods-INLA and FRK-do scale well computationally, however the resulting model fits are very sensitive to the user's modelling assumptions and parameter choices. The binomial observation distribution commonly used for disease prevalence mapping with INLA fails to account for small-scale overdispersion present in the malaria prevalence data, which can lead to poor predictions. Selection of an appropriate alternative such as the Beta-binomial distribution is required to produce a reliable model fit. The small-scale random effect term in FRK overcomes this pitfall, but FRK model estimates are very reliant on providing a sufficient number and appropriate configuration of basis functions. Unfortunately the computation time for FRK increases rapidly with increasing basis resolution.

CONCLUSIONS

INLA and FRK both enable scalable geostatistical modelling of malaria prevalence data. However care must be taken when using both methods to assess the fit of the model to data and plausibility of predictions, in order to select appropriate model assumptions and parameters.

Spatio-temporal modelling of routine health facility data for malaria risk micro-stratification in mainland Tanzania

As malaria transmission declines, the need to monitor the heterogeneity of malaria risk at finer scales becomes critical to guide community-based targeted interventions. Although routine health facility (HF) data can provide epidemiological evidence at high spatial and temporal resolution, its incomplete nature of information can result in lower administrative units without empirical data. To overcome geographic sparsity of data and its representativeness, geo-spatial models can leverage routine information to predict risk in un-represented areas as well as estimate uncertainty of predictions. Here, a Bayesian spatio-temporal model was applied on malaria test positivity rate (TPR) data for the period 2017-2019 to predict risks at the ward level, the lowest decision-making unit in mainland Tanzania. To quantify the associated uncertainty, the probability of malaria TPR exceeding programmatic threshold was estimated. Results showed a marked spatial heterogeneity in malaria TPR across wards. 17.7 million people resided in areas where malaria TPR was high (≥ 30; 90% certainty) in the North-West and South-East parts of Tanzania. Approximately 11.7 million people lived in areas where malaria TPR was very low (< 5%; 90% certainty). HF data can be used to identify different epidemiological strata and guide malaria interventions at micro-planning units in Tanzania. These data, however, are imperfect in many settings in Africa and often require application of geo-spatial modelling techniques for estimation.

Mapping malaria incidence using routine health facility surveillance data in Uganda

INTRODUCTION

Maps of malaria risk are important tools for allocating resources and tracking progress. Most maps rely on cross-sectional surveys of parasite prevalence, but health facilities represent an underused and powerful data source. We aimed to model and map malaria incidence using health facility data in Uganda.

METHODS

Using 24 months (2019-2020) of individual-level outpatient data collected from 74 surveillance health facilities located in 41 districts across Uganda (n=445 648 laboratory-confirmed cases), we estimated monthly malaria incidence for parishes within facility catchment areas (n=310) by estimating care-seeking population denominators. We fit spatio-temporal models to the incidence estimates to predict incidence rates for the rest of Uganda, informed by environmental, sociodemographic and intervention variables. We mapped estimated malaria incidence and its uncertainty at the parish level and compared estimates to other metrics of malaria. To quantify the impact that indoor residual spraying (IRS) may have had, we modelled counterfactual scenarios of malaria incidence in the absence of IRS.

RESULTS

Over 4567 parish-months, malaria incidence averaged 705 cases per 1000 person-years. Maps indicated high burden in the north and northeast of Uganda, with lower incidence in the districts receiving IRS. District-level estimates of cases correlated with cases reported by the Ministry of Health (Spearman's r=0.68, p<0.0001), but were considerably higher (40 166 418 cases estimated compared with 27 707 794 cases reported), indicating the potential for underreporting by the routine surveillance system. Modelling of counterfactual scenarios suggest that approximately 6.2 million cases were averted due to IRS across the study period in the 14 districts receiving IRS (estimated population 8 381 223).

CONCLUSION

Outpatient information routinely collected by health systems can be a valuable source of data for mapping malaria burden. National Malaria Control Programmes may consider investing in robust surveillance systems within public health facilities as a low-cost, high benefit tool to identify vulnerable regions and track the impact of interventions.

Fine-scale mapping of urban malaria exposure under data scarcity: an approach centred on vector ecology

BACKGROUND

Although malaria transmission has experienced an overall decline in sub-Saharan Africa, urban malaria is now considered an emerging health issue due to rapid and uncontrolled urbanization and the adaptation of vectors to urban environments. Fine-scale hazard and exposure maps are required to support evidence-based policies and targeted interventions, but data-driven predictive spatial modelling is hindered by gaps in epidemiological and entomological data. A knowledge-based geospatial framework is proposed for mapping the heterogeneity of urban malaria hazard and exposure under data scarcity. It builds on proven geospatial methods, implements open-source algorithms, and relies heavily on vector ecology knowledge and the involvement of local experts.

METHODS

A workflow for producing fine-scale maps was systematized, and most processing steps were automated. The method was evaluated through its application to the metropolitan area of Dakar, Senegal, where urban transmission has long been confirmed. Urban malaria exposure was defined as the contact risk between adult Anopheles vectors (the hazard) and urban population and accounted for socioeconomic vulnerability by including the dimension of urban deprivation that is reflected in the morphology of the built-up fabric. Larval habitat suitability was mapped through a deductive geospatial approach involving the participation of experts with a strong background in vector ecology and validated with existing geolocated entomological data. Adult vector habitat suitability was derived through a similar process, based on dispersal from suitable breeding site locations. The resulting hazard map was combined with a population density map to generate a gridded urban malaria exposure map at a spatial resolution of 100 m.

RESULTS

The identification of key criteria influencing vector habitat suitability, their translation into geospatial layers, and the assessment of their relative importance are major outcomes of the study that can serve as a basis for replication in other sub-Saharan African cities. Quantitative validation of the larval habitat suitability map demonstrates the reliable performance of the deductive approach, and the added value of including local vector ecology experts in the process. The patterns displayed in the hazard and exposure maps reflect the high degree of heterogeneity that exists throughout the city of Dakar and its suburbs, due not only to the influence of environmental factors, but also to urban deprivation.

CONCLUSIONS

This study is an effort to bring geospatial research output closer to effective support tools for local stakeholders and decision makers. Its major contributions are the identification of a broad set of criteria related to vector ecology and the systematization of the workflow for producing fine-scale maps. In a context of epidemiological and entomological data scarcity, vector ecology knowledge is key for mapping urban malaria exposure. An application of the framework to Dakar showed its potential in this regard. Fine-grained heterogeneity was revealed by the output maps, and besides the influence of environmental factors, the strong links between urban malaria and deprivation were also highlighted.

Spatial and spatio-temporal epidemiological approaches to inform COVID-19 surveillance and control: a systematic review of statistical and modelling methods in Africa

OBJECTIVE

Various studies have been published to better understand the underlying spatial and temporal dynamics of COVID-19. This review sought to identify different spatial and spatio-temporal modelling methods that have been applied to COVID-19 and examine influential covariates that have been reportedly associated with its risk in Africa.

DESIGN

Systematic review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

DATA SOURCES

Thematically mined keywords were used to identify refereed studies conducted between January 2020 and February 2022 from the following databases: PubMed, Scopus, MEDLINE via Proquest, CINHAL via EBSCOhost and Coronavirus Research Database via ProQuest. A manual search through the reference list of studies was also conducted.

ELIGIBILITY CRITERIA FOR SELECTING STUDIES

Peer-reviewed studies that demonstrated the application of spatial and temporal approaches to COVID-19 outcomes.

DATA EXTRACTION AND SYNTHESIS

A standardised extraction form based on critical appraisal and data extraction for systematic reviews of prediction modelling studies checklist was used to extract the meta-data of the included studies. A validated scoring criterion was used to assess studies based on their methodological relevance and quality.

RESULTS

Among 2065 hits in five databases, title and abstract screening yielded 827 studies of which 22 were synthesised and qualitatively analysed. The most common socioeconomic variable was population density. HIV prevalence was the most common epidemiological indicator, while temperature was the most common environmental indicator. Thirteen studies (59%) implemented diverse formulations of spatial and spatio-temporal models incorporating unmeasured factors of COVID-19 and the subtle influence of time and space. Cluster analyses were used across seven studies (32%) to explore COVID-19 variation and determine whether observed patterns were random.

CONCLUSION

COVID-19 modelling in Africa is still in its infancy, and a range of spatial and spatio-temporal methods have been employed across diverse settings. Strengthening routine data systems remains critical for generating estimates and understanding factors that drive spatial variation in vulnerable populations and temporal variation in pandemic progression.

PROSPERO REGISTRATION NUMBER

CRD42021279767.

Malaria transmission in Nepal under climate change: anticipated shifts in extent and season, and comparison with risk definitions for intervention

BACKGROUND

Climate and climate change affect the spatial pattern and seasonality of malaria risk. Season lengths and spatial extents of mapped current and future malaria transmission suitability predictions for Nepal were assessed for a combination of malaria vector and parasites: Anopheles stephensi and Plasmodium falciparum (ASPF) and An. stephensi and Plasmodium vivax (ASPV) and compared with observed estimates of malaria risk in Nepal.

METHODS

Thermal bounds of malaria transmission suitability for baseline (1960-1990) and future climate projections (RCP 4.5 and RCP 8.5 in 2030 and 2050) were extracted from global climate models and mapped for Nepal. Season length and spatial extent of suitability between baseline and future climate scenarios for ASPF and ASPV were compared using the Warren's I metric. Official 2010 DoHS risk districts (DRDs) and 2021 DoHS risk wards (DRWs), and spatiotemporal incidence trend clusters (ITCs) were overlaid on suitability season length and extent maps to assess agreement, and potential mismatches.

RESULTS

Shifts in season length and extent of malaria transmission suitability in Nepal are anticipated under both RCP 4.5 and RCP 8.5 scenarios in 2030 and 2050, compared to baseline climate. The changes are broadly consistent across both future climate scenarios for ASPF and ASPV. There will be emergence of suitability and increasing length of season for both ASPF and ASPV and decreasing length of season for ASPV by 2050. The emergence of suitability will occur in low and no-risk DRDs and outside of high and moderate-risk DRWs, season length increase will occur across all DRD categories, and outside of high and moderate-risk DRWs. The high and moderate risk DRWs of 2021 fall into ITCs with decreasing trend.

CONCLUSIONS

The study identified areas of Nepal where malaria transmission suitability will emerge, disappear, increase, and decrease in the future. However, most of these areas are anticipated outside of the government's current and previously designated high and moderate-risk areas, and thus outside the focus of vector control interventions. Public health officials could use these anticipated changing areas of malaria risk to inform vector control interventions for eliminating malaria from the country, and to prevent malaria resurgence.

Mapping under-five child malaria risk that accounts for environmental and climatic factors to aid malaria preventive and control efforts in Ghana: Bayesian geospatial and interactive web-based mapping methods

BACKGROUND

Under-five child malaria is one of the leading causes of morbidity and mortality globally, especially among sub-Saharan African countries like Ghana. In Ghana, malaria is responsible for about 20,000 deaths in children annually of which 25% are those aged < 5 years. To provide opportunities for efficient malaria surveillance and targeted control efforts amidst limited public health resources, the study produced high resolution interactive web-based spatial maps that characterized geographical differences in malaria risk and identified high burden communities.

METHODS

This modelling and web-based mapping study utilized data from the 2019 Malaria Indicators Survey (MIS) of the Demographic and Health Survey Program. A novel and advanced Bayesian geospatial modelling and mapping approaches were utilized to examine predictors and geographical differences in under-five malaria. The model was validated via a cross-validation approach. The study produced an interactive web-based visualization map of the malaria risk by mapping the predicted malaria prevalence at both sampled and unsampled locations.

RESULTS

In 2019, 718 (25%) of 2867 under-five children surveyed had malaria. Substantial geographical differences in under-five malaria risk were observed. ITN coverage (log-odds 4.5643, 95% credible interval = 2.4086-6.8874), travel time (log-odds 0.0057, 95% credible interval = 0.0017-0.0099) and aridity (log-odds = 0.0600, credible interval = 0.0079-0.1167) were predictive of under-five malaria in the spatial model. The overall predicted national malaria prevalence was 16.3% (standard error (SE) 8.9%) with a range of 0.7% to 51.4% in the spatial model with covariates and prevalence of 28.0% (SE 13.9%) with a range of 2.4 to 67.2% in the spatial model without covariates. Residing in parts of Central and Bono East regions was associated with the highest risk of under-five malaria after adjusting for the selected covariates.

CONCLUSION

The high-resolution interactive web-based predictive maps can be used as an effective tool in the identification of communities that require urgent and targeted interventions by programme managers and implementers. This is key as part of an overall strategy in reducing the under-five malaria burden and its associated morbidity and mortality in a country with limited public health resources where universal intervention is practically impossible.

The use of routine health facility data for micro-stratification of malaria risk in mainland Tanzania

Background

Current efforts to estimate the spatially diverse malaria burden in malaria-endemic countries largely involve the use of epidemiological modelling methods for describing temporal and spatial heterogeneity using sparse interpolated prevalence data from periodic cross-sectional surveys. However, more malaria-endemic countries are beginning to consider local routine data for this purpose. Nevertheless, routine information from health facilities (HFs) remains widely under-utilized despite improved data quality, including increased access to diagnostic testing and the adoption of the electronic District Health Information System (DHIS2). This paper describes the process undertaken in mainland Tanzania using routine data to develop a high-resolution, micro-stratification risk map to guide future malaria control efforts.

Methods

Combinations of various routine malariometric indicators collected from 7098 HFs were assembled across 3065 wards of mainland Tanzania for the period 2017–2019. The reported council-level prevalence classification in school children aged 5–16 years (PfPR5–16) was used as a benchmark to define four malaria risk groups. These groups were subsequently used to derive cut-offs for the routine indicators by minimizing misclassifications and maximizing overall agreement. The derived-cutoffs were converted into numbered scores and summed across the three indicators to allocate wards into their overall risk stratum.

Results

Of 3065 wards, 353 were assigned to the very low strata (10.5% of the total ward population), 717 to the low strata (28.6% of the population), 525 to the moderate strata (16.2% of the population), and 1470 to the high strata (39.8% of the population). The resulting micro-stratification revealed malaria risk heterogeneity within 80 councils and identified wards that would benefit from community-level focal interventions, such as community-case management, indoor residual spraying and larviciding.

Conclusion

The micro-stratification approach employed is simple and pragmatic, with potential to be easily adopted by the malaria programme in Tanzania. It makes use of available routine data that are rich in spatial resolution and that can be readily accessed allowing for a stratification of malaria risk below the council level. Such a framework is optimal for supporting evidence-based, decentralized malaria control planning, thereby improving the effectiveness and allocation efficiency of malaria control interventions.

Small-scale spatiotemporal epidemiology of notifiable infectious diseases in China: a systematic review

Background

The prevalence of infectious diseases remains one of the major challenges faced by the Chinese health sector. Policymakers have a tremendous interest in investigating the spatiotemporal epidemiology of infectious diseases. We aimed to review the small-scale (city level, county level, or below) spatiotemporal epidemiology of notifiable infectious diseases in China through a systematic review, thus summarizing the evidence to facilitate more effective prevention and control of the diseases.

Methods

We searched four English language databases (PubMed, EMBASE, Cochrane Library, and Web of Science) and three Chinese databases (CNKI, WanFang, and SinoMed), for studies published between January 1, 2004 (the year in which China’s Internet-based disease reporting system was established) and December 31, 2021. Eligible works were small-scale spatial or spatiotemporal studies focusing on at least one notifiable infectious disease, with the entire territory of mainland China as the study area. Two independent reviewers completed the review process based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Results

A total of 18,195 articles were identified, with 71 eligible for inclusion, focusing on 22 diseases. Thirty-one studies (43.66%) were analyzed using city-level data, 34 (47.89%) were analyzed using county-level data, and six (8.45%) used community or individual data. Approximately four-fifths (80.28%) of the studies visualized incidence using rate maps. Of these, 76.06% employed various spatial clustering methods to explore the spatial variations in the burden, with Moran’s I statistic being the most common. Of the studies, 40.85% explored risk factors, in which the geographically weighted regression model was the most commonly used method. Climate, socioeconomic factors, and population density were the three most considered factors.

Conclusions

Small-scale spatiotemporal epidemiology has been applied in studies on notifiable infectious diseases in China, involving spatiotemporal distribution and risk factors. Health authorities should improve prevention strategies and clarify the direction of future work in the field of infectious disease research in China.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-022-07669-9.

Spatial Analysis of Mosquito-Borne Diseases in Europe: A Scoping Review

Mosquito-borne infections are increasing in endemic areas and previously unaffected regions. In 2020, the notification rate for Dengue was 0.5 cases per 100,000 population, and for Chikungunya <0.1/100,000. In 2019, the rate for Malaria was 1.3/100,000, and for West Nile Virus, 0.1/100,000. Spatial analysis is increasingly used in surveillance and epidemiological investigation, but reviews about their use in this research topic are scarce. We identify and describe the methodological approaches used to investigate the distribution and ecological determinants of mosquito-borne infections in Europe. Relevant literature was extracted from PubMed, Scopus, and Web of Science from inception until October 2021 and analysed according to PRISMA-ScR protocol. We identified 110 studies. Most used geographical correlation analysis (n = 50), mainly applying generalised linear models, and the remaining used spatial cluster detection (n = 30) and disease mapping (n = 30), mainly conducted using frequentist approaches. The most studied infections were Dengue (n = 32), Malaria (n = 26), Chikungunya (n = 26), and West Nile Virus (n = 24), and the most studied ecological determinants were temperature (n = 39), precipitation (n = 24), water bodies (n = 14), and vegetation (n = 11). Results from this review may support public health programs for mosquito-borne disease prevention and may help guide future research, as we recommended various good practices for spatial epidemiological studies.

Hotspots and super-spreaders: Modelling fine-scale malaria parasite transmission using mosquito flight behaviour

Malaria hotspots have been the focus of public health managers for several years due to the potential elimination gains that can be obtained from targeting them. The identification of hotspots must be accompanied by the description of the overall network of stable and unstable hotspots of malaria, especially in medium and low transmission settings where malaria elimination is targeted. Targeting hotspots with malaria control interventions has, so far, not produced expected benefits. In this work we have employed a mechanistic-stochastic algorithm to identify clusters of super-spreader houses and their related stable hotspots by accounting for mosquito flight capabilities and the spatial configuration of malaria infections at the house level. Our results show that the number of super-spreading houses and hotspots is dependent on the spatial configuration of the villages. In addition, super-spreaders are also associated to house characteristics such as livestock and family composition. We found that most of the transmission is associated with winds between 6pm and 10pm although later hours are also important. Mixed mosquito flight (downwind and upwind both with random components) were the most likely movements causing the spread of malaria in two out of the three study areas. Finally, our algorithm (named MALSWOTS) provided an estimate of the speed of malaria infection progression from house to house which was around 200-400 meters per day, a figure coherent with mark-release-recapture studies of Anopheles dispersion. Cross validation using an out-of-sample procedure showed accurate identification of hotspots. Our findings provide a significant contribution towards the identification and development of optimal tools for efficient and effective spatio-temporal targeted malaria interventions over potential hotspot areas.

Spatial pattern and predictors of malaria in Ethiopia: Application of auto logistics regression

Introduction

Malaria is a severe health threat in the World, mainly in Africa. It is the major cause of health problems in which the risk of morbidity and mortality associated with malaria cases are characterized by spatial variations across the county. This study aimed to investigate the spatial patterns and predictors of malaria distribution in Ethiopia.

Methods

A weighted sample of 15,239 individuals with rapid diagnosis test obtained from the Central Statistical Agency and Ethiopia malaria indicator survey of 2015. Global Moran’s I and Moran scatter plots were used in determining the distribution of malaria cases, whereas the local Moran’s I statistic was used in identifying exposed areas. The auto logistics spatial binary regression model was used to investigate the predictors of malaria.

Results

The final auto logistics regression model was reported that male clients had a positive significant effect on malaria cases as compared to female clients [AOR = 2.401, 95% CI: (2.125–2.713) ]. The distribution of malaria across the regions was different. The highest incidence of malaria was found in Gambela [AOR = 52.55, 95%CI: (40.54–68.12)] followed by Beneshangul [AOR = 34.95, 95%CI: (27.159–44.963)]. Similarly, individuals in Amhara [AOR = 0.243, 95% CI:(0.195–0.303], Oromiya [AOR = 0.197, 955 CI: (0.158–0.244)], Dire Dawa [AOR = 0.064, 95%CI(0.049–0.082)], Addis Ababa[AOR = 0.057,95%CI:(0.044–0.075)], Somali[AOR = 0.077,95%CI:(0.059–0.097)], SNNPR[OR = 0.329, 95%CI: (0.261–0.413)] and Harari [AOR = 0.256, 95%CI:(0.201–0.325)] were less likely to had low incidence of malaria as compared with Tigray. Furthermore, for one meter increase in altitude, the odds of positive rapid diagnostic test (RDT) decreases by 1.6% [AOR = 0.984, 95% CI: (0.984–0.984)]. The use of a shared toilet facility was found as a protective factor for malaria in Ethiopia [AOR = 1.671, 95% CI: (1.504–1.854)]. The spatial autocorrelation variable changes the constant from AOR = 0.471 for logistic regression to AOR = 0.164 for auto logistics regression.

Conclusions

This study found that the incidence of malaria in Ethiopia had a spatial pattern which is associated with socio-economic, demographic, and geographic risk factors. Spatial clustering of malaria cases had occurred in all regions, and the risk of clustering was different across the regions. The risk of malaria was found to be higher for those who live in soil floor-type houses as compared to those who lived in cement or ceramics floor type. Similarly, households with thatched, metal and thin, and other roof-type houses have a higher risk of malaria than ceramics tiles roof houses. Moreover, using a protected anti-mosquito net was reducing the risk of malaria incidence.

Epidemiological characteristics of mumps from 2004 to 2020 in Jiangsu, China: a flexible spatial and spatiotemporal analysis

The mumps resurgence has frequently been reported around the world in recent years, especially in many counties mumps vaccines have been widely used. This study aimed to describe the spatial epidemiological characteristics of mumps in Jiangsu, and provide a scientific basis for the implementation and adjustment of strategies to prevent and control mumps. The epidemiological characteristics were described with ratio or proportion. Spatial autocorrelation, Tango's flexible spatial scan statistics, and Kulldorff's elliptic spatiotemporal scan statistics were applied to identify the spatial autocorrelation, detect hot and cold spots of mumps incidence, and aggregation areas. A total of 172 775 cases were reported from 2004 to 2020 in Jiangsu. The general trend of mumps incidence is declining with a bimodal seasonal distribution identified mainly in summer and winter, respectively. Children aged 5–10 years old are the main risk group. A migration trend of hot spots from southeast to northwest over time was found. Similar high-risk aggregations were detected in the northwestern parts through spatial-temporal analysis with the most likely cluster time frame around 2019. Local medical and health administrations should formulate and implement targeted health care policies and allocate health resources more appropriately corresponding to the epidemiological characteristics of mumps.

Plasmodium falciparum parasite prevalence in East Africa: Updating data for malaria stratification

The High Burden High Impact (HBHI) strategy for malaria encourages countries to use multiple sources of available data to define the sub-national vulnerabilities to malaria risk, including parasite prevalence. Here, a modelled estimate of Plasmodium falciparum from an updated assembly of community parasite survey data in Kenya, mainland Tanzania, and Uganda is presented and used to provide a more contemporary understanding of the sub-national malaria prevalence stratification across the sub-region for 2019. Malaria prevalence data from surveys undertaken between January 2010 and June 2020 were assembled form each of the three countries. Bayesian spatiotemporal model-based approaches were used to interpolate space-time data at fine spatial resolution adjusting for population, environmental and ecological covariates across the three countries. A total of 18,940 time-space age-standardised and microscopy-converted surveys were assembled of which 14,170 (74.8%) were identified after 2017. The estimated national population-adjusted posterior mean parasite prevalence was 4.7% (95% Bayesian Credible Interval 2.6-36.9) in Kenya, 10.6% (3.4-39.2) in mainland Tanzania, and 9.5% (4.0-48.3) in Uganda. In 2019, more than 12.7 million people resided in communities where parasite prevalence was predicted ≥ 30%, including 6.4%, 12.1% and 6.3% of Kenya, mainland Tanzania and Uganda populations, respectively. Conversely, areas that supported very low parasite prevalence (<1%) were inhabited by approximately 46.2 million people across the sub-region, or 52.2%, 26.7% and 10.4% of Kenya, mainland Tanzania and Uganda populations, respectively. In conclusion, parasite prevalence represents one of several data metrics for disease stratification at national and sub-national levels. To increase the use of this metric for decision making, there is a need to integrate other data layers on mortality related to malaria, malaria vector composition, insecticide resistance and bionomic, malaria care-seeking behaviour and current levels of unmet need of malaria interventions.

Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict

This paper provides statistical guidance on the development and application of model-based geostatistical methods for disease prevalence mapping. We illustrate the different stages of the analysis, from exploratory analysis to spatial prediction of prevalence, through a case study on malaria mapping in Tanzania. Throughout the paper, we distinguish between predictive modelling, whose main focus is on maximizing the predictive accuracy of the model, and explanatory modelling, where greater emphasis is placed on understanding the relationships between the health outcome and risk factors. We demonstrate that these two paradigms can result in different modelling choices. We also propose a simple approach for detecting over-fitting based on inspection of the correlation matrix of the estimators of the regression coefficients. To enhance the interpretability of geostatistical models, we introduce the concept of domain effects in order to assist variable selection and model validation. The statistical ideas and principles illustrated here in the specific context of disease prevalence mapping are more widely applicable to any regression model for the analysis of epidemiological outcomes but are particularly relevant to geostatistical models, for which the separation between fixed and random effects can be ambiguous.

Estimating the local spatio-temporal distribution of malaria from routine health information systems in areas of low health care access and reporting

BACKGROUND

Reliable surveillance systems are essential for identifying disease outbreaks and allocating resources to ensure universal access to diagnostics and treatment for endemic diseases. Yet, most countries with high disease burdens rely entirely on facility-based passive surveillance systems, which miss the vast majority of cases in rural settings with low access to health care. This is especially true for malaria, for which the World Health Organization estimates that routine surveillance detects only 14% of global cases. The goal of this study was to develop a novel method to obtain accurate estimates of disease spatio-temporal incidence at very local scales from routine passive surveillance, less biased by populations' financial and geographic access to care.

METHODS

We use a geographically explicit dataset with residences of the 73,022 malaria cases confirmed at health centers in the Ifanadiana District in Madagascar from 2014 to 2017. Malaria incidence was adjusted to account for underreporting due to stock-outs of rapid diagnostic tests and variable access to healthcare. A benchmark multiplier was combined with a health care utilization index obtained from statistical models of non-malaria patients. Variations to the multiplier and several strategies for pooling neighboring communities together were explored to allow for fine-tuning of the final estimates. Separate analyses were carried out for individuals of all ages and for children under five. Cross-validation criteria were developed based on overall incidence, trends in financial and geographical access to health care, and consistency with geographic distribution in a district-representative cohort. The most plausible sets of estimates were then identified based on these criteria.

RESULTS

Passive surveillance was estimated to have missed about 4 in every 5 malaria cases among all individuals and 2 out of every 3 cases among children under five. Adjusted malaria estimates were less biased by differences in populations' financial and geographic access to care. Average adjusted monthly malaria incidence was nearly four times higher during the high transmission season than during the low transmission season. By gathering patient-level data and removing systematic biases in the dataset, the spatial resolution of passive malaria surveillance was improved over ten-fold. Geographic distribution in the adjusted dataset revealed high transmission clusters in low elevation areas in the northeast and southeast of the district that were stable across seasons and transmission years.

CONCLUSIONS

Understanding local disease dynamics from routine passive surveillance data can be a key step towards achieving universal access to diagnostics and treatment. Methods presented here could be scaled-up thanks to the increasing availability of e-health disease surveillance platforms for malaria and other diseases across the developing world.

Improving national level spatial mapping of malaria through alternative spatial and spatio-temporal models

The most common approach to create spatial prediction of malaria in the literature is to approximate a Gaussian process model using stochastic partial differential equation (SPDE). We compared SPDE to computationally faster alternatives, generalized additive model (GAM) and state-of-the-art machine learning method gradient boosted trees (GBM), with respect to their predictive skill for country-level malaria prevalence mapping. We also evaluated the intuition that incorporation of past data and the use of spatio-temporal models may improve predictive accuracy of present spatial distribution of malaria. Model performances varied among the countries and setting with SPDE and GAM performed well generally. The inclusion of past data is beneficial for GAM and GBM, but not for SPDE. We further investigated the weaknesses of SPDE at spatio-temporal setting and GAM at the edges of the countries. Taken together, we believe that spatial/spatio-temporal SPDE models should be evaluated alongside with the alternatives or at least GAM.