A Dynamical and Zero-Inflated Negative Binomial Regression Modelling of Malaria Incidence in Limpopo Province, South Africa

Stratification and Adaptation of Malaria Control Interventions in Chad

Malaria remains the leading cause of morbidity and mortality in Chad. The World Health Organization (WHO) has recommended that endemic countries stratify malaria to guide interventions. Thus, the Republic of Chad has initiated a stratification process based on malaria incidence with the aim of defining transmission risk and proposing interventions. We collected routine malaria data from health facilities from 2017-2021, the national survey on malaria indicators, the entomological data of NMCP operational research, the demographic and health surveys, and remote sensing of environmental data. Stratification was based on the adjusted incidence of malaria to guide interventions. The adjusted incidence of malaria was, on average, 374 cases per 1000 people in the country. However, it varied according to health districts. Health districts were stratified into very low malaria incidence (n = 25), low malaria incidence (n = 20), moderate malaria incidence (n = 46) and high malaria incidence (n = 38). Micro-stratification in health districts with very low incidence was carried out to identify districts with incidence <10 cases per 1000 person with a view to a malaria pre-elimination programme. Appropriate malaria control interventions were proposed based on the strata identified. Stratification enables the country to target interventions to accelerate the reduction of the burden caused by malaria with a pre-elimination goal.

Geo-epidemiology of malaria incidence in the Vhembe District to guide targeted elimination strategies, South-Africa, 2015-2018: a local resurgence

In South Africa, the population at risk of malaria is 10% (around six million inhabitants) and concern only three provinces of which Limpopo Province is the most affected, particularly in Vhembe District. As the elimination approaches, a finer scale analysis is needed to accelerate the results. Therefore, in the process of refining local malaria control and elimination strategies, the aim of this study was to identify and describe malaria incidence patterns at the locality scale in the Vhembe District, Limpopo Province, South Africa. The study area comprised 474 localities in Vhembe District for which smoothed malaria incidence curve were fitted with functional data method based on their weekly observed malaria incidence from July 2015 to June 2018. Then, hierarchical clustering algorithm was carried out considering different distances to classify the 474 smoothed malaria incidence curves. Thereafter, validity indices were used to determine the number of malaria incidence patterns. The cumulative malaria incidence of the study area was 4.1 cases/1000 person-years. Four distinct patterns of malaria incidence were identified: high, intermediate, low and very low with varying characteristics. Malaria incidence increased across transmission seasons and patterns. The localities in the two highest incidence patterns were mainly located around farms, and along the rivers. Some unusual malaria phenomena in Vhembe District were also highlighted as resurgence. Four distinct malaria incidence patterns were found in Vhembe District with varying characteristics. Findings show also unusual malaria phenomena in Vhembe District that hinder malaria elimination in South Africa. Assessing the factors associated with these unusual malaria phenome would be helpful on building innovative strategies that lead South Africa on malaria elimination.

Generalized Linear Models to Forecast Malaria Incidence in Three Endemic Regions of Senegal

Affecting millions of individuals yearly, malaria is one of the most dangerous and deadly tropical diseases. It is a major global public health problem, with an alarming spread of parasite transmitted by mosquito (Anophele). Various studies have emerged that construct a mathematical and statistical model for malaria incidence forecasting. In this study, we formulate a generalized linear model based on Poisson and negative binomial regression models for forecasting malaria incidence, taking into account climatic variables (such as the monthly rainfall, average temperature, relative humidity), other predictor variables (the insecticide-treated bed-nets (ITNs) distribution and Artemisinin-based combination therapy (ACT)) and the history of malaria incidence in Dakar, Fatick and Kedougou, three different endemic regions of Senegal. A forecasting algorithm is developed by taking the meteorological explanatory variable Xj at time t-𝓁j, where t is the observation time and 𝓁j is the lag in Xj that maximizes its correlation with the malaria incidence. We saturated the rainfall in order to reduce over-forecasting. The results of this study show that the Poisson regression model is more adequate than the negative binomial regression model to forecast accurately the malaria incidence taking into account some explanatory variables. The application of the saturation where the over-forecasting was observed noticeably increases the quality of the forecasts.

Towards malaria elimination: analysis of travel history and case forecasting using the SARIMA model in Limpopo Province

Despite various efforts and policy implementation aimed at controlling and eliminating malaria, imported malaria remains a major factor posing challenges in places that have made progress in malaria elimination. The persistence of malaria in Limpopo Province has largely been attributed to imported cases, thus reducing the pace of achieving the malaria-free target by 2025. Data from the Limpopo Malaria Surveillance Database System (2010-2020) was analyzed, and a seasonal auto-regressive integrated moving average (SARIMA) model was developed to forecast malaria incidence based on the incidence data's temporal autocorrelation. The study found that out of 57,288 people that were tested, 51,819 (90.5%) cases were local while 5469 (9.5%) cases were imported. Mozambique (44.9%), Zimbabwe (35.7%), and Ethiopia (8.5%) were the highest contributors of imported cases. The month of January recorded the highest incidence of cases while the least was in August. Analysis of the yearly figures showed an increasing trend and seasonal variation of recorded malaria cases. The SARIMA (3,1,1) X (3,1,0) [12] model used in predicting expected malaria case incidences for three consecutive years showed a decline in malaria incidences. The study demonstrated that imported malaria accounted for 9.5% of all cases. There is a need to re-focus on health education campaigns on malaria prevention methods and strengthening of indoor residual spray programs. Bodies collaborating toward malaria elimination in the Southern Africa region need to ensure a practical delivery of the objectives.

Modeling the spatial patterns of antenatal care utilization in Nigeria with inference based on Pólya-Gamma mixtures

Despite the vast advantages of making antenatal care visits, the service utilization among pregnant women in Nigeria is suboptimal. A five-year monitoring estimate indicated that about 24% of the women who had live births made no visit. The non-utilization induced excessive zeroes in the outcome of interest. Thus, this study adopted a zero-inflated negative binomial model within a Bayesian framework to identify the spatial pattern and the key factors hindering antenatal care utilization in Nigeria. We overcome the intractability associated with posterior inference by adopting a Pólya-Gamma data-augmentation technique to facilitate inference. The Gibbs sampling algorithm was used to draw samples from the joint posterior distribution. Results revealed that type of place of residence, maternal level of education, access to mass media, household work index, and woman's working status have significant effects on the use of antenatal care services. Findings identified substantial state-level spatial disparity in antenatal care utilization across the country. Cost-effective techniques to achieve an acceptable frequency of utilization include the creation of a community-specific awareness to emphasize the importance and benefits of the appropriate utilization. Special consideration should be given to older pregnant women, women in poor antenatal utilization states, and women residing in poor road network regions.

Weather integrated malaria prediction system using Bayesian structural time series model for northeast states of India

Malaria is an endemic disease in India and targeted to eliminate by the year 2030. The present study is aimed at understanding the epidemiological patterns of malaria transmission dynamics in Assam and Arunachal Pradesh followed by the development of a malaria prediction model using monthly climate factors. A total of 144,055 cases in Assam during 2011-2018 and 42,970 cases in Arunachal Pradesh were reported during the 2011-2019 period observed, and Plasmodium falciparum (74.5%) was the most predominant parasite in Assam, whereas Plasmodium vivax (66%) in Arunachal Pradesh. Malaria transmission showed a strong seasonal variation where most of the cases were reported during the monsoon period (Assam, 51.9%, and Arunachal Pradesh, 53.6%). Similarly, the malaria incidence was highest in the male population in both states (Asam, 55.75%, and Arunachal Pradesh, 51.43%), and the disease risk is also higher among the > 15 years age group (Assam, 61.7%, and Arunachal Pradesh, 67.9%). To predict the malaria incidence, Bayesian structural time series (BSTS) and Seasonal Auto-Regressive Integrated Moving Average with eXogenous factors (SARIMAX) models were implemented. A statistically significant association between malaria cases and climate variables was observed. The most influencing climate factors are found to be maximum and mean temperature with a 6-month lag, and it showed a negative association with malaria incidence. The BSTS model has shown superior performance on the optimal auto-correlated dataset (OAD) which contains auto-correlated malaria cases, cross-correlated climate variables besides malaria cases in both Assam (RMSE, 0.106; MAE, 0.089; and SMAPE, 19.2%) and Arunachal Pradesh (RMSE, 0.128; MAE, 0.122; and SMAPE, 22.6%) than the SARIMAX model. The findings suggest that the predictive performance of the BSTS model is outperformed, and it may be helpful for ongoing intervention strategies by governmental and nongovernmental agencies in the northeast region to combat the disease effectively.

Bayesian spatio-temporal distributed lag modeling for delayed climatic effects on sparse malaria incidence data

Background

In many areas of the Greater Mekong Subregion (GMS), malaria endemic regions have shrunk to patches of predominantly low-transmission. With a regional goal of elimination by 2030, it is important to use appropriate methods to analyze and predict trends in incidence in these remaining transmission foci to inform planning efforts. Climatic variables have been associated with malaria incidence to varying degrees across the globe but the relationship is less clear in the GMS and standard methodologies may not be appropriate to account for the lag between climate and incidence and for locations with low numbers of cases.

Methods

In this study, a methodology was developed to estimate the spatio-temporal lag effect of climatic factors on malaria incidence in Thailand within a Bayesian framework. A simulation was conducted based on ground truth of lagged effect curves representing the delayed relation with sparse malaria cases as seen in our study population. A case study to estimate the delayed effect of environmental variables was used with malaria incidence at a fine geographic scale of sub-districts in a western province of Thailand.

Results

From the simulation study, the model assumptions which accommodated both delayed effects and excessive zeros appeared to have the best overall performance across evaluation metrics and scenarios. The case study demonstrated lagged climatic effect estimation of the proposed modeling with real data. The models appeared to be useful to estimate the shape of association with malaria incidence.

Conclusions

A new method to estimate the spatiotemporal effect of climate on malaria trends in low transmission settings is presented. The developed methodology has potential to improve understanding and estimation of past and future trends in malaria incidence. With further development, this could assist policy makers with decisions on how to more effectively distribute resources and plan strategies for malaria elimination.

Analysis of zero inflated dichotomous variables from a Bayesian perspective: application to occupational health

BACKGROUND

Zero-inflated models are generally aimed to addressing the problem that arises from having two different sources that generate the zero values observed in a distribution. In practice, this is due to the fact that the population studied actually consists of two subpopulations: one in which the value zero is by default (structural zero) and the other is circumstantial (sample zero).

METHODS

This work proposes a new methodology to fit zero inflated Bernoulli data from a Bayesian approach, able to distinguish between two potential sources of zeros (structural and non-structural).

RESULTS

The proposed methodology performance has been evaluated through a comprehensive simulation study, and it has been compiled as an R package freely available to the community. Its usage is illustrated by means of a real example from the field of occupational health as the phenomenon of sickness presenteeism, in which it is reasonable to think that some individuals will never be at risk of suffering it because they have not been sick in the period of study (structural zeros). Without separating structural and non-structural zeros one would be studying jointly the general health status and the presenteeism itself, and therefore obtaining potentially biased estimates as the phenomenon is being implicitly underestimated by diluting it into the general health status.

CONCLUSIONS

The proposed methodology is able to distinguish two different sources of zeros (structural and non-structural) from dichotomous data with or without covariates in a Bayesian framework, and has been made available to any interested researcher in the form of the bayesZIB R package ( https://cran.r-project.org/package=bayesZIB ).

A model of malaria population dynamics with migrants

We present a compartmental model in ordinary differential equations of malaria disease transmission, accommodating the effect of indoor residual spraying on the vector population. The model allows for influx of infected migrants into the host population and for outflow of recovered migrants. The system is shown to have positive solutions. In the special case of no infected immigrants, we prove global stability of the disease-free equilibrium. Existence of a unique endemic equilibrium point is also established for the case of positive influx of infected migrants. As a case study we consider the combined South African malaria region. Using data covering 31 years, we quantify the effect of malaria infected immigrants on the South African malaria region.

Comparison of Predictive Models and Impact Assessment of Lockdown for COVID-19 over the United States

The novel Coronavirus Disease 2019 (COVID-19) remains a worldwide threat to community health, social stability, and economic development. Since the first case was recorded on December 29, 2019, in Wuhan of China, the disease has rapidly extended to other nations of the world to claim many lives, especially in the USA, the United Kingdom, and Western Europe. To stay ahead of the curve consequent of the continued increase in case and mortality, predictive tools are needed to guide adequate response. Therefore, this study aims to determine the best predictive models and investigate the impact of lockdown policy on the USA’ COVID-19 incidence and mortality. This study focuses on the statistical modelling of the USA daily COVID-19 incidence and mortality cases based on some intuitive properties of the data such as overdispersion and autoregressive conditional heteroscedasticity. The impact of the lockdown policy on cases and mortality was assessed by comparing the USA incidence case with that of Sweden where there is no strict lockdown. Stochastic models based on negative binomial autoregressive conditional heteroscedasticity [NB INGARCH (p,q)], the negative binomial regression, the autoregressive integrated moving average model with exogenous variables (ARIMAX) and without exogenous variables (ARIMA) models of several orders are presented, to identify the best fitting model for the USA daily incidence cases. The performance of the optimal NB INGARCH model on daily incidence cases was compared with the optimal ARIMA model in terms of their Akaike Information Criteria (AIC). Also, the NB model, ARIMA model and without exogenous variables are formulated for USA daily COVID-19 death cases. It was observed that the incidence and mortality cases show statistically significant increasing trends over the study period. The USA daily COVID-19 incidence is autocorrelated, linear and contains a structural break but exhibits autoregressive conditional heteroscedasticity. Observed data are compared with the fitted data from the optimal models. The results further indicate that the NB INGARCH fits the observed incidence better than ARIMA while the NB models perform better than the optimal ARIMA and ARIMAX models for death counts in terms of AIC and root mean square error (RMSE). The results show a statistically significant relationship between the lockdown policy in the USA and incidence and death counts. This suggests the efficacy of the lockdown policy in the USA.

COVID-19 incidence and mortality in Nigeria: gender based analysis

BACKGROUND

Coronavirus Disease 2019 (COVID-19) has been surging globally. Risk strata in medical attention are of dynamic significance for apposite assessment and supply distribution. Presently, no known cultured contrivance is available to fill this gap of this pandemic. The aim of this study is to develop a predictive model based on vector autoregressive moving average (VARMA) model of various orders for gender based daily COVID-19 incidence in Nigeria. This study also aims to proffer empirical evidence that compares incidence between male and female for COVID-19 risk factors.

METHODS

Wilcoxon signed-rank test is employed to investigate the significance of the difference in the gender distributions of the daily incidence. A VARMA model of various orders is formulated for the gender based daily COVID-19 incidence in Nigeria. The optimal VARMA model is identified using Bayesian information criterion. Also, a predictive model based on univariate autoregressive moving average model is formulated for the daily death cases in Nigeria. Fold change is estimated based on crude case-fatality risk to investigate whether there is massive underreporting and under-testing of COVID-19 cases in Nigeria.

RESULTS

Daily incidence is higher in males on most days from 11 April 2020 to 12 September 2020. Result of Wilcoxon signed-rank test shows that incidence among male is significantly higher than female (p-value < 2.22 × 10-16). White neural network test shows that daily female incidence is not linear in mean (p-value = 0.00058746) while daily male incidence is linear in mean (p-value = 0.4257). McLeod-Li test shows that there is autoregressive conditional heteroscedasticity in the female incidence (Maximum p-value = 1.4277 × 10-5) and male incidence (Maximum p-value = 9.0816 × 10-14) at 5% level of significance. Ljung-Box test (Tsay, 2014) shows that the daily incidence cases are not random (p-value=0.0000). The optimal VARMA model for male and female daily incidence is VARMA (0,1). The optimal model for the Nigeria's daily COVID-19 death cases is identified to be ARIMA (0,1,1). There is no evidence of massive underreporting and under-testing of COVID-19 cases in Nigeria.

CONCLUSIONS

Comparison of the observed incidence with fitted data by gender shows that the optimal VARMA and ARIMA models fit the data well. Findings highlight the significant roles of gender on daily COVID-19 incidence in Nigeria.

Evolution of Malaria Incidence in Five Health Districts, in the Context of the Scaling Up of Seasonal Malaria Chemoprevention, 2016 to 2018, in Mali

Context: In Mali, malaria transmission is seasonal, exposing children to high morbidity and mortality. A preventative strategy called Seasonal Malaria Chemoprevention (SMC) is being implemented, consisting of the distribution of drugs at monthly intervals for up to 4 months to children between 3 and 59 months of age during the period of the year when malaria is most prevalent. This study aimed to analyze the evolution of the incidence of malaria in the general population of the health districts of Kati, Kadiolo, Sikasso, Yorosso, and Tominian in the context of SMC implementation. Methods: This is a transversal study analyzing the routine malaria data and meteorological data of Nasa Giovanni from 2016 to 2018. General Additive Model (GAM) analysis was performed to investigate the relationship between malaria incidence and meteorological factors. Results: From 2016 to 2018, the evolution of the overall incidence in all the study districts was positively associated with the relative humidity, rainfall, and minimum temperature components. The average monthly incidence and the relative humidity varied according to the health district, and the average temperature and rainfall were similar. A decrease in incidence was observed in children under five years old in 2017 and 2018 compared to 2016. Conclusion: A decrease in the incidence of malaria was observed after the SMC rounds. SMC should be applied at optimal periods.

Impact of COVID-19 prevalence and mode of transmission on mortality cases over WHO regions

With the current outbreak of coronavirus disease 2019 (COVID-19), countries have been on rising preparedness to detect and isolate any imported and locally transmitted cases of the disease. It is observed that mode of transmission of the disease varies from one country to the other. Recent studies have shown that COVID-19 cases are not influenced by race and weather conditions. In this study, effect of modes of transmission of COVID-19 is considered with respect to prevalence and mortality counts in World Health Organisation (WHO) regions. Also, a negative binomial model is formulated for new death cases in all WHO regions as a function of confirmed cases, confirmed new cases, total deaths and modes of transmission, with the goal of identifying a model that predicts the total new death cases the best. Results from this study show that there is strong linear relationship among the COVID-19 confirmed cases, total new deaths and mode of transmission in all WHO regions. Findings highlight the significant roles of modes of transmission on total new death cases over WHO regions. Mode of transmission based on community transmission and clusters of cases significantly affects the number of new deaths in WHO regions. Vuong test shows that the formulated negative binomial model fits the data better than the null model.

Epidemic Changes and Spatio-Temporal Analysis of Japanese Encephalitis in Shaanxi Province, China, 2005-2018

Japanese encephalitis (JE) is a mosquito-borne viral disease, which is the most serious viral encephalitis in China and other countries of the Asia-Pacific region. Since 2005, the epidemic patterns of JE have changed dramatically in China because of the vaccination of children younger than 15 years old, and JE is expanding geographically along with global warming. This retrospective epidemiological study analyzed dynamic environmental factors and the spatio-temporal distribution of human cases of JE in Shaanxi Province—one of the most severely affected areas of China—from 2005 to 2018. The results demonstrated that the high-risk population changed rapidly as the annual rate of JE cases increased by more than 40% in the age group >60 years during the study period, and endemic areas expanded northward in Shaanxi. Hotspot analysis detected four hotspots accounting for 52.38% the total cases, and the panel negative binomial regression model revealed that the spatio-temporal distribution of JE was significantly affected by temperature, relative humidity, wind velocity, El Niño-Southern Oscillation, coniferous forest coverage, and urban areas. These findings can provide useful information for improving current strategies and measures to reduce disease incidence.

Modelling Malaria Incidence in the Limpopo Province, South Africa: Comparison of Classical and Bayesian Methods of Estimation

Malaria infects and kills millions of people in Africa, predominantly in hot regions where temperatures during the day and night are typically high. In South Africa, Limpopo Province is the hottest province in the country and therefore prone to malaria incidence. The districts of Vhembe, Mopani and Sekhukhune are the hottest districts in the province. Malaria cases in these districts are common and malaria is among the leading causes of illness and deaths in these districts. Factors contributing to malaria incidence in Limpopo Province have not been deeply investigated, aside from the general knowledge that the province is the hottest in South Africa. Bayesian and classical methods of estimation have been applied and compared on the effect of climatic factors on malaria incidence. Credible and confidence intervals from a negative binomial model estimated via Bayesian estimation and maximum likelihood estimation, respectively, were utilized in the comparison process. Overall assumptions underpinning each method were given. The Bayesian method appeared more robust than the classical method in analysing malaria incidence in Limpopo Province. The classical method identified rainfall and temperature during the night to be significant predictors of malaria incidence in Mopani, Vhembe and Waterberg districts. However, the Bayesian method found rainfall, normalised difference vegetation index, elevation, temperatures during the day and night to be the significant predictors of malaria incidence in Mopani, Sekhukhune and Vhembe districts of Limpopo Province. Both methods affirmed that Vhembe district is more susceptible to malaria incidence, followed by Mopani district. We recommend that the Department of Health and Malaria Control Programme of South Africa allocate more resources for malaria control, prevention and elimination to Vhembe and Mopani districts of Limpopo Province.

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.

Modeling an association between malaria cases and climate variables for Keonjhar district of Odisha, India: a Bayesian approach

Malaria, a vector-borne disease, is a significant public health problem in Keonjhar district of Odisha (the malaria capital of India). Prediction of malaria, in advance, is an urgent need for reporting rolling cases of disease throughout the year. The climate condition do play an essential role in the transmission of malaria. Hence, the current study aims to develop and assess a simple and straightforward statistical model of an association between malaria cases and climate variates. It may help in accurate predictions of malaria cases given future climate conditions. For this purpose, a Bayesian Gaussian time series regression model is adopted to fit a relationship of the square root of malaria cases with climate variables with practical lag effects. The model fitting is assessed using a Bayesian version of R² (RsqB). Whereas, the predictive ability of the model is measured using a cross-validation technique. As a result, it is found that the square root of malaria cases with lag 1, maximum temperature, and relative humidity with lag 3 and 0 (respectively), are significantly positively associated with the square root of the cases. However, the minimum and average temperatures with lag 2, respectively, are observed as negatively (significantly) related. The considered model accounts for moderate amount of variation in the square root of malaria cases as received through the results for RsqB. We also present Absolute Percentage Errors (APE) for each of the 12 months (January-December) for a better understanding of the seasonal pattern of the predicted (square root of) malaria cases. Most of the APEs obtained corresponding to test data points is reasonably low. Further, the analysis shows that the considered model closely predicted the actual (square root of) malaria cases, except for some peak cases during the particular months. The output of the current research might help the district to develop and strengthen early warning prediction of malaria cases for proper mitigation, eradication, and prevention in similar settings.