Multi-step polynomial regression method to model and forecast malaria incidence

Determining suitable machine learning classifier technique for prediction of malaria incidents attributed to climate of Odisha

This study investigated the influence of climate factors on malaria incidence in the Sundargarh district, Odisha, India. The WEKA machine learning tool was used with two classifier techniques, Multi-Layer Perceptron (MLP) and J48, with three test options, 10-fold cross-validation, percentile split, and supplied test. A comparative analysis was carried out to ascertain the superior model among malaria prediction accuracy techniques in varying climate contexts. The results suggested that J48 had exhibited better skill than MLP with the 10-fold cross-validation method over the percentile split and supplied test options. J48 demonstrated less error (RMSE = 0.6), better kappa = 0.63, and higher accuracy = 0.71), suggesting it as most suitable model. Seasonal variation of temperature and humidity had a better association with malaria incidents than rainfall, and the performance was better during the monsoon and post-monsoon when the incidents are at the peak.

Predicting COVID-19 statistics using machine learning regression model: Li-MuLi-Poly

In this paper, linear regression (LR), multi-linear regression (MLR) and polynomial regression (PR) techniques are applied to propose a model Li-MuLi-Poly. The model predicts COVID-19 deaths happening in the United States of America. The experiment was carried out on machine learning model, minimum mean square error model, and maximum likelihood ratio model. The best-fitting model was selected according to the measures of mean square error, adjusted mean square error, mean square error, root mean square error (RMSE) and maximum likelihood ratio, and the statistical t-test was used to verify the results. Data sets are analyzed, cleaned up and debated before being applied to the proposed regression model. The correlation of the selected independent parameters was determined by the heat map and the Carl Pearson correlation matrix. It was found that the accuracy of the LR model best-fits the dataset when all the independent parameters are used in modeling, however, RMSE and mean absolute error (MAE) are high as compared to PR models. The PR models of a high degree are required to best-fit the dataset when not much independent parameter is considered in modeling. However, the PR models of low degree best-fits the dataset when independent parameters from all dimensions are considered in modeling.

Epidemiology and Risk Factors for Notifiable Scrub Typhus in Taiwan during the Period 2010-2019

Scrub typhus is a zoonotic disease caused by the bacterium Orientia tsutsugamushi. In this study, the epidemiological characteristics of scrub typhus in Taiwan, including gender, age, seasonal variation, climate factors, and epidemic trends from 2010 to 2019 were investigated. Information about scrub typhus in Taiwan was extracted from annual summary data made publicly available on the internet by the Taiwan Centers for Disease Control. From 2010 to 2019, there were 4352 confirmed domestic and 22 imported cases of scrub typhus. The incidence of scrub typhus ranged from 1.39 to 2.30 per 100,000 from 2010-2019, and peaked in 2013 and 2015-2016. Disease incidence varied between genders, age groups, season, and residence (all p < 0.001) from 2010 to 2019. Risk factors were being male (odds ratio (OR) =1.358), age 40 to 64 (OR = 1.25), summer (OR = 1.96) or fall (OR = 1.82), and being in the Penghu islands (OR = 1.74) or eastern Taiwan (OR = 1.92). The occurrence of the disease varied with gender, age, and place of residence comparing four seasons (all p < 0.001). Weather, average temperature (°C) and rainfall were significantly correlated with confirmed cases. The number of confirmed cases increased by 3.279 for every 1 °C (p = 0.005) temperature rise, and 0.051 for every 1 mm rise in rainfall (p = 0.005). In addition, the total number of scrub typhus cases in different geographical regions of Taiwan was significantly different according to gender, age and season (all p < 0.001). In particular, Matsu islands residents aged 20-39 years (OR = 2.617) and residents of the Taipei area (OR = 3.408), northern Taiwan (OR = 2.268) and eastern Taiwan (OR = 2.027) were affected during the winter. Males and females in the 50-59 age group were at high risk. The total number of imported cases was highest among men, aged 20-39, during the summer months, and in Taipei or central Taiwan. The long-term trend of local cases of scrub typhus was predicted using the polynomial regression model, which predicted the month of most cases in a high-risk season according to the seasonal index (1.19 in June by the summer seasonal index, and 1.26 in October by the fall seasonal index). The information in this study will be useful for policy-makers and clinical experts for direct prevention and control of chigger mites with O. tsutsugamushi that cause severe illness and are an economic burden to the Taiwan medical system. These data can inform future surveillance and research efforts in Taiwan.

Climate based malaria forecasting system for Andhra Pradesh, India

Malaria is a major public health problem in tropical and subtropical countries of the World. During the year 1999, Visakhapatnam district of Andhra Pradesh, India experienced a major epidemic of malaria, and nearly 41,805 cases were reported. Hence, a retrospective malaria surveillance study was conducted from 2001 to 2016 and reported nearly a total of 149,317 malaria cases during the study period. Of which, Plasmodium vivax contributes 32%, and Plasmodium falciparum contributes 68% of the total cases. Malaria cases follow a strong seasonal variation and 70% of cases are reported during the monsoon periods. In the present study, we exploited multi step polynomial regression and seasonal autoregressive integrated moving average (SARIMA) models to forecast the malaria cases in the study area. The polynomial model predicted malaria cases with high predictive power and found that malaria cases at lag one, and population played a vital role in malaria transmission. Similarly, mean temperature, rainfall and Normalized Difference Vegetation Index build a significant impact on malaria cases. The best fit model was SARIMA (1, 1, 2) (2, 1, 1)12 which was used for forecasting monthly malaria incidence for the period of January 2015 to December 2016. The performance accuracy of both models are similar, however lowest Akaike information criterion score was observed by the polynomial model, and this approach can be helpful further for forecasting malaria incidence to implement effective control measures in advance for combating malaria in India.

Devising a method towards development of early warning tool for detection of malaria outbreak

Background & objectives

Uncertainty often arises in differentiating seasonal variation from outbreaks of malaria. The present study was aimed to generalize the theoretical structure of sine curve for detecting an outbreak so that a tool for early warning of malaria may be developed.

Methods

A 'case/mean-ratio scale' system was devised for labelling the outbreak in respect of two diverse districts of Assam and Rajasthan. A curve-based method of analysis was developed for determining outbreak and using the properties of sine curve. It could be used as an early warning tool for Plasmodium falciparum malaria outbreaks.

Result

In the present method of analysis, the critical C_max(peak value of sine curve) value of seasonally adjusted curve for P. falciparum malaria outbreak was 2.3 for Karbi Anglong and 2.2 for Jaisalmer districts. On case/mean-ratio scale, the C_max value of malaria curve between C_maxand 3.5, the outbreak could be labelled as minor while >3.5 may be labelled as major. In epidemic years, with mean of case/mean ratio of ≥1.00 and root mean square (RMS) ≥1.504 of case/mean ratio, outbreaks can be predicted 1-2 months in advance.

Interpretation & conclusions

The present study showed that in P. falciparum cases in Karbi Anglong (Assam) and Jaisalmer (Rajasthan) districts, the rise in C_maxvalue of curve was always followed by rise in average/RMS or both and hence could be used as an early warning tool. The present method provides better detection of outbreaks than the conventional method of mean plus two standard deviation (mean+2 SD). The identified tools are simple and may be adopted for preparedness of malaria outbreaks.

Real-time predictive seasonal influenza model in Catalonia, Spain

Influenza surveillance is critical to monitoring the situation during epidemic seasons and predictive mathematic models may aid the early detection of epidemic patterns. The objective of this study was to design a real-time spatial predictive model of ILI (Influenza Like Illness) incidence rate in Catalonia using one- and two-week forecasts. The available data sources used to select explanatory variables to include in the model were the statutory reporting disease system and the sentinel surveillance system in Catalonia for influenza incidence rates, the official climate service in Catalonia for meteorological data, laboratory data and Google Flu Trend. Time series for every explanatory variable with data from the last 4 seasons (from 2010–2011 to 2013–2014) was created. A pilot test was conducted during the 2014–2015 season to select the explanatory variables to be included in the model and the type of model to be applied. During the 2015–2016 season a real-time model was applied weekly, obtaining the intensity level and predicted incidence rates with 95% confidence levels one and two weeks away for each health region. At the end of the season, the confidence interval success rate (CISR) and intensity level success rate (ILSR) were analysed. For the 2015–2016 season a CISR of 85.3% at one week and 87.1% at two weeks and an ILSR of 82.9% and 82% were observed, respectively. The model described is a useful tool although it is hard to evaluate due to uncertainty. The accuracy of prediction at one and two weeks was above 80% globally, but was lower during the peak epidemic period. In order to improve the predictive power, new explanatory variables should be included.

Predicting factors for malaria re-introduction: an applied model in an elimination setting to prevent malaria outbreaks

BACKGROUND

Malaria re-introduction is a challenge in elimination settings. To prevent re-introduction, receptivity, vulnerability, and health system capacity of foci should be monitored using appropriate tools. This study aimed to design an applicable model to monitor predicting factors of re-introduction of malaria in highly prone areas.

METHODS

This exploratory, descriptive study was conducted in a pre-elimination setting with a high-risk of malaria transmission re-introduction. By using nominal group technique and literature review, a list of predicting indicators for malaria re-introduction and outbreak was defined. Accordingly, a checklist was developed and completed in the field for foci affected by re-introduction and for cleared-up foci as a control group, for a period of 12 weeks before re-introduction and for the same period in the previous year. Using field data and analytic hierarchical process (AHP), each variable and its sub-categories were weighted, and by calculating geometric means for each sub-category, score of corresponding cells of interaction matrices, lower and upper threshold of different risks strata, including low and mild risk of re-introduction and moderate and high risk of malaria outbreaks, were determined. The developed predictive model was calibrated through resampling with different sets of explanatory variables using R software. Sensitivity and specificity of the model were calculated based on new samples.

RESULTS

Twenty explanatory predictive variables of malaria re-introduction were identified and a predictive model was developed. Unpermitted immigrants from endemic neighbouring countries were determined as a pivotal factor (AHP score: 0.181). Moreover, quality of population movement (0.114), following malaria transmission season (0.088), average daily minimum temperature in the previous 8 weeks (0.062), an outdoor resting shelter for vectors (0.045), and rainfall (0.042) were determined. Positive and negative predictive values of the model were 81.8 and 100 %, respectively.

CONCLUSIONS

This study introduced a new, simple, yet reliable model to forecast malaria re-introduction and outbreaks eight weeks in advance in pre-elimination and elimination settings. The model incorporates comprehensive deterministic factors that can easily be measured in the field, thereby facilitating preventive measures.

Malaria incidence from 2005-2013 and its associations with meteorological factors in Guangdong, China

BACKGROUND

The temporal variation of malaria incidence has been linked to meteorological factors in many studies, but key factors observed and corresponding effect estimates were not consistent. Furthermore, the potential effect modification by individual characteristics is not well documented. This study intends to examine the delayed effects of meteorological factors and the sub-population's susceptibility in Guangdong, China.

METHODS

The Granger causality Wald test and Spearman correlation analysis were employed to select climatic variables influencing malaria. The distributed lag non-linear model (DLNM) was used to estimate the non-linear and delayed effects of weekly temperature, duration of sunshine, and precipitation on the weekly number of malaria cases after controlling for other confounders. Stratified analyses were conducted to identify the sub-population's susceptibility to meteorological effects by malaria type, gender, and age group.

RESULTS

An incidence rate of 1.1 cases per 1,000,000 people was detected in Guangdong from 2005-2013. High temperature was associated with an observed increase in malaria incidence, with the effect lasting for four weeks and a maximum relative risk (RR) of 1.57 (95% confidence interval (CI): 1.06-2.33) by comparing 30°C to the median temperature. The effect of sunshine duration peaked at lag five and the maximum RR was 1.36 (95% CI: 1.08-1.72) by comparing 24 hours/week to 0 hours/week. A J-shaped relationship was found between malaria incidence and precipitation with a threshold of 150 mm/week. Over the threshold, precipitation increased malaria incidence after four weeks with the effect lasting for 15 weeks, and the maximum RR of 1.55 (95% CI: 1.18-2.03) occurring at lag eight by comparing 225 mm/week to 0 mm/week. Plasmodium falciparum was more sensitive to temperature and precipitation than Plasmodium vivax. Females had a higher susceptibility to the effects of sunshine and precipitation, and children and the elderly were more sensitive to the change of temperature, sunshine duration, and precipitation.

CONCLUSION

Temperature, duration of sunshine and precipitation played important roles in malaria incidence with effects delayed and varied across lags. Climatic effects were distinct among sub-groups. This study provided helpful information for predicting malaria incidence and developing the future warning system.

The use of quantile regression to forecast higher than expected respiratory deaths in a daily time series: a study of New York City data 1987-2000

Forecasting higher than expected numbers of health events provides potentially valuable insights in its own right, and may contribute to health services management and syndromic surveillance. This study investigates the use of quantile regression to predict higher than expected respiratory deaths.

Data taken from 70,830 deaths occurring in New York were used. Temporal, weather and air quality measures were fitted using quantile regression at the 90th-percentile with half the data (in-sample). Four QR models were fitted: an unconditional model predicting the 90th-percentile of deaths (Model 1), a seasonal / temporal (Model 2), a seasonal, temporal plus lags of weather and air quality (Model 3), and a seasonal, temporal model with 7-day moving averages of weather and air quality. Models were cross-validated with the out of sample data. Performance was measured as proportionate reduction in weighted sum of absolute deviations by a conditional, over unconditional models; i.e., the coefficient of determination (R1).

The coefficient of determination showed an improvement over the unconditional model between 0.16 and 0.19. The greatest improvement in predictive and forecasting accuracy of daily mortality was associated with the inclusion of seasonal and temporal predictors (Model 2). No gains were made in the predictive models with the addition of weather and air quality predictors (Models 3 and 4). However, forecasting models that included weather and air quality predictors performed slightly better than the seasonal and temporal model alone (i.e., Model 3 > Model 4 > Model 2)

This study provided a new approach to predict higher than expected numbers of respiratory related-deaths. The approach, while promising, has limitations and should be treated at this stage as a proof of concept.

A scoping review of malaria forecasting: past work and future directions

OBJECTIVES

There is a growing body of literature on malaria forecasting methods and the objective of our review is to identify and assess methods, including predictors, used to forecast malaria.

DESIGN

Scoping review. Two independent reviewers searched information sources, assessed studies for inclusion and extracted data from each study.

INFORMATION SOURCES

Search strategies were developed and the following databases were searched: CAB Abstracts, EMBASE, Global Health, MEDLINE, ProQuest Dissertations & Theses and Web of Science. Key journals and websites were also manually searched.

ELIGIBILITY CRITERIA FOR INCLUDED STUDIES

We included studies that forecasted incidence, prevalence or epidemics of malaria over time. A description of the forecasting model and an assessment of the forecast accuracy of the model were requirements for inclusion. Studies were restricted to human populations and to autochthonous transmission settings.

RESULTS

We identified 29 different studies that met our inclusion criteria for this review. The forecasting approaches included statistical modelling, mathematical modelling and machine learning methods. Climate-related predictors were used consistently in forecasting models, with the most common predictors being rainfall, relative humidity, temperature and the normalised difference vegetation index. Model evaluation was typically based on a reserved portion of data and accuracy was measured in a variety of ways including mean-squared error and correlation coefficients. We could not compare the forecast accuracy of models from the different studies as the evaluation measures differed across the studies.

CONCLUSIONS

Applying different forecasting methods to the same data, exploring the predictive ability of non-environmental variables, including transmission reducing interventions and using common forecast accuracy measures will allow malaria researchers to compare and improve models and methods, which should improve the quality of malaria forecasting.

INDIA'S APPROACH TO ELIMINATING PLASMODIUM FALCIPARUM MALARIA: A MODELING PERSPECTIVE

Approximately one-third of the world's population that is at risk to malaria lives in India. Plasmodium falciparum, a deadly form of malaria, accounts for about 50% of the cases there. Since 1940s, India has used a number of programs to combat the disease with variable success. Since 1998, the total numbers of malaria cases, and in particular P. falciparum cases, have been steadily declining, making India one of the success stories among the countries supported by the Roll Back Malaria (RBM) Partnership. This article considers India's P. falciparum control methods from the perspective of a Ross-MacDonald type model. The model is fitted to the P. falciparum cases in India over the period 1983–2009. We focus on the disease reproduction number as being a measure of success of programs. Before the start of RBM measures, the disease reproduction number was [Formula: see text], meaning that the incidence of disease was increasing among the population. With the new control measures [Formula: see text], suggesting that P.falciparum cases may be declining to zero but extremely slowly. The model here projects 0.734 million cases of P. falciparum malaria for 2010, down from 1.14 million cases in 2000. This impressive 36% decrease falls somewhat short of the RBM's goal of 50% reduction. However, a sensitivity analysis of the disease reproduction number done here suggests that India's control programs do apply controls at the most critical points in the disease cycle; namely, mosquito biting rates, mosquito mortality, and treatment of infected humans. This suggests that as more resources become available, they should be applied to strengthen these controls. The novelty here is in fitting recent data on malaria from India to derive current values of the disease reproduction number.

Theoretical investigation of malaria prevalence in two Indian cities using the response surface method

Background

Elucidation of the relationships between malaria incidence and climatic and non-climatic factors in a region is of utmost importance in understanding the causative factors of disease spread and design of control strategies. Very often malaria prevalence data is restricted to short time scales (months to few years). This demands application of rigorous statistical modelling techniques for analysis and prediction. The monthly malaria prevalence data for three to five years from two cities in southern India, situated in two different climatic zones, are studied to capture their dependence on climatic factors.

Methods

The statistical technique of response surface method (RSM) is applied for the first time to study any epidemiological data. A new step-by-step model reduction technique is proposed to refine the initial model obtained from RSM. This provides a simpler structure and gives better fit. This combined approach is applied to two types of epidemiological data (Slide Positivity Rates values and Total Malaria cases), for two cities in India with varying strengths of disease prevalence and environmental conditions.

Results

The study on these data sets reveals that RSM can be used successfully to elucidate the important environmental factors influencing the transmission of the disease by analysing short epidemiological time series. The proposed approach has high predictive ability over relatively long time horizons.

Conclusions

This method promises to provide reliable forecast of malaria incidence across varying environmental conditions, which may help in designing useful control programmes for malaria.

Mathematical models of malaria--a review

Mathematical models have been used to provide an explicit framework for understanding malaria transmission dynamics in human population for over 100 years. With the disease still thriving and threatening to be a major source of death and disability due to changed environmental and socio-economic conditions, it is necessary to make a critical assessment of the existing models, and study their evolution and efficacy in describing the host-parasite biology. In this article, starting from the basic Ross model, the key mathematical models and their underlying features, based on their specific contributions in the understanding of spread and transmission of malaria have been discussed. The first aim of this article is to develop, starting from the basic models, a hierarchical structure of a range of deterministic models of different levels of complexity. The second objective is to elaborate, using some of the representative mathematical models, the evolution of modelling strategies to describe malaria incidence by including the critical features of host-vector-parasite interactions. Emphasis is more on the evolution of the deterministic differential equation based epidemiological compartment models with a brief discussion on data based statistical models. In this comprehensive survey, the approach has been to summarize the modelling activity in this area so that it helps reach a wider range of researchers working on epidemiology, transmission, and other aspects of malaria. This may facilitate the mathematicians to further develop suitable models in this direction relevant to the present scenario, and help the biologists and public health personnel to adopt better understanding of the modelling strategies to control the disease.