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

Abstract

ObjectivesThere 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.DesignScoping review. Two independent reviewers searched information sources, assessed studies for inclusion and extracted data from each study.Information sourcesSearch 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 studiesWe 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.ResultsWe 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.ConclusionsApplying 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.

Use of Space Time Model for Forecasting Mortality due to Malaria: A Case of Ifakara and Rufiji Health and Demographic Surveillance System Sites

“…In this study, the models included covariates information such as rainfall, temperature and NDVI which are important factors for malaria transmission [42], and mosquito net which is reduces malaria transmission. The inclusion of these covariates in the model may be not weaken the accuracy of forecasts because malaria transmission-reducing interventions are considered in the models [7]. Rainfall and ownership of mosquito net were independently associated with mortality due to malaria in the study areas.…”

“…Malaria incidence forecasting models have been developed in many endemic countries [5][6][7][8][9]. These models mostly used time series models and other use climate related predictors, such as rainfall, temperature and normalized difference vegetation index.…”

“…In addition, these models did not include spatial effect which could capture heterogeneity in malaria transmission. The accuracy of forecasts may be weakened due to transmission-reducing predictors were not considered in the models [7]. Health and demographic surveillance system (HDSS) data, however, are becoming increasingly available with the routinely collecting data and adopting the verbal autopsy approach [8].…”

A Seasonal Autoregressive Integrated Moving Average (SARIMA) forecasting model to predict monthly malaria cases in KwaZulu-Natal, South Africa

“…An understanding of the assumptions underlying a predictive model and its advantage(s) and disadvantage(s) is vital when developing a forecast model. [5] The SARIMA approach can exhibit temporal trends such as seasonality and autocorrelation (which is a correlation of a time series with its own past and future values) [6] that are actualised by eliminating high-frequency noise in the data. Furthermore, owing to the model's ability to perform automated model determination over a time series, predictions can be said to be reliable if longer time series data are employed.…”

“…The formulated models are easy to interpret in a retrospective study. [5] Nevertheless, the formulation of the models requires general mathematical and statistical skills, and an understanding of a relevant statistical package/software for the execution of analysis. The required mathematical and statistical skills are not limited to trigonometry, complex numbers, calculus, linear regression (multiple regression and weighted least square) and basic probability.…”