Abstract
INTRODUCTION: Climate change and population dynamics have been postulated as driving hand-food and mouth disease (HFMD) transmission. This study aims to develop a forecasting tool utilising climatic predictors and internet search queries in developing preventive strategies that would alleviate the burden of HFMD in Sabah. METHODS: Nine years of data was collected, consisting of temperature, humidity and rainfall from the Malaysia meteorological department, HFMD cases from the Sabah State Health Department and internet search queries from Google trends of years 2010-2018. Correlations between dependent and independent variables and their lagged functions were executed and integrated into a Seasonal auto-regressive moving average (SARIMA) model and subsequently, in measuring fit, the Akaikes Information Criterion (AIC) and log-likelihood metrics were utilised to select the best model. All statistical analysis was carried out using R. RESULTS: Google search trends evinced moderate positive correlations to HFMD cases (r0-6weeks: 0.47-0.56) with temperature revealing weaker positive correlations (r0-3weeks: 0.17-0.22). The autocorrelation functions revealed moderately positive correlations (r=0.15-1.0) at lags of zero and five weeks. Fit and parsimony were prioritised in selection, with a single model integrating mean temperature at lag zero-week and google search trends at lag one-week producing best fit (AIC: 4077.22, log-likelihood: -2030.61). DISCUSSION: Trajectorial forecasting oscillations of the model are stable up to four weeks in advance with accuracy being highest at one and two weeks justifying it as a low-cost, time-sensitive tool to be used in outbreak preparedness and mitigation. However, the model still requires validation and will carried out in the near future.
Links
BibTeX (Download)
@proceedings{APCPH-2019-55,
title = {Forecasting Hand-Foot and Mouth Disease (HFMD) Cases Using Weather Variables and Google Search Queries in Sabah, Malaysia},
author = {Vivek Jason Jayaraj and Victor CW Hoe},
year = {2019},
date = {2019-07-22},
urldate = {2019-07-22},
journal = {6th Asia-Pacific Conference on Public Health 2019 Proceedings},
issue = {6},
abstract = {INTRODUCTION: Climate change and population dynamics have been postulated as driving hand-food and mouth disease (HFMD) transmission. This study aims to develop a forecasting tool utilising climatic predictors and internet search queries in developing preventive strategies that would alleviate the burden of HFMD in Sabah. METHODS: Nine years of data was collected, consisting of temperature, humidity and rainfall from the Malaysia meteorological department, HFMD cases from the Sabah State Health Department and internet search queries from Google trends of years 2010-2018. Correlations between dependent and independent variables and their lagged functions were executed and integrated into a Seasonal auto-regressive moving average (SARIMA) model and subsequently, in measuring fit, the Akaikes Information Criterion (AIC) and log-likelihood metrics were utilised to select the best model. All statistical analysis was carried out using R. RESULTS: Google search trends evinced moderate positive correlations to HFMD cases (r0-6weeks: 0.47-0.56) with temperature revealing weaker positive correlations (r0-3weeks: 0.17-0.22). The autocorrelation functions revealed moderately positive correlations (r=0.15-1.0) at lags of zero and five weeks. Fit and parsimony were prioritised in selection, with a single model integrating mean temperature at lag zero-week and google search trends at lag one-week producing best fit (AIC: 4077.22, log-likelihood: -2030.61). DISCUSSION: Trajectorial forecasting oscillations of the model are stable up to four weeks in advance with accuracy being highest at one and two weeks justifying it as a low-cost, time-sensitive tool to be used in outbreak preparedness and mitigation. However, the model still requires validation and will carried out in the near future.},
note = {Type: ORAL PRESENTATION; Organisation: Department of Social and Preventive Medicine, Faculty of Medicine, University Malaya, Ministry of Health, Malaysia, Centre for Occupational and Environmental Health-UM, Faculty of Medicine, University Malaya},
keywords = {ARIMA, Coxsackie, EV71, google trends, prediction model, weather},
pubstate = {published},
tppubtype = {proceedings}
}