Prediction of dengue outbreaks based on disease surveillance, meteorological and socio-economic data

Abstract

Background: The goal of this research is to create a system that can use the available relevant information about the factors responsible for the spread of dengue and; use it to predict the occurrence of dengue within a geographical region, so that public health experts can prepare for, manage and control the epidemic. Our study presents new geospatial insights into our understanding and management of health, disease and health-care systems.

Methods: We present a machine learning-based methodology capable of providing forecast estimates of dengue prediction in each of the fifty districts of Thailand by leveraging data from multiple data sources. Using a set of prediction variables, we show an increase in prediction accuracy of the model with an optimal combination of predictors which include: meteorological data, clinical data, lag variables of disease surveillance, socioeconomic data and the data encoding spatial dependence on dengue transmission. We use Generalized Additive Models (GAMs) to fit the relationships between the predictors (with a lag of one month) and the clinical data of Dengue hemorrhagic fever (DHF) using the data from 2008 to 2012. Using the data from 2013 to 2015 and a comparative set of prediction models, we evaluate the predictive ability of the fitted models according to RMSE and SRMSE as well as using adjusted R-squared value, deviance explained and change in AIC.

Results: The model allows for combining different predictors to make forecasts with a lead time of one month and also describe the statistical significance of the variables used to characterize the forecast. The discriminating ability of the final model was evaluated against Bangkok specific constant threshold and WHO moving threshold of the epidemic in terms of specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV).

Conclusions: The out-of-sample validation showed poorer results than the in-sample validation, however it demonstrated ability in detecting outbreaks up-to one month ahead. We also determine that for the predicting dengue outbreaks within a district, the influence of dengue incidences and socioeconomic data from the surrounding districts is statistically significant. This validates the influence of movement patterns of people and spatial heterogeneity of human activities on the spread of the epidemic.

Keywords: Data-driven epidemiology; Dengue forecasting; Disease surveillance; Generalized additive models (GAMs).

Fig. 1

Dependency chart for the occurence of DH/DHF/DSS Cases

Fig. 2

Mean Monthly Rainfall within Bangkok for different the years

Fig. 3

Monthly DTR within Bangkok for different years

Fig. 4

Data suggests that the outbreak increases every alternate year

Fig. 5

The DHF incidence peaked in October and November. Graph shows the amount of dengue patients for each month for the years 2008–2015. Rainy season continues from mid-May to mid-October

Fig. 6

The DHF incidence peaks in the months of October and November. However in the year 2013, DHF outbreak happened for several months continuously

Fig. 7

Monthly observed and predicted dengue cases from 2008-2012.

Fig. 8

Association between the meteorological variables and dengue over lags of 0–3 months. Solid lines represent relative risks (RR) of dengue cases whereas the dotted lines depict the upper and lower limits of 95% confidence intervals. P-values and the initial analyses are listed in Table 1

Fig. 9

Cross-correlation between the Outcome and the Predictors

Fig. 10

Retrospective transmission period is calculated to account for the influence of dengue incidences in each of the target districts (Fig 10) and lag-response curve for an increase in various units of dengue incidences (Fig 10)

Fig. 11

Association between past dengue count over optimal lags (lag 1,2 and 23) for each target district and lag 1,12 for surrounding district count data and garbage data with dengue outbreak. Solid lines represent relative risks (RR) of dengue cases whereas the dotted lines depict the upper and lower limits of 95% confidence intervals

Fig. 12

Residual Diagnosis. Top-left is the histogram of residuals; top-right is the partial ACF plot. On bottom-left is the Q-Q plot for the deviance residuals whereas the relationship between the reported and predicted cases is shown on bottom-right

Fig. 13

The final model (F) was trained on the data from 2008−2014 and the validation is performed on external data of year 2015