Estimation of reproduction number and non stationary spectral analysis of Dengue epidemic

TL;DR

This study estimates the Dengue reproduction number and analyzes its non-stationary spectral patterns in Ahmedabad, revealing phase locking with climatic factors and increasing post-monsoon outbreak durations despite decreasing monsoon intensity.

Contribution

It introduces a combined approach of reproduction number estimation and wavelet-based spectral analysis to understand Dengue dynamics and environmental influences in a specific urban setting.

Findings

Reproduction number $R_p$ estimated with Monte Carlo simulations.

Dengue peaks lag monsoon by approximately 9 weeks.

Post-monsoon Dengue outbreaks are increasing in duration.

Abstract

In this work we analyze the post monsoon Dengue outbreaks by analyzing the transient and long term dynamics of Dengue incidences and its environmental correlates in Ahmedabad city in western India from 2005-2012. We calculate the reproduction number $R_p$ using the growth rate of post monsoon Dengue outbreaks and biological parameters like host and vector incubation periods and vector mortality rate, and its uncertainties are estimated through Monte-Carlo simulations by sampling parameters from their respective probability distributions. Reduction in Female Aedes mosquito density required for an effective prevention of Dengue outbreaks is also calculated. The non stationary pattern of Dengue incidences and its climatic correlates like rainfall temperature is analyzed through Wavelet based methods. We find that the mean time lag between peak of monsoon and Dengue is 9 weeks. Monsoon and Dengue cases are phase locked from 2008-2012 in the 16-32 weeks band. The duration of post monsoon outbreak has been increasing every year, especially post 2008, even though the intensity and duration of monsoon has been decreasing. Temperature and Dengue incidences show correlations in the same band, but phase lock is not stationary.