An Association between Rainy Days with Clinical Dengue Fever in Dhaka, Bangladesh: Findings from a Hospital Based Study

Abstract

Background: Dengue, a febrile illness, is caused by a Flavivirus transmitted by Aedes aegypti and Aedes albopictus mosquitoes. Climate influences the ecology of the vectors. We aimed to identify the influence of climatic variability on the occurrence of clinical dengue requiring hospitalization in Zone-5, a high incidence area of Dhaka City Corporation (DCC), Bangladesh.

Methods and findings: We retrospectively identified clinical dengue cases hospitalized from Zone-5 of DCC between 2005 and 2009. We extracted records of the four major catchment hospitals of the study area. The Bangladesh Meteorological Department (BMD) provided data on temperature, rainfall, and humidity of DCC for the study period. We used autoregressive integrated moving average (ARIMA) models for the number of monthly dengue hospitalizations. We also modeled all the climatic variables using Poisson regression. During our study period, dengue occurred throughout the year in Zone-5 of DCC. The median number of hospitalized dengue cases was 9 per month. Dengue incidence increased sharply from June, and reached its peak in August. One additional rainy day per month increased dengue cases in the succeeding month by 6% (RR = 1.06, 95% CI: 1.04-1.09).

Conclusions: Dengue is transmitted throughout the year in Zone-5 of DCC, with seasonal variation in incidence. The number of rainy days per month is significantly associated with dengue incidence in the subsequent month. Our study suggests the initiation of campaigns in DCC for controlling dengue and other Aedes mosquito borne diseases, including Chikunguniya from the month of May each year. BMD rainfall data may be used to determine campaign timing.

Keywords: climatic variability; dengue; disease control campaigns; prediction; rainfall; seasonal variation; year-long transmission.

Figure 1

Median polish of the number of dengue cases over months during years 2005–2009 in Zone-5 of Dhaka City Corporation, Dhaka, Bangladesh. Y-axis represents the monthly additive effect of hospitalized dengue cases to the overall median in square root scale. X-axis represents the months. To get the median dengue cases in a specific month, add the overall median to the month’s effect and take the square. For example, the median number of dengue cases in June is (3 − 0)² = 9 cases (approximately).

Figure 2

Monthly aggregated observations for hospitalized dengue cases from Zone-5 and weather parameters from 2005 to 2009 in Dhaka city.

Figure 3

Two seasonal ARIMA models for the monthly number of dengue cases during years 2005–2009 in Zone-5 of Dhaka City Corporation, Dhaka, Bangladesh. X-axis indicates the index of the months, and Y-axis indicates the total number of dengue cases in square root scale. The black line is the observed cases and the red line is the modeled cases. The red line with pink region indicates the forecast of the next 20 months with 95% confidence band for prediction.

Figure 4

Prediction of total number of monthly dengue hospitalizations by Poisson regression model.