Ambient Temperature and Dengue Hospitalisation in Brazil: A 10-year period two-stage case-design time series analysis

Abstract

Background: Climate factors are known to influence seasonal patterns of dengue transmission However, little is known about the effect of extreme heat on the severity of dengue infection, such as hospital admission. We aimed to quantify the effect of ambient temperature on dengue hospitalisation risk in Brazil Methods: We retrieved daily dengue hospitalisation counts by each of 5,570 municipalities across the 27 states of Brazil from 1st of January 2010 to 31st of December 2019, from the Brazilian Public Hospital Admission System (SIH). We obtained average daily ambient temperature for each municipality from the ERA5-land product reanalysis. We combined distributed lag non-linear models with time stratified design model framework to pool an estimate for dose-response and lag-response structures for the association of Dengue hospitalisation relative risk (RR) and temperature. We estimated the overall dengue hospitalisation RR for the whole country as well as for each of the five macro-regions by meta- analysing state level estimates. Findings: 579,703 hospital admissions due to dengue occurred over the 10 years period of 2010 to 2019. We observed a positive association between high temperatures and high risk of hospitalisation for Brazil and each of the five macro-regions. The overall RR for dengue hospitalisation was at the 50th percentile of temperature distribution 1.40 (95% IC 1.27-1.54) and at 95th percentile of temperature the RR was 1.50 (1.38-1.66) for Brazil, relative to the minimum temperature, which was the one with the lowest risk. We also found lagged effects of heat on hospitalisation, particularly observed on the same day (lag 0) both at the 50th percentile and 95th. Interpretation: High temperatures are associated with an increase in the risk of hospitalisation by dengue infection. These findings may guide preparedness and mitigation policies during dengue season outbreaks, particularly on the health-care demand.