Cross-sectional entomological data reveals an increased risk of arboviral transmission in a year of record-breaking heat in Southern Europe

Abstract

Background: In 2023, an unprecedentedly high number of locally acquired cases of dengue virus occurred in Europe, including the first large outbreak in Rome (Italy). Globally, 2023 was warmer than any previous year on record since 1850, with each month from June to December being warmer than the corresponding month in previous years.

Methods: We conducted an extensive analysis of entomological data in Rome to investigate how the unprecedentedly high temperatures seen in 2023 affected the abundance and seasonal dynamics of the local vector Aedes albopictus population and the risk of autochthonous transmission of both dengue and chikungunya viruses. We applied a mathematical dynamic temperature-based model to analyse adult mosquito collections in 2023 compared to those available for 2012 and assess the consequent changes in the risk of arboviral transmission. Simulations of potential local transmission triggered by disease importation were used to evaluate the effectiveness of human and entomological surveillance for monitoring transmission risks, providing insights for future action plans RESULTS: Results highlight a higher abundance and longer active season of Ae. albopictus as well as a higher risk of local arbovirus transmission in 2023.

Conclusions: Our findings raise concern in light of the predictions for more frequent extremely warm years in Europe in the near future.

Fig. 1. Sampling sites within the metropolitan city of Rome monitored in 2012 (N = 10) and 2023 (N = 14).

Orange = sites monitored both in 2012 and 2023. Blue = sites monitored only in 2023. Bold dashed line = Grande Raccordo Anulare, the 68-km long ring-shaped highway encircling the city most urbanised area. This map incorporates data from © OpenStreetMap contributors, which is made available at www.openstreetmap.org/copyright under the Open Database License (ODbL).

Fig. 2

Average posterior distribution of Ae. albopictus females’ density (females/hectare) in 10 (2012) and 14 (2023) sampling sites within the most urbanised area of the city of Rome (Italy).

Fig. 3. Estimated basic reproduction number (R0) of Dengue and Chikungunya in 9 (2012) and 12 (2023) sites within the most urbanised area of the city of Rome (Italy).

R0 represents the basic reproduction number, i.e. the number of human secondary cases caused by a primary human infector. White = days in which R0 is not above epidemic threshold (R0 < 1). Gradient of Blue = days in which R0 is above epidemic threshold (R0 > 1).

Fig. 4. Probability (%) of observing an outbreak with >10 cases of dengue or chikungunya in 2012 and 2023.

Probability (%) of observing an outbreak with >10 cases of dengue a or chikungunya b in 2012 (red) and 2023 (yellow) following the introduction of a single imported human case at different weeks from May 1st to November 1st across different study sites located in the metropolitan city of Rome. Estimates were derived for sites associated with more than 30 residents per hectare. Boxplots show the probability (%) of observing an outbreak with >10 cases. The box represents the upper and lower quartiles, the centre line the median and the whiskers the 95% prediction interval. N = 9 independent samples for both years. The horizontal dashed line indicates the 5% probability used as a reference in the text. Dots = probability (%) of observing an outbreak with >10 cases of dengue or chikungunya in each site in 2012 and 2023.

Fig. 5. Probability of observing an outbreak with at least one secondary human case, probability of detecting at least one mosquito infected and human infected, and the mean reporting (solid lines) delays.

Probability of observing an outbreak with at least one secondary human case (yellow) and at least ten human cases (brown) for DENV (a) and CHIKV (b). The probability of detecting at least one mosquito infected and human-infected calculated during outbreaks with at least one autochthonous human case for DENV (c) and CHIKV d calculated as the percentage of simulations producing secondary transmission detected by the surveillance system. Red = human surveillance; light blue = mosquito lower capture rate (1/1000); dark blue = mosquito higher capture rate (1/100). The mean reporting (solid lines) delays (day) and 95% CI (dashed lines) from a case importation and the first detection of autochthonous transmission either by the collection of an infected mosquito under different capture rates (light and dark blue) or by notification of a symptomatic case (red) for DENV (e) and CHIKV (f), based on simulations producing at least one secondary transmission events. N = 12 independent samples.