Natural experiment on airport activity and hospital admissions in Northern Italy
Abstract
Background: Assessing the health impact of environmental exposures such as airport activity is challenging due to strong confounding and non-random exposure. The collapse in air traffic during the COVID-19 pandemic provides a natural experiment to evaluate causal effects in a real-world setting.
Methods: We analyzed age-standardized hospitalizations (total, respiratory, cardiovascular) in people directly exposed to Treviso airport-related environmental factors. Bayesian changepoint detection (BEAST) and counterfactual modeling (CausalImpact) were used to assess structural changes and deviations from a counterfactual trajectory. The pre-period was 2002–2019; post-periods was 2020–2024. Two alternative post-periods (2021-2024 and 2022-2024) were used for temporal falsification. Covariates included hospitalizations in residents in a control area and a placebo outcome (appendicitis and cholelithiasis admissions) to account for systemic healthcare dynamics.
Results: Air traffic showed a clear structural break in 2020. No corresponding changepoint was observed in hospital admissions. Only respiratory hospitalizations showed a statistically significant reduction −14% [95%CI −22% −6%], whereas total −3.7% [−14% +8%], circulatory −12% [−25% +3%] and placebo outcomes −8% [−23% +12%] showed no evidence of change. However, the respiratory effect was not temporally consistent: it attenuated and became non-significant in 2021–2024 −12% [−22% +6%], but re-emerged in 2022–2024 −17% [−27% to + 2%] despite the recovery of air traffic.
Conclusions: Findings do not support a robust causal relationship between airport activity and hospital admissions. Observed patterns are more consistent with post-pandemic healthcare dynamics, including possible “immunity debt”. These results highlight the importance of combining counterfactual methods with temporal consistency checks when informing environmental health policies and interpreting observational data.
Key messages: • The 2020 collapse in air traffic provided a natural experiment, but no temporally coherent effect on hospital admissions was observed.
• Combining counterfactual models with temporal falsification improves causal inference in environmental epidemiology.
Related
- Associations of short-term exposure to multiple air pollutants with hospital admissions for ischemic stroke in Lanzhou, China
- Multi-season evaluation and analysis of categorical trend forecasts of influenza hospital admissions in the United States
- Etiology and Severity Profiles of Acute Pancreatitis in Consecutive Admissions at a Japanese Acute-care Hospital: Comparison with the 2016 Nationwide Survey
- Long-term Rotarix Vaccine Effectiveness in Coastal Kenya, Estimated From Child Hospital Admissions Data From July 2014 to December 2022.
- Brownfield land and emergency hospital admissions in England: a small-area ecological study using contamination hazard score as an exposure proxy
BibTeX
@article{ramigni2026natural,
title = {{Natural experiment on airport activity and hospital admissions in Northern Italy}},
author = {M Ramigni and V Gobbetto and C Piovesan},
journal = {The European Journal of Public Health},
year = {2026},
month = oct,
doi = {10.1093/eurpub/ckag169.1054},
url = {https://doi.org/10.1093/eurpub/ckag169.1054},
note = {CC BY-NC 4.0},
}
RIS
TY - JOUR
TI - Natural experiment on airport activity and hospital admissions in Northern Italy
AU - M Ramigni
AU - V Gobbetto
AU - C Piovesan
PY - 2026
DA - 2026/10/01
JO - The European Journal of Public Health
DO - 10.1093/eurpub/ckag169.1054
UR - https://doi.org/10.1093/eurpub/ckag169.1054
AB - Background: Assessing the health impact of environmental exposures such as airport activity is challenging due to strong confounding and non-random exposure. The collapse in air traffic during the COVID-19 pandemic provides a natural experiment to evaluate causal effects in a real-world setting. Methods: We analyzed age-standardized hospitalizations (total, respiratory, cardiovascular) in people directly exposed to Treviso airport-related environmental factors. Bayesian changepoint detection (BEAST) and counterfactual modeling (CausalImpact) were used to assess structural changes and deviations from a counterfactual trajectory. The pre-period was 2002–2019; post-periods was 2020–2024. Two alternative post-periods (2021-2024 and 2022-2024) were used for temporal falsification. Covariates included hospitalizations in residents in a control area and a placebo outcome (appendicitis and cholelithiasis admissions) to account for systemic healthcare dynamics. Results: Air traffic showed a clear structural break in 2020. No corresponding changepoint was observed in hospital admissions. Only respiratory hospitalizations showed a statistically significant reduction −14% [95%CI −22% −6%], whereas total −3.7% [−14% +8%], circulatory −12% [−25% +3%] and placebo outcomes −8% [−23% +12%] showed no evidence of change. However, the respiratory effect was not temporally consistent: it attenuated and became non-significant in 2021–2024 −12% [−22% +6%], but re-emerged in 2022–2024 −17% [−27% to + 2%] despite the recovery of air traffic. Conclusions: Findings do not support a robust causal relationship between airport activity and hospital admissions. Observed patterns are more consistent with post-pandemic healthcare dynamics, including possible “immunity debt”. These results highlight the importance of combining counterfactual methods with temporal consistency checks when informing environmental health policies and interpreting observational data. Key messages: • The 2020 collapse in air traffic provided a natural experiment, but no temporally coherent effect on hospital admissions was observed. • Combining counterfactual models with temporal falsification improves causal inference in environmental epidemiology.
SN - 1101-1262
SN - 1464-360X
ER -