Avraham Ebenstein PhD, Eyal Frank and Yaniv Reingewertz PhD
Background: Exposure to air pollution in the form of particulate matter smaller than 10 µm in diameter (PM10) has been associated with increased morbidity and mortality. However, since air pollution is correlated with confounding factors that might otherwise affect health, identifying the causal link has proven challenging.
Objectives: To identify the effect of PM10 on hospital admissions due to respiratory illnesses.
Methods: We used the Instrumental Variable (IV) methodology to control for confounding factors affecting hospital admissions. Exploiting the timing of sandstorms as an instrumental variable allows for a better estimate of the relationship between PM10 and hospital admissions. Data on PM concentrations and hospital admissions rates were compiled for Israel’s two largest cities, Jerusalem and Tel Aviv, for 2007–2009. We compared our IV estimates to those derived from a Poisson regression, which is commonly used in the existing literature.
Results: Sandstorms led to an increase of 307 µg/m3 of PM10 concentrations. A 10 µg/m3 increase in PM10 is associated with a 0.8% increase in hospital admissions due to respiratory conditions, using Poisson regression. The same finding was noted using the IV methodology.
Conclusions: The association between PM10 and hospital admission reflects a primarily causal relationship. Instrumental variable methodology could be applied to analysis of the effect of air pollution on hospital admissions.