Objectives To assess the short-term impact of expected exposure to PM10 due to a major construction site (Jan 2013–Jan 2015) on the health of the population residing in the seven towns nearby (N = 235 000).

Method Estimates of PM10 short-term effects on all-cause and cause-specific mortality and on selected causes of hospital admissions were estimated for a pre-construction period (2007–2011) using Poisson regression models. Expected PM10 concentrations at ground level were estimated applying the ISCST3 Gaussian dispersion model to forecast PM10 emission rates due to the site. Mean counts of the 2007–2011 deaths and hospitalizations were taken to estimate the expected numbers of health events. The 2013–2015 impact was evaluated in terms of numbers of attributable deaths and hospitalizations during the construction site progress, under several counterfactual scenarios.

Results Between 2013 and 2015, PM10 levels exceeding the mean PM10 pre-construction concentrations would be responsible for 0.54 attributable deaths (0.13 cardiovascular and 0.04 respiratory) and for 0.14 cardiac, 0.05 cerebrovascular and 0.51 hospital admissions. If considering the EU limit of 40 µg/m³, PM10 levels would be responsible for 11.06 attributable deaths (2.69 cardiovascular and 0.90 respiratory) and for 2.81 cardiac, 1.17 cerebrovascular and 10.89 respiratory hospital admissions. Corresponding values above the WHO threshold of 20 µg/m³ would be 51.73 attributable deaths (12.58 cardiovascular and 4.17 respiratory) 13.60 cardiac, 5.37 cerebrovascular and 49.13 respiratory hospital admissions.

Conclusions The expected exposure appears to have a limited impact on health. Future monitoring of the actual exposure levels during the progress of the works will allow evaluating the accuracy of those estimates.