Background: Spontaneous pneumothoraces (SP) tend to occur in clusters which have been related to atmospheric pressure variations and thunderstorm insurgence. We examined the influence of standard meteorological parameter variations and concentrations of the major air pollutants on incidence of spontaneous pneumothorax (SP) in a highly developed industrial area (Turin, Italy).
Methods: From October 2002 to December 2007, 591 SP patients were prospectively evaluated. For each day, standard weather parameters and concentration of air pollutants were recorded.
Results: The total number of admissions for SP was 591. The number of days with admissions was 363, which represents the 19% of the total number of days in the study period (1918). Eighty-one percent of days with SP admissions were clusterized. Results of statistical analysis showed that the sequence of SP events was not random. There was relationship between SP and daily wind speed (WS) minimum, daily standard deviation of NO(2), NO(2), CO(2) daily maximum and minimum, O3 daily minimum, daily mean CO(2) (p = 0.01), daily NO(2) minimum (p = 0.001). Multiple regression analysis has shown relationship between number of SP admissions and increase of daily mean and minimum NO(2) (p = 0.001), decrease of NO(2) standard deviation (p = 0.01), decrease of daily mean and minimum O(3) (p = 0.01), and of maximum of NO (p = 0.001), increase of daily O(3) standard deviation (p = 0.05). Daily decrement of standard deviation of temperature (p = 0.01) and increment ofWS anomalies and minima (p = 0.01) were also significant.
Conclusions: Meteorological parameters and atmospheric pollutants might explain cluster hospitalization.
Keywords: Meteorological parameters; atmospheric pollutants; bivariate and multiple regression; clustering; spectral analysis; spontaneous pneumothorax.