摘要：Figures See all figures Authors Imed Gargouri , Ahmed Rebaï , Moncef Khadhraoui , Monem Kallel , Kathouar Jmal-Hammami , Khaled Medhioub , Mohamed Larbi Masmoudi , Boubaker Elleuch , Jean-Marie Haguenoer Service de médecine du travail et de pathologie professionnelle, CHU Hédi Chaker, Faculté de médecine, Université de Sfax, Tunisie, Centre de biotechnologie de Sfax, BP K, 3038 Sfax, Tunisie, Laboratoire Eau, énergie, et environnement, École nationale des ingénieurs de Sfax, Université de Sfax, Tunisie, Unité de recherche « Étude et gestion des environnements côtiers et urbains », École nationale des ingénieurs de Sfax, Université de Sfax, Tunisie, Laboratoire de toxicologie professionnelle et de l’environnement, Faculté de médecine, Université Lille 2, France Key words: air pollution, epidemiology, epidemiologic factors, epidemiologic studies, feasibility studies, hospitalisation, morbidity DOI : 10.1684/ers.2006.0021 Page(s) : 467-75 Published in: 2006 Introduction : In the Tunisian region of Sfax, pollution and its environmental impact have been studied since the 1990s, but little work has been done on its health effects. Here we examine the feasibility of assessing the health impact of air pollution in Sfax. Material and methods : Since 1996, the city of Sfax has been equipped with one stationary air quality measurement station that automatically records the data collected. Here we examined data for 13 months (from October 1996 through June 1998 over three separate temperature periods: one cold and two hot for the following indicators: carbon monoxide (CO), sulfur dioxide (SO 2), nitrogen dioxide (NO 2), ozone (O 3) and particulate matter with an aerodynamic diameter less than 10 μm (PM 10). The number of daily hospital admissions (for cardiovascular or respiratory signs) for longer than 24 hours was selected as the health indicator. The time series analysis of the associations between daily variations in pollution indicators and health indicators was performed with SPSS software, version 10.0.5. Results: Daily mean pollution levels from October 1, 1996, to June 30, 1997, and March 1-June 1998 varied from 0.1 to 7.2 mg/m 3 for CO, 11.0 to 163.6 μg/m 3 for NO 2, 17.7 to 83.6 μg/m 3 for O 3, 5.8 to 181.1 μg/m 3 for SO 2 and 10.0 to 214.8 μg/m 3 for PM 10. Cardiovascular admissions were significantly associated with PM 10 only (p < 0.03), and respiratory admissions with SO 2 only (p<0.04). Cardiovascular admissions also showed seasonal trends. Hospital morbidity was strongly influenced by days of the week but not associated with temperature, relative humidity, or holidays. The month of Ramadan, however, appeared to affect cardiovascular but not respiratory morbidity. Discussion: This study showed that hospital morbidity increased with the daily levels of certain pollutants: cardiovascular hospitalization with PM 10 and respiratory hospitalization with SO 2. Hospital admissions were highest for cardiovascular and respiratory causes on days of low and intermediate pollutant concentrations. The hospital data are an important source of information but must be restructured for use in epidemiologic surveillance. Pollution data were recorded for a relatively short period, and information for April 1997 is missing. Conclusion: The analysis showed a significant association between acid and particulate pollution (SO 2, PM 10) and hospital morbidity, despite pollution levels that were not very elevated in Sfax and despite the presence of only a single measuring station, which limits interpretation. This feasibility study allowed us to show that a network to monitor the short-term health effects of air pollution can and should be established. It should include healthcare professionals, public health personnel, and specialists in pollution measurement.
