Influenza outbreaks and weather conditions in Argentina

RUSTICUCCI, MATILDE; BETTOLLI, MARÍA LAURA; HARRIS, MARÍA DE LOS ANGELES; MARTÍNEZ, LAURA; PODESTÁ, OSCAR

Paris, Francia

Congreso; Third International Workshop on Geomedical Systems Geomed 2001; 2001

The present study examined associations between weather variables and influenza outbreaks by analysing data recorded in the Flu Epidemiological Vigilance (VEG in Spanish) program. During the years 1999 and 2000, the strains were the following ones: A/Sydney/05/97 (H3N2); A/Bayern/07/95 (H1N1); B/Beijing/184/93 in 1999, and A/New Caledonia/20/99(H1N1) y B/Yamanashi/166/98 like in 2000. The periods considered were the cold season of 1999, from Sunday 6 June (week 23) to Saturday 30 October (week 43) and the cold season of 2000, from Sunday 30 April (week 18) to Saturday 25 November (week 47). The influenza data were sorted geographically and into three broad age groups: between 0 to 18 years, 18 to 64 years, and over 64 years. The influenza peak shows a preferential occurrence direction. In 1999, they appear successively in Bariloche, Santa Rosa, Bahía Blanca, Córdoba, Tucumán, Buenos Aires and Mar del Plata. In 2000, once more it starts in Bariloche and there are some new centres. It remains the succession Santa Rosa, Bahía Blanca, Córdoba, while the weeks of the maximum number of influenza cases appears earlier than they did in 1999 in Mar del Plata and Buenos Aires. The following weather parameters were considered in weekly values: maximum and minimum temperatures, mean temperature, mean dew-point temperature, mean relative humidity, mean sea-level pressure, mean wind speed and mean calm frequencies, as well as four mean wind-direction frequencies (North, East, South and West). The daily temperature difference was also included and the number of days a week into cold spells was calculated at every locality. Most localities showed significant and negative correlations with temperature. High relative humidity seemed to be a significant parameter in influenza outbreaks, even in localities without correlations with temperature. Several lagged correlation with weather variables resulted significant. This result increases the potential forecast value of meteorological parameters for prevention purposes. Non significant associations appeared to be between the daily temperature differences and the virus outbreaks. Buenos Aires city did not show significant correlations with weather variables. This fact may be related to the heterogeneity of the centres considered in this city. The group of adult patients (18 to 64 years) seemed to be the most weather-sensitive group. The most significant variable, the temperature, showed that the peak of influenza cases occurred in the coldest week of the year or in the next one in every locality. The occurrence of cold events were not a determinant factor for the emergence of the influenza peak, except in the cases where the cold event occurred during the coldest week of the year, strengthening the coldness. The determinant factor for the influenza outbreaks is the relativity of the temperature value in every locality, consequently every city presents its own threshold. Since only two years of data were available it would not be appropriate to establish a climatic threshold value.