Revista Cubana de Salud Pública
versão impressa ISSN 0864-3466
COUTIN MARIE, Gisele. Methods for detection of seasonal variations in Cuba applied to health surveillance. Rev. cub. salud pública [online]. 2007, vol.33, n.1, pp. 0-0. ISSN 0864-3466. http://dx.doi.org/10.1590/S0864-34662007000100006.
Seasonal variation is a component of time series identified in a number of health events and describied in morbidity, mortality among others. Identification of seasonality, the analysis of its variability and detection of its period are key elements to discover deviations from the regular behaviour of health phenomena, which allows early alert in the face of possible outbreaks or epidemics and adequate health surveillance. Graphic methods are useful tools to examine seasonality since they provide a more rapid general overview than the simple direct data observation. This paper showed the results of the application of various graphs to detect and analyze seasonality, that is, the box and moustache graphs, the correlogram or autocorrelation function graph and the periodgram that is the main detecting tool of the significant spectral analysis frequencies. A descriptive study was performed to scrutinize seasonal variation in monthly series of infant mortality (rate per 1000 live births), births(absolute number), Aedes Aegypti infestation (absolute number of foci) in Cuba from 1998 to 2005. These three examined health events exhibited manifest seasonality with different variability degrees and similar periods. The graphs were advantageous for health surveillance since they offered an improved visual representation and demonstrated more clearly the short-term variation within the times series
Palavras-chave : Seasonality; time series; graphs; correlogram; periodgram; surveillance.