Uso da análise de séries temporais em estudos epidemiológicos

Abstract

This paper aims to introduce and raise interest on time series analysis in epidemiological studies. Conceptual aspects are described and methodological guidelines for this type of analysis are systematized. We defined the key concepts of time series analysis (trends, seasonal and cyclical variation, association and random variation), and how they can be applied to epidemiology. Methods to assess trends (annual percent change), based on Prais-Winsten regression models, and to quantify seasonal variation (Serfling model) were presented. We also introduced segmented regression analysis for interrupted time series, as a strategy for assessing effects of health interventions.