Abstract

Global health surveillance systems collect vast amounts of information to monitor seasonal infections, like influenza that can form pandemics given the confluence of favorable social, environmental, and epidemiological conditions. Growing analytic capabilities could transform these surveillance systems into interactive platforms for outbreak forecasting and real-time tracking of public health interventions. This task requires further advancing the understanding of global disease dynamics and the development of the statistical methodology to emulate the dynamics with available data. We developed a longitudinal research framework to characterize the seasonality of infectious diseases with public global health surveillance data utilizing a 2-decadal set of 2422 weekly time series of 14 reported outcomes for 173 Member States from the World Health Organization’s (WHO) international influenza virological surveillance system, FluNet. Specifically, we estimated the seasonal characteristics (such as peak timing and amplitude) and their uncertainty using mixed effects models with harmonic components and the δ-method. First, after developing a traditional model for a set of time-referenced observations for each country and exploring temporal patterns for variables of interest, we described the model as the entire vector of all observations. Next, we formulated a set of research hypotheses and emphasized critical aspects of the model, specifically the interplay of random and fixed effects. Finally, we clarified the context for using the term linear in the model, which refers to the additivity of the fixed and random effects and the application of harmonic functions reflecting nonlinear seasonal periodic changes over time. We developed multi-panel visualizations to present a complex interplay of seasonal peaks across geographic locations. We offered ways to quantify average seasonal peaks for homogeneous climatic zones and WHO regions. We produced a compilation, or an analecta of data visualizations to describe the global traveling waves of influenza across geographic locations. We also discussed the challenges of data completeness and credibility, especially in countries with limited public health resources. The methodology developed for longitudinal studies offers new insights and flexibility to accommodate the challenges of surveillance data. Our findings and interpretations suggest ways to improve data collection, reporting, and analysis methods and guide the development of statistical methodology, predictive modeling, and decision-making efforts to protect public health.