Application of time series control charts to model and monitor the seasonality of norovirus

Abstract

INTRODUCTION: Time series control charts are tools that allow one to identify the nature of certain variations in a process, whether they are variations due to common causes, deemed of little importance, or to special causes. The goal of this study was to apply time series control charts to analyse the seasonality observed in norovirus infections in children admitted from May 2008 to April 2011 to a paediatric hospital in Belém City, Pará State, Amazon Region, Brazil. MATERIALS AND METHODS: To apply the models, it was necessary to make the stationary series, which was performed by applying the first difference of monthly norovirus positivity rates. To construct the charts, Kolmogorov-Smirnov normality test was used to verify if the data were normally distributed. RESULTS: During the three years, the highest prevalence was observed between September and October 2008 (63.6%) and February 2010 (62.1%). The study series was under statistical control, with a tendency for stability and no evidence of seasonality or cycle. No outlying values were observed, which strongly suggest that no atypical factor that could affect the stability of the series was present. CONCLUSION: This study demonstrates the applicability of time series control charts in the seasonality analysis of several pathogens studied in the biomedical sciences, and it is the first time that this statistical method has been applied to assess the time distribution of norovirus.