Analysis of a time series of viral meningoencephalitis

Abstract

Introduction: the infectious meningoencephalitis is a medical emergency of diverse etiology, for the most part viral, that vary according to the geographical localization and the station of the year. It constitutes a fundamental problem of morbimortality and the related studies are reduced in number and present time, and they don't contribute resources that allow to predict the appearance of future hyperendemic or epidemic events.

Objective: to characterize a temporary series of viral meningoencephalitis and their presage.

Methods: an epidemiological, observational, descriptive, longitudinal, retrospective study was carried out in which it was organized the monthly incidence of viral meningoencephalitis chronologically during 10 years in Havana and the temporary series was analyzed.

Results: in the sequence graph tendency is not observed and the repetition of the picks suggests seasonality, and it is demonstrated in the ghastly analysis that happens to the 12 months and in the analysis preservative of the seasonal component the highest positive seasonal factors belong to May (7,102), June (5,037), July (4,459) and December (3,459); the noise of the series has a normal distribution. They originated predictions until December of 2026 by means of the pattern Simple moving average of 2 terms where the biggest incidence of cases of viral meningoencephalitis in Havana will continue happening in the months of May to July and in December, being able to arrive respectively in May and June to 27 and 25 cases.

Conclusions: the monthly incidence of viral meningoencephalitis in the time and place of the study has a stationary character in stocking, tendency was not observed. The seasonality existence was demonstrated in the months of May to July and December; it is a non random series that allowed to emit similar predictions in the following years.