Time series analysis in predicting the oro-maxillo-facial traumas

Abstract

A subject of high interest in epidemiology is to register monthly and yearly the prevalence of certain diseases, in order to make prognosis regarding the expected number of new patients. It is suitable to solve these sorts of problems by treating these monthly reports as time series and making forecasts through the modelling methods of Time Series Analysis. Our paper propose this approach in making forecasts about the expected number of 4 types of oro-maxillo-facial traumas at patients from Iasi, during a period of 8 months (May 2015 -December 2015) based on 11 years records made within the Oro-Maxillo-Facial Surgery Clinic, "Sf. Spiridon" Emergency Hospital Iasi. The forecasts were made in SPSS 16.0, using the Expert Modeler module, and the best suitable models found were Simple Seasonal Exponential Smoothing and Simple Exponential Smoothing. The paper presents the results of this analysis and the predicted values, with 95% confidence intervals. (Abstract)