Bronchiolitis Hospitalization in Southern Brazil from 2002 to 2012: A Count Time-series Approach

Abstract

Abstract INTRODUCTION: Because the antibody neutralizing respiratory syncytial virus (anti-RSV) has a short period of immunization and high cost, the identification of regions and months of highest occurrence of bronchiolitis is very important. METHODS: An Autoregressive Conditional Poisson model was constructed for count data and compared to the standard time-series Poisson regression model. RESULTS: The metropolitan area of Paraná presented the highest average occurrence from May to July. CONCLUSIONS: The constructed model presented a better fit and allowed prediction of when and where the bronchiolitis hospitalizations are distributed.

Keywords

• Autoregressive Conditional Poisson
• Bronchiolitis Hospitalizations
• Count Time Series