A retrospective analysis of the surveillance data on laboratory confirmed cases of influenza in 4 post pandemic seasons in Serbia was performed to evaluate predictors of hospitalization and admission to intensive care units (ICU). The specimens, including nasal and throat swabs were tested for influenza. Univariate and multivariate logistic regression analyses were performed. Data of a total of 777 confirmed influenza cases were analyzed. Age > 65 years, the presence of any co-morbidity or the presence of ≥ 2 comorbidities, infection with influenza virus subtype A (H1) pdm09, and an interval greater than 3 days between symptom onset and the first physician visit, were independently associated with hospital admission. These variables, as well as infection with non-subtype influenza virus A, were predictors for ICU admission. Obesity and chronic neurological disease were independent predictors for ICU admission but not hospitalization. Overall, 41.7% of patients with influenza had at least one co-morbidity, but only 3% of all patients were vaccinated against influenza. Identification of high risk groups and education of these groups regarding their increased susceptibility to severe forms of influenza, and in particular regarding the importance of influenza vaccination, is essential.
Keywords: Serbia; hospitalization; influenza; predictors; surveillance.