Excess hospitalizations and in-hospital mortality associated with seasonal influenza in Italy: a 11-year retrospective study

Background: Influenza and flu-like syndromes are difficult to monitor because the symptoms are not specific, laboratory tests are not routinely performed, and diagnosis codes are often lacking or incompletely registered in medical records. This may result in an underestimation of hospital admissions, associated costs, and in-hospital mortality. Therefore, this study aimed to estimate the public health and economic burden of hospitalisations associated with influenza in Italy, at the national and regional levels. Methods: This 11-year retrospective study included patients admitted to hospitals for influenza or diagnoses associated with influenza (including respiratory and cardiocirculatory conditions) from 2008/09 to 2018/19. Data on hospitalisations were extracted from the Italian Hospital Discharge Records. Information on weekly influenza-like syndrome incidence and weekly average temperature were used to estimate the burden of influenza in terms of hospital admissions in every Italian region and for different age groups by applying a negative binomial model. The model was also applied to estimate in-hospital mortality and the total costs of influenza and influenza-like hospital admissions. Results: Over the study period, in addition to 3,970 average seasonal admissions coded as influenza, we estimated an average of 21,500 excess hospitalization associated with influenza per season, which corresponds to 36.4 cases per 100,000. Most of the excess hospitalisations concerned older individuals (> 65 years) and children (0-4 years) with 86 and 125 cases per 100,000, respectively. Large variations were observed across regions. Overall, the total estimated hospital burden associated with influenza (including respiratory and cardiocirculatory conditions) was approximately €123 m per year. While the in-hospital mortality for admissions with a primary diagnosis of influenza was very low (~ 150 cases per season), cases increased dramatically for primary diagnoses of influenza and pneumonia (about 9,500 cases per season). The average seasonal in-hospital deaths attributable to influenza were equal to 2,775 cases. Conclusions: Our findings suggest a remarkable underestimation of the burden of influenza, mostly in the older population but not neglectable in younger individuals. Our results may aid the management of current and future flu seasons and should be used for policy making (e.g., vaccine strategies) and operation management choices (e.g., planning and staffing beds during influenza peaks). Overall, the present study supports the need for increased testing for influenza in Italy to tackle the current underestimation of influenza burden.