Influenza-attributable mortality in the Iberian Peninsula, seasons 2012/13 - 2017/18

Mazagatos, Clara; Silva, Susana Pereira; Delgado-Sanz, Concha; Rodrigues, Ana Paula; Oliva, Jesus; Nunes, Baltazar; Larrauri, Amparo

http://hdl.handle.net/10400.18/6009

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
SEE2018_flumomo_final.pdf		743.34 KB	Adobe PDF	Download

Send Feedback

Authors

Mazagatos, Clara

Silva, Susana Pereira

Abstract(s)

Introduction / Objectives: Influenza has a considerable impact on mortality. Significant excess all-cause mortality is often observed during winter seasons, mainly among the elderly. Quantifying the mortality associated to influenza is important to estimate the real impact of seasonal influenza epidemics on the population. Our objective is to estimate the influenza-attributable mortality in Portugal and Spain for seasons 2012-13 to 2017-18 Methodology: To estimate influenza-attributable mortality in both countries, a multiplicative Poisson regression time-series model (FluMOMO) with over-dispersion corrected for trend and seasonality, with ISO-week as time unit was used. We estimated influenza-attributable deaths and mortality rates (and their 95% CIs), for all ages and by age group (<5, 5–14, 15–64 and ≥65 years). Weekly all-cause deaths were obtained from computerized civil registers, covering 100% of the Portuguese population and 92% of the Spanish population. Information on influenza activity is included in the model as the Goldstein proxy (weekly ILI rates x weekly positivity rate – For Portugal the same over-all data was used for each age group) and temperature as the difference between the observed and predicted average weekly mean temperatures (NOAA database). Mortality rates (by 100 000 persons) were calculated using national population data, by January 1st every year, linearly interpolated through the year to obtain weekly data Results: Influenza-attributable mortality showed variation by season and age. In both Portugal and Spain, the highest influenza-attributable mortality rates for the ≥65 age group were observed in seasons 2014-15 and 2016-17, when A(H3N2) was dominant or in codominance (219 and 204 in Portugal, and 228 and 169 in Spain, respectively). For the 15-64 age group, in Portugal we observed the highest and lowest mortality rates in seasons 2016-17 and 2013-14 (6 and 3, respectively), in which A(H3N2) was dominant or codominant; in Spain, highest and lowest rates were estimated in seasons 2017-18 and 2012-13 (4 and 1, respectively), in which B was dominant. For all ages, seasons 2014-15, 2016-17 and 2017-18 yielded the highest influenza-attributable mortality rates in both countries (48, 48 and 34 in Portugal and 41, 31 and 30 in Spain, respectively). These results are provisional and will be updated by the end of the season Discussion/Conclusions: The FluMOMO model approach provides a valuable estimation of the impact of influenza-attributable mortality by season and age group. The quantification of the mortality attributable to influenza at national level is essential for the evaluation of the impact of control measures, such as influenza vaccination programs.