Years of life lost due to influenza-attributable mortality in older adults in the Netherlands: a competing risks approach

Abstract

We estimated the influenza mortality burden in adults 60 years of age and older in the Netherlands in terms of years of life lost, taking into account competing mortality risks. Weekly laboratory surveillance data for influenza and other respiratory pathogens and weekly extreme temperature served as covariates in Poisson regression models fitted to weekly age-group specific mortality data for the period 1999/2000 through 2012/13. Burden for age-groups 60-64 through 85-89 years was computed as years of life lost before age 90 (YLL90) using restricted mean lifetimes survival analysis and accounting for competing risks. Influenza-attributable mortality burden was greatest for persons aged 80-84 years, at 914 YLL90 per 100,000 persons (95% uncertainty interval:867, 963), followed by 85-89 years (787 YLL90/100,000; 95% uncertainty interval:741, 834). Ignoring competing mortality risks in the computation of influenza-attributable YLL90 would lead to substantial over-estimation of burden, from 3.5% for 60-64 years to 82% for persons aged 80-89 years at death. Failure to account for competing mortality risks has implications for accuracy of disease burden estimates, especially among persons aged 80 years and older. As the mortality burden borne by the elderly is notably high, prevention initiatives may benefit from being redesigned to more effectively prevent infection in the oldest age-groups.