Within-season influenza vaccine waning suggests potential net benefits to delayed vaccination in older adults in the United States.

ABSTRACT

BACKGROUND: There is growing evidence that there is within (intra-) season waning of influenza vaccine protection in older adults, suggesting there may be a benefit to giving influenza vaccine closer to the time of increased infection risk. We aimed to quantitatively evaluate the impact of modifying the timing of influenza vaccination in U.S. older adults. METHODS: Using historical data (2010/2011-2015/2016, inclusive) on influenza activity and vaccine uptake, we explore the optimal time to begin vaccinating older adults (≥65 years) in the U.S. to maximize prevention of influenza. We modelled the effect of changing the timing of vaccination by estimating the percentage change to the current disease burden and used this to calculate the estimated optimal week to begin vaccination in the U.S. RESULTS: When we assumed a relatively slower waning protection rate (over 52 weeks), the estimated optimal time to begin vaccinating those aged ≥65 years varied between mid-August (week 34, 2012-2013) and mid-late October (week 43, 2011-2012) depending on the season, resulting in 0.44% and 5.11% of the current disease burden prevented respectively. Under faster waning (over 26 weeks), the estimated optimal week varied between early September (week 37, 2012-2013) and mid-November (week 47, 2011-2012), resulting in 3.69% and 11.97% of the current disease burden prevented respectively. CONCLUSIONS: While it is difficult to determine the ideal time to start to vaccinate due to substantial variation in timing of individual seasons, we found that there are potentially substantial benefits to minimizing the time between vaccination and influenza activity in U.S. older adults. Modest delays in immunization were beneficial in the seasons we evaluated. If further evidence suggests fast waning, longer delays may be warrant as in these scenarios the timing of the current vaccination was often very suboptimal.