Waning in influenza vaccine effectiveness against influenza A(H1N1)pdm09-associated hospitalization in children in 2012/2013.

Measuring waning in vaccine effectiveness (VE) is challenging due to potential depletion-of-susceptibles bias. Some SARS-CoV-2 studies excluded individuals with prior infection and adjusted for the probability of remaining uninfected. We applied this approach to assess waning influenza VE in Hong Kong during the 2012/2013 season. First, we estimated the infection risk for unvaccinated children using published serological and surveillance data. Next, we derived infection risk for vaccinated children, assuming VE against infection of 57%. Uncorrected VE from 14 to 270 days post-vaccination was estimated from hospitalized children. We calculated the rate of depletion of susceptibles given infection risk and VE corrected for depletion-of-susceptibles bias. Waning rates for uncorrected and bias-corrected VE were measured by comparing VE at day 270 versus day 14. Bias was assessed as the absolute difference between two waning rates in percentage points. Waning rate of uncorrected VE was overestimated by 5.9 percentage points or 1.3 percentage points when assessed up to day 120. Bias was substantial when assuming 80% unvaccinated, and all vaccinated children were initially uninfected, but minimal when these proportions were similar. The observed waning in 2012/2013 was unlikely due to depletion-of-susceptibles bias. Further studies across various conditions are needed to confirm our findings.