High-Dose vs Standard-Dose Influenza Vaccine in Older Adults According to Prior Hospitalization Burden
J Christensen, N D Johansen, D Modin, M Loiacono, R C Harris, A Huang, K V Bartholdy, K F Bernholm, J I M Borchsenius, F S Davidovski, L W Davodian, M Dons, A M R Jensen, C Espersen, T Biering-SoerensenAbstract
Introduction
High-dose inactivated influenza vaccine (HD-IIV) has demonstrated superior protection against hospitalization outcomes compared with standard-dose (SD-IIV) among older adults. However, whether prior hospitalization burden modifies this effect has not previously been assessed.
Purpose
To investigate whether prior hospitalization burden modifies the relative vaccine effectiveness (rVE) of HD-IIV vs. SD-IIV against severe clinical outcomes.
Methods
This is a post-hoc exploratory analysis of DANFLU-2, an open-label, pragmatic, individually randomized trial conducted in Denmark from 2022-2025 to assess the rVE of HD-IIV vs. SD-IIV in the older general population. Hospitalization burden was defined as the cumulative number of all-cause hospitalizations (NoH) in the 5-year period prior to randomization. rVE was calculated as 1 minus the relative risk in the HD-IIV group compared with the SD-IIV group. Firstly, rVE was assessed against the primary and secondary endpoints of DANFLU-2 according to continuous NoH with restricted cubic splines to identify potentially relevant thresholds for NoH at which to dichotomize the population for subgroup analysis. Effect modification was assessed using interaction terms in log-binomial regression for continuous NoH and with the Cochran-Mantel-Haenszel test for homogeneity for dichotomous NoH. In this analysis, continuous NoH was log2-transformed to account for diminishing marginal effects of additional events.
Results
Among 332,438 participants (mean age 73.7±5.8 years; 48.6% female), 131,121 had at least one hospitalization in the 5-year period prior to randomization (median NoH: 0, IQR: [0, 1]). Those with NoH ≥1 were on average older and more likely to be male or suffer from comorbidities. For the primary diagnoses of these hospitalizations, the most frequent ICD-10 codes were cardiovascular (I; 11.9%), infectious (A or B; 11.7%), and musculoskeletal (M; 9.5%). NoH modified the effectiveness of HD-IIV compared with SD-IIV against hospitalizations for pneumonia or influenza, influenza, and cardiorespiratory disease, with p for interaction suggestive of increasing rVE in those with higher NoH (figure 1). Subjects were stratified according to NoH >2 for subgroup analysis. Here, we similarly found significant effect modification against the same endpoints, with higher rVE observed among those with NoH >2 (figure 2). Pneumonia and all-cause hospitalization along with all-cause mortality outcomes were not modified by NoH (p>0.05).
Conclusion
Our findings were suggestive that previous hospitalization burden might modify the effect of HD-IIV compared with SD-IIV against several outcomes, with higher effectiveness observed in participants with >2 hospitalizations in the 5-year-period prior to vaccination. Given the post-hoc exploratory nature of this analysis, findings should be considered hypothesis generating.Figure 1Figure 2