How does vaccine effectiveness measured against hospitalization differ from effectiveness against infection for influenza H3N2?
Executive summary
Vaccine effectiveness (VE) against influenza A(H3N2) typically shows higher protection against severe outcomes like hospitalization than against any infection or mild clinical disease, a pattern observed across multiple seasons and networks where VE against H3N2 outpatient illness has often been lower and more variable while VE against H3N2-associated hospitalization is generally higher and more consistent in younger adults and children [1] [2] [3]. This divergence reflects differences in outcome definitions, host factors (age, comorbidities, immunocompromise), viral antigenic match, and study design, and it means a vaccine that only partly prevents infection can still substantially reduce the risk of severe disease requiring hospital care [4] [5] [6].
1. Why hospitalization and infection give different VE numbers
Hospitalization-focused VE studies measure protection against a severe, clinically significant endpoint—laboratory-confirmed influenza that leads to hospital admission—whereas infection or outpatient VE studies capture any symptomatic or laboratory-confirmed infection presenting to outpatient care or detected in surveillance, and those broader endpoints include mild cases that vaccines are less able to prevent; as a result, VE estimates against hospitalization are often higher because vaccines can blunt disease severity even when they do not fully prevent infection [1] [6] [5].
2. Real-world examples: H3N2 seasons and the pattern
Interim U.S. networks reported H3N2 VE around 42% in outpatient settings but about 55% against H3N2-associated hospitalization in the same surveillance network (NVSN), and other IVY network estimates showed 51% VE against H3N2 hospitalization while outpatient estimates were lower and sometimes not statistically significant [1]. Southern Hemisphere data and multicountry sentinel SARI studies likewise found lower VE against outpatient H3N2 illness (e.g., mid-30s to 40s) and somewhat higher protection against hospitalization in many groups, though exact numbers vary by season and population [6] [3].
3. Role of antigenic drift and subclades (subclade K example)
When H3N2 drifts antigenically—examples include recent subclade K—serologic reactivity to vaccine strains can fall, producing reduced VE against infection because neutralizing antibodies are less well matched, yet multiple data sources and WHO/Pan-American guidance emphasize that even in drifted seasons vaccination still tends to protect against severe outcomes and hospital attendance, likely via partial immunity, cross-reactive responses, or boosted cell-mediated immunity [4] [3] [7].
4. Age, waning, vaccine formulation and effect size
VE against hospitalization differs by age: studies have repeatedly shown stronger hospital protection in children and younger adults but weaker or inconsistent protection in older adults and immunocompromised patients, and waning of protection over months further reduces VE against both infection and hospitalization—often with a greater apparent drop for infection endpoints measured late in the season [2] [8] [3]. High‑dose or enhanced vaccines for older adults have produced better protection against both infection and hospitalization in randomized or large observational studies, underscoring formulation effects on these differing endpoints [9] [10].
5. Methodological drivers: test‑negative design and bias
Most VE estimates come from test‑negative case–control studies comparing vaccinated proportions among those testing positive versus negative; endpoint selection, case ascertainment, healthcare‑seeking behavior, and correlated vaccination behaviors (e.g., with COVID vaccines) can confound outpatient and hospitalization VE differently, sometimes inflating or attenuating VE for either outcome depending on how controls are chosen and how comorbidities are adjusted for (p1_s1; [6]; [1] snippets referencing confounding).
6. What this means for public health messaging and individual risk
Practically, the recurring empirical pattern—modest-to-moderate VE against H3N2 infection but stronger protection against hospitalization—supports recommendations to vaccinate every season because vaccines reduce the chance of severe disease even when they do not fully block transmission or mild infections; nevertheless, stakeholders should acknowledge group-specific limitations (older adults, immunocompromised) and consider enhanced vaccines and nonpharmaceutical measures during high H3N2 circulation or when a drifted subclade predominates [4] [5] [9] [10].