RESUMO
BACKGROUND: There is little information on relationships between indirect and direct protection by COVID-19 vaccination on close contacts of the vaccinees. Here, we assessed effect modification of direct-indirect action influencing the protective effects of vaccination. METHODS: Secondary attack rates (SARs) in household contacts (n = 2422) depending on vaccination status of the index cases (n = 1112) with known vaccination history during the delta variant-dominant period (August 2-November 2, 2021) in two public health jurisdictions were calculated using multivariable logistic regression analysis to assess indirect protection by COVID-19 vaccination as adjusted odds ratios (aORs) for SARs. The impact of the time of index case vaccination on indirect-direct protective effects was also assessed. FINDINGS: Contacts of index cases receiving 2× COVID-19 vaccinations showed significantly lower SARs than contacts of unvaccinated index cases (aOR:0.48, 95 %CI = 0.32-0.74). Relative to contacts where neither index cases nor contacts themselves were vaccinated (0,0), those with (2,0), (0,2) and (2,2) had lower SARs (0.45, 95 %CI = 0.24-0.82, 0.24, 95 %CI = 0.17-0.032, 0.11, 95 %CI = 0.06-0.20, respectively. No significant interactions on the SARs regarding times of vaccination between index cases and household contacts were observed, indicating additive but not synergistic protection. INTERPRETATION: The indirect protective effects of COVID-19 vaccination were attributed to an additive effect together with the direct effect on onward transmission in the household setting. These findings emphasize the importance of herd immunity by COVID-19 vaccination not only for unvaccinated but also vaccinated individuals.