Background
We investigated the real-world effectiveness of a heterologous versus homologous primeboost
vaccination scheme against severe clinical outcomes of SARS-CoV-2 Omicron
infection among hospitalized patients in Belgium, and if this effect is homogenous across
different Omicron sub-lineages.
Methods
Observational data from existing national surveillance systems and administrative sources
were leveraged to estimate a causal effect. A Directed Acyclic Graph was formalized to
identify confounding factors. The target population consisted of adult COVID-19 patients
admitted to a Belgian hospital after the start of the booster vaccination campaign. Only
patients with a SARS-CoV-2 Omicron infection (BA.1, BA.2, or BA.4 and BA.5 combined
(BA.4/5) sub-lineages) confirmed by whole genome sequencing and vaccinated with a
homologous (mRNA for prime/mRNA for boost) or heterologous (viral vector for
prime/mRNA for boost) prime-boost vaccination schedule were included in the study
population. An adjusted logistic model was build and g-computation was executed to
estimate counterfactual risks of different clinical severity outcomes (intensive care unit (ICU)
transfer, in-hospital mortality, and severe COVID-19 – defined as experiencing an acute
respiratory distress syndrome (ARDS) event, or an ICU transfer, or in-hospital mortality) for
patients that received a homologous versus heterologous prime-boost vaccination, and to
contrast those. The role of the Omicron sub-lineage as an effect modifier was investigated by
estimating and contrasting these counterfactual risks across different sub-lineages.
Results
756 patients (499 homologous, 257 heterologous prime-boost vaccination) were included in
the study population. No significant difference in counterfactual risks for severe COVID-19
(RD=-0.04, 95%CI [-0.10–0.02]), ICU transfer (RD=-0.03, 95%CI [-0.07–0.02]), or inhospital
mortality (RD=0.00, 95%CI [-0.05–0.04]) was found between homologous versus
heterologous prime-boost vaccination, and no significant effect modification of the Omicron
sub-lineage could be observed.
Conclusions
Our results suggest that there is no difference in protection against severe disease among
hospitalized COVID-19 patients offered by the contrasted prime-boost vaccination schemes.
These results can inform vaccine policy, and contribute to knowledge used for the
development of strategies for booster campaigns during the continuous COVID-19 crisis.
