Moving towards a precision approach for prevention of severe COVID-19

Replication of results in science is always reassuring, so we were pleased to see that Utkarsh Agrawal and colleagues,¹ using data from all regions of the UK, identified nearly identical risk factors as we did for severe COVID-19 despite vaccination among a nationwide cohort of US veterans.² An advantage of our study was the analysis of multiple subgroups, which allowed estimation of absolute risks on the basis of age and specific details about immune-compromised status. Advantages of the study by Agrawal and colleagues include the use of a variable that summarises the number of severe comorbidities at the patient level, a study population with large numbers of patients who had received booster vaccines, and a subanalysis limited to patients who had received boosters—which showed similar relative risks to what had been observed in analysis of the entire vaccinated population.

Data increasingly support the hypothesis that there are so-called vaccine complete responders, who most likely do not require frequent re-dosing; partial responders, who would benefit from re-dosing; and limited or non-responders, for whom we desperately need alternative prevention options, such as effective pre-exposure or post-exposure prophylaxis.³ Monoclonal antibody strategies have been shown to be ephemeral, and more advances are needed in this space.⁴ The large UK dataset used by Agrawal and colleagues might be useful for identifying these vaccine response phenotypes, through subanalyses stratified simultaneously by age and the number of comorbidities, with separate analyses for boosted or unboosted patients. Findings could be used to inform practice regarding vaccine distribution campaigns, targeting those who are likely to derive substantial clinical benefit from additional vaccine doses.