A massive real-world study tracking over 3 million adults has delivered what might be the most definitive evidence yet that COVID-19 booster vaccines work exactly as advertised: they cut the risk of hospitalization and death by roughly half.
The research, published in the journal Vaccine, analyzed 3,464,877 adults eligible for booster vaccination during England's autumn 2022 campaign. Researchers at the University of Bristol and University of Oxford matched each boosted individual with an unboosted counterpart of similar age, vaccination history, clinical vulnerability, and location—then tracked them for a year.
The numbers are striking. Among boosted adults, the 350-day COVID-19 hospitalization risk was 3.78 per 1,000 people, compared to 6.81 per 1,000 among the unboosted. For deaths, the gap was even more dramatic: 0.29 per 1,000 versus 0.61 per 1,000. That translates to protection that halves your risk of the worst outcomes.
What makes this study particularly valuable is its scale and methodology. This isn't a carefully controlled clinical trial with cherry-picked participants—it's real-world data from NHS records covering millions of people going about their lives. The researchers used the OpenSAFELY platform, which links GP and hospital records, to track actual hospitalizations and deaths.
Both vaccines studied—Moderna's BA.1 mRNA-1273 and Pfizer-BioNTech's BA.1 BNT162b2—showed similar effectiveness. Protection was strongest in the first 70 days after vaccination, then gradually declined, which is consistent with what we know about immune responses.
Now, the study isn't perfect. Dr. Paul Madley-Dowd, the corresponding author, notes they couldn't account for every confounding factor. The researchers actually spotted one curious finding that highlights this limitation: boosted individuals showed a small reduction in fracture risk. Unless booster shots are somehow strengthening bones (they're not), this suggests some unmeasured health behaviors differ between groups.
That said, the core findings align with biological plausibility, previous smaller studies, and clinical observations. The study tracked 14,436 COVID-19 hospitalizations and 1,152 COVID-19 deaths—enough events to draw meaningful conclusions.
