Victoria's 2021 Vaccine Drive Averted 316,000 Infections. Half the Credit Belongs to People Who Skipped the Shot.

Direct protection from a COVID-19 vaccine is the easy part to measure. Track who got a shot, track who got infected, compare the two groups. Indirect protection, the shielding that spreads outward when enough people nearby are vaccinated, has resisted that kind of accounting for years. A new analysis in the Journal of the Royal Society Interface says it isolated that hidden effect in a single Australian state's 2021 outbreak. The team, led by mathematicians at RMIT University in Melbourne, put a population-wide number on it for the first time.

The researchers were led by RMIT's Lewi Stone, working with Lixin Lin and Haydar Demirhan. James Trauer of Monash University joined the team. So did Peter Eizenberg, a general practitioner at Melbourne's Doctors of Ivanhoe. They modeled Victoria's Delta-variant wave from September 26 to November 21, 2021, across the state's 6.44 million residents. Then they compared that real outbreak to a modeled version with no vaccination at all. The campaign, they estimate, averted roughly 316,000 infections, 33,500 hospitalizations, and 4,900 deaths. That is four to eight times what Victoria actually recorded that fall.

Splitting credit between two kinds of protection

Close to half of those averted infections, the paper reports, trace to indirect protection rather than a person's own vaccination status. Eighty-five percent of that indirect share landed on people who were themselves unvaccinated. They were protected because the virus ran out of vaccinated carriers to pass through on its way to them. For hospitalizations and deaths, indirect protection's share drops to about a quarter. The authors say the gap makes sense. Direct vaccination does two jobs at once: it blocks infection, and it reduces how sick a breakthrough case turns out to be. Indirect protection only does the first job.

Picture the virus moving through a population the way water moves through a field of channels. A vaccinated person acts like a patch of dry ground. A chain of infection that reaches one stops there, whether or not that person would have gotten sick anyway. Scatter enough dry patches through the field and more channels run dry before reaching anyone, vaccinated or not. That is the mechanism the Victoria data let the researchers measure directly instead of just assuming.

Why this took an unusual state to measure

Population-level evidence for indirect vaccination effects has been thin. Earlier attempts in Israel and in Brazil tried to isolate the same signal. Both came back weaker or more uncertain, limited by how much natural immunity already existed in those populations before vaccines arrived. Victoria was different. Prolonged lockdowns had kept the state's COVID exposure close to zero heading into the Delta wave. By late 2021, almost all of the population's immunity came from vaccination rather than prior infection. That gap let the RMIT-led team separate the two signals in a way earlier studies could not.

Where the doses go matters too

The paper also models a more even rollout. Actual coverage ranged from about 40 percent to more than 60 percent, depending on the region, across the 67 local government areas in the analysis. What if Victoria had spread those doses evenly instead? Rebalancing toward equal coverage, the model suggests, would have cut infections, hospitalizations, and deaths by another 23 to 25 percent. Indirect protection delivers its biggest returns in places where vaccination still has room to climb, not in places already well covered.

The headline numbers rest on a model, not a direct count. To estimate what would have happened with no vaccination, the researchers fit a curve to real vaccination levels. Those levels ranged from roughly 40 to 60 percent across Victoria's regions. The team then extended that curve down to zero, a point the data never actually reached. The authors describe their estimates as conservative. The result has not yet been replicated by another team at the same population-wide scale.

If the method holds up under scrutiny from other research groups, it gives health departments a new argument for vaccination funding. It also makes a case for evening out coverage between neighborhoods, using a share of the benefit that direct-protection numbers alone have always left out.