The (VE = ARU - ARV) / ARU Formula: How Vaccine Efficacy is Calculated from Attack Rates in Unvaccinated and Vaccinated Groups

On March 10, 2025, the World Health Organization published a standardized framework detailing how a vaccine's performance is quantified, drawing a strict line between controlled clinical trials and real-world deployment. The foundation of this measurement rests on a single mathematical expression used by epidemiologists worldwide to determine exactly how much protection an immunization provides.[2]

That expression is VE = (ARU - ARV) / ARU. In this formula, VE stands for Vaccine Efficacy, ARU represents the Attack Rate in the Unvaccinated population, and ARV represents the Attack Rate in the Vaccinated population. By comparing these two rates, researchers can isolate the specific biological impact of the vaccine against a baseline of natural infection.[1]

An attack rate is not a raw count of infections; it is a proportion. The Centers for Disease Control and Prevention defines it as the number of new cases of a disease in an at-risk population divided by the total number of people in that population over a specific time period. If 1,000 unvaccinated people are exposed to a virus and 100 contract the disease, the ARU is 10 percent.[1]

Applying the formula reveals the relative risk reduction. If the ARU is 10 percent and the ARV is 1 percent, the calculation becomes (0.10 - 0.01) / 0.10. This yields 0.90, or a 90 percent vaccine efficacy. This figure indicates that the vaccine reduced the risk of contracting the disease by 90 percent relative to the unvaccinated group.[1][6]

A common misconception is that a 90 percent efficacy means 10 percent of vaccinated individuals will inevitably get sick. In reality, it means an individual's personal risk of infection is cut by 90 percent compared to what it would be if they had not received the shot. The absolute risk depends entirely on the background prevalence of the disease in the community.[6]

The mathematical purity of the VE formula is strictly applicable only in Phase III randomized controlled trials. The WHO explicitly separates "efficacy" from "effectiveness." As the organization notes, "Vaccine efficacy measures how well a vaccine works when given under ideal circumstances," where exposure, dosage, and participant health are tightly monitored.[2]

Effectiveness, conversely, measures "how well a vaccine performs when it is used in routine circumstances in the community." Once a vaccine leaves the trial environment, the variables multiply. Different age demographics, underlying health conditions, and varying levels of exposure to the pathogen all alter the attack rates.[2][3]

The CDC's annual influenza vaccine monitoring provides a clear example of this shift from efficacy to effectiveness. Because the influenza virus mutates rapidly and randomized placebo trials are no longer ethical for an approved seasonal vaccine, the agency continuously measures how well the annual shot works using observational studies rather than the strict VE formula.[7]

To calculate effectiveness without a controlled trial, epidemiologists often turn to the screening method. This observational technique compares the proportion of disease cases occurring in vaccinated individuals against the overall proportion of the population that is vaccinated. It allows researchers to estimate protection levels using existing public health registries.[4]

The mathematics behind the screening method rely heavily on conditional probability. Researchers modeling these dynamics have demonstrated that effectiveness can be calculated using Bayes' theorem to adjust for varying exposure risks in the general population, compensating for the lack of a randomized control group.[8]

Proving efficacy in the first place requires precise statistical planning. Determining the necessary sample size for a Phase III trial depends entirely on the expected attack rate. If a disease is relatively rare, researchers must enroll massive cohorts—sometimes tens of thousands of participants—to observe enough natural infections to reach statistical significance.[5]

Field evaluations face the additional challenge of behavioral confounding. The basic VE formula assumes equal exposure to the pathogen across both the vaccinated and unvaccinated groups. If vaccinated individuals engage in riskier behavior because they feel protected—a phenomenon known as risk compensation—the ARV artificially inflates, lowering the calculated effectiveness.[3][8]

To account for these discrepancies, modern epidemiological models are moving beyond static percentages. Researchers are developing dynamic frameworks that integrate real-time mobility data and localized transmission rates to continuously adjust the ARU and ARV, tracking a vaccine's protective power as it evolves during an active outbreak.