Outbreak Attack Rate

Understanding the Attack Rate in Outbreak Investigation

The attack rate tells you the cumulative incidence of a disease in a defined exposed population over the course of one outbreak. You work it out with attack_rate = cases / exposed_population × 100%, applied to a closed cohort across the outbreak window. Where an incidence rate leans on person-time, this one doesn't. It's a plain proportion that sums up the cumulative risk carried by everyone who was exposed.

In epidemiology, attack rates do most of the heavy lifting in an outbreak investigation, since they put a number on how readily a pathogen spreads in a given setting. To give you a feel for the range: norovirus at parties or on cruises runs 60–80%, influenza within households sits around 30–50%, the COVID-19 household secondary attack rate reached 50–70% with Omicron, foodborne Salmonella from shared meals lands at 30–60%, and measles in unvaccinated school cohorts climbs above 90%.

Applications

SUS surveillance, CDC EIS officers and WHO outbreak teams reach for it to pin down vehicles of transmission, whether food, water or person-to-person, to compute attack rate ratios in cohort studies, to gauge vaccine effectiveness, and to decide where control measures should land. When the attack rate is much higher among people who ate a particular food than among those who didn't, that contrast is what cracks open a foodborne outbreak.

FAQ

What is the difference between attack rate and incidence rate? The attack rate is a proportion with no time in the denominator, which suits a closed outbreak. The incidence rate brings in person-time and fits open populations followed over longer stretches.

What is a secondary attack rate? It's the share of susceptible contacts of a primary case who go on to get sick, which gives you a direct read on how the pathogen moves within a household or among close contacts.

How is the exposed population defined? That changes with each outbreak. It might be the guests at a wedding, the students in a classroom, the staff on a shift, or the household members of an index case. Whatever the setting, the denominator has to take in everyone who could plausibly have been exposed to the source.