incidence rate

The US Centers for Disease Control and Prevention, in its Principles of Epidemiology in Public Health Practice, defines an incidence rate, or person-time rate, as a measure of incidence that incorporates time directly into the denominator. The numerator counts only new cases, in people who did not have the condition when counting began. The denominator is person-time: the sum of the time each person was observed, totalled for all persons, which the course describes as the total time the population was at risk of and being watched for disease. Its worked example follows 2,100 women for four years, finds 16 new cases of heart disease across 6,400 person-years, and reports the rate as 2.5 per 1,000 person-years.

Two other measures sit close to it and answer different questions. Incidence proportion is the proportion of an initially disease-free population that develops disease, becomes injured, or dies during a specified period of time, so in the same example 16 cases among 2,100 women is 7.6 per 1,000 over four years. Prevalence is the proportion of persons in a population who have a particular disease or attribute at a specified point in time, and unlike either incidence measure it counts cases that already existed. A rate and a proportion from the same study will not match, because one divides by time and the other divides by people.

The practical use of an incidence rate is that it puts a number on how common something actually is, which a relative measure never does. A hazard ratio of 1.29 says one group's rate was 1.29 times the other's and stops there. The incidence rates underneath it say whether that means 361 cases per 100,000 person-years against 218, or 36,100 against 21,800. Health reporting tends to print the ratio and leave the rates out, and the gap between how those two sound is the reason this site prints both.