Two new cases and two different denominators
A fictional primary-care team is reviewing a closed cohort of 1,000 adults who were free of a specified condition at the start of a year. During follow-up, ten people develop the condition. The records also report exactly 980 person-years of time at risk because entry, event and follow-up dates were accounted for individually. A trainee labels both ten divided by 1,000 and ten divided by 980 as the same incidence rate and then adds a separate group of prevalent patients to the numerator. The supervisor asks for the quantities to be named correctly before any comparison with another service. The exercise supplies its own data and does not make a claim about the actual disease rate in any country.
1. What is the one-year incidence proportion in this cohort?
Read the worked answer
Ten divided by 1,000 equals 0.01, or 1%, under the supplied cohort definition. Existing cases from another group do not belong in the new-event numerator.
2. What is the incidence rate using the supplied person-time?
Read the worked answer
Ten divided by 980 equals approximately 0.010204 events per person-year, or 10.204 per 1,000 person-years. This is a rate using actual supplied person-time, not the same denominator as the incidence proportion.