Incidence Rate Calculator (cases per person-time with exact 95% CI)

Understanding the incidence rate calculator — concept and principle

This calculator works out an incidence rate, the number of new cases of a disease divided by the total person-time at risk, with an exact 95% confidence interval. Epidemiologists, students and public health staff use it to summarise cohort studies, surveillance data and clinical follow-up.

Cases per person-time: with the default inputs the result is 4.24 per 1,000 (95% CI 2.51–6.69).

How it is calculated

Incidence rate = new cases ÷ person-time × multiplier

95% CI lower = χ²₀.₀₂₅(2k) ÷ 2 ÷ person-time × multiplier

95% CI upper = χ²₀.₉₇₅(2k + 2) ÷ 2 ÷ person-time × multiplier

k is the number of new cases; person-time is in person-years.
The multiplier expresses the rate per 1,000, 10,000 or 100,000 person-years.
When k = 0 the lower limit is 0.

The principle

People are followed for different lengths of time: some join late, some leave or die, and those who develop the disease stop contributing time at risk. Summing each person's time at risk gives a fair denominator. The count of new cases in that time is treated as a Poisson variable. Because the Poisson distribution is linked to the chi-square distribution, exact confidence limits for the expected count can be read from chi-square quantiles (Garwood's method), then divided by person-time.

Worked example

Default inputs: 18 new cases, 4,250 person-years, per 1,000 person-years.

Rate = 18 ÷ 4,250 × 1,000 = 4.24 per 1,000 person-years.
Lower limit = χ²₀.₀₂₅(36) ÷ 2 = 10.67 cases → 10.67 ÷ 4,250 × 1,000 = 2.51.
Upper limit = χ²₀.₉₇₅(38) ÷ 2 = 28.45 cases → 6.69.
Rate per person-year = 0.004235.

Each point is the calculator re-run with only “Total person-time at risk (person-years)” changed; the gold point is the default example. Change the inputs above and this chart follows your numbers.

Total person-time at risk (person-years)	Incidence rate
1,062	16.94
2,125	8.47
3,188	5.65
4,250	4.24
5,312	3.39
6,375	2.82
8,500	2.12

How to read the result

The rate tells you how fast new cases appear in the population. The confidence interval reflects uncertainty from the number of cases: with few cases it is wide. Compare rates only when they are measured in similar populations and periods, or after age standardisation.

Common mistakes

Using the number of people instead of person-time.
Counting time after a person already had the disease.
Confusing an incidence rate with incidence proportion (attack rate), which has no time in the denominator.

Keep in mind: This is a statistical tool. Interpreting differences between groups needs attention to bias, confounding and study design.

Frequently asked questions

What is a person-year?

One person followed for one year, or two people for six months each, and so on.

Why use an exact interval?

It works well with small counts, where the normal approximation can give poor or negative limits.

What is the difference between incidence and prevalence?

Incidence counts new cases over time; prevalence counts all existing cases at a point in time.

Sources: Centers for Disease Control and Prevention, Principles of Epidemiology in Public Health Practice, 3rd ed., Lesson 3 "Measures of Risk" · Garwood F, "Fiducial limits for the Poisson distribution", Biometrika 1936;28:437–442

Checked 9 Nov 2026. Centers for Disease Control and Prevention, Principles of Epidemiology in Public Health Practice, 3rd edition, Lesson 3 “Measures of Risk”, Section 2 “Morbidity Frequency Measures” (incidence proportion, incidence rate, person-time); exact Poisson confidence limits via the chi-square relation (Garwood, Biometrika 1936).

Your numbers