Incidence & Prevalence Calculator
Calculate the three basic measures of how common a disease is: cumulative incidence (risk), the incidence rate per person-time, and point or period prevalence. Each comes with a 95% confidence interval — Wilson for proportions, exact Poisson for rates — expressed per 100, 1,000, 10,000 or 100,000, with the working shown. Free, no sign-up, and nothing leaves your browser.
Type numbers without thousands separators: 9400, not 9,400.
Enter the number of cases and the population to see the result.
How to Use the Incidence & Prevalence Calculator
- 1. Pick the measure. Cumulative incidence for new cases among a group followed from a fixed start; incidence rate when people were followed for different lengths of time; prevalence for existing cases at a moment or over a period.
- 2. Enter the counts. New or existing cases, and the matching denominator: people at risk at the start, person-time at risk (in years, months, weeks or days), or the population.
- 3. Choose the base. Express the result per 100 (a percentage), 1,000, 10,000 or 100,000. Rare conditions read best per 100,000; common ones per 100.
- 4. Read the estimate and interval. The calculator shows the 95% confidence interval and the working, and copies a one-line result for your report.
Incidence vs Prevalence: Three Measures
| Measure | Numerator | Denominator | Answers |
|---|---|---|---|
| Cumulative incidence (risk, attack rate) | New cases during a period | People at risk at the start | What is the chance of getting the disease over this period? |
| Incidence rate | New cases | Person-time at risk | How fast do new cases arise? |
| Point prevalence | Existing cases at one moment | Population at that moment | What share has the disease now? |
| Period prevalence | Anyone with the disease during a period | Average or mid-period population | What share had it at any time in the period? |
- Cumulative incidence = new cases / people at risk at the start. It is a proportion between 0 and 1 and must always be quoted with its time period — "6% over five years". An attack rate is the same measure for a short outbreak.
- Incidence rate = new cases / person-time at risk, where person-time adds up each person's time free of the disease and under observation. Its unit is "per person-year", so 12.77 per 1,000 person-years is a rate, not a percentage.
- Prevalence = existing cases / population. It mixes how often the disease starts with how long it lasts.
Proportions — cumulative incidence and prevalence — get Wilson score intervals. The incidence rate treats the number of cases x as a Poisson count and uses the exact (Garwood) limits, χ²(0.025; 2x)/2 and χ²(0.975; 2x + 2)/2, divided by the person-time; that interval is valid even for very few cases.
Worked Example: Diabetes in a Five-Year Cohort
A cohort of 2,000 adults without diabetes is followed for five years, and 120 develop it. Press Load example in each mode to follow along.
| Measure | Calculation | Result | 95% CI |
|---|---|---|---|
| Cumulative incidence | 120 / 2,000 = 0.0600 | 60.00 per 1,000 over 5 years | 50.41–71.28 |
| Incidence rate | 120 / 9,400 person-years = 0.012766 | 12.77 per 1,000 person-years | 10.58–15.26 |
| Point prevalence | 340 / 4,000 = 0.0850 | 85.00 per 1,000 (8.50%) | 76.75–94.05 |
The cohort contributes 9,400 person-years rather than the full 10,000, because people stop contributing time when they are diagnosed or lost to follow-up. For the rate, the exact Poisson limits for 120 cases are 99.49 and 143.49 cases; divided by 9,400 person-years they give 10.58 to 15.26 per 1,000 person-years. The prevalence row comes from a separate survey of the same town on one day, in which 340 of 4,000 adults had diabetes.
How Incidence and Prevalence Are Linked
When a population is stable, prevalence odds equal incidence rate times average duration: P / (1 − P) ≈ I × D. A disease that starts often but resolves quickly, like the common cold, has a high incidence and a low prevalence. A chronic disease such as diabetes accumulates: even a modest incidence produces a high prevalence because people live with it for decades. That is also why better treatment that prolongs survival raises prevalence without any change in incidence.
Common Mistakes
- Leaving prevalent cases in the incidence denominator. People who already have the disease are not at risk of developing it; remove them first.
- Quoting a risk without its time period. "6%" means nothing until you add "over five years".
- Calling an incidence rate a percentage. A rate has units of person-time and can exceed 1 per person-year for recurrent events.
- Using cumulative incidence with heavy losses to follow-up. If many people leave early, the risk is underestimated; an incidence rate uses the time they did contribute.
- Comparing crude rates between populations with different age structures. Age-standardize before comparing.
Frequently Asked Questions
- What is the difference between incidence and prevalence?
- Incidence counts new cases arising in a period among people who did not have the disease; prevalence counts everyone who has the disease at a moment or during a period, however long ago it started. Incidence measures risk and is the measure for studying causes; prevalence measures burden and is the measure for planning services.
- How do I calculate person-years?
- Add up the time each person was observed and at risk. Someone followed for five years without the disease contributes 5 person-years; someone diagnosed after two years contributes 2, and someone lost to follow-up after three and a half years contributes 3.5. With grouped data, the population at the midpoint multiplied by the length of the period is a common approximation.
- What is the difference between incidence rate and cumulative incidence?
- Cumulative incidence divides new cases by the number of people at risk at the start, giving a risk over a stated period. The incidence rate divides new cases by person-time, giving a speed: cases per person-year. They agree closely when the disease is rare and follow-up is complete; the rate is the better choice when people are followed for different lengths of time.
- Should I report per 1,000 or per 100,000?
- Choose the base that gives readable numbers, usually one to three digits before the decimal point. Common conditions read naturally as percentages or per 1,000; rare ones, such as many cancers, are conventionally reported per 100,000. Keep the same base when you compare groups.
- Why are there different confidence interval methods for rates and proportions?
- A proportion counts how many of n people have the outcome, so it follows a binomial distribution and gets a Wilson interval. A rate counts events over person-time with no fixed upper limit, so the count is treated as Poisson and the exact chi-square-based interval applies. Both stay valid with small counts, where the simple normal approximation fails.
- What is period prevalence?
- Period prevalence counts everyone who had the condition at any time during a period — those who already had it at the start plus new cases — divided by the population, usually the average or mid-period population. Lifetime prevalence is the extreme case: anyone who has ever had the condition.