Age-Adjusted Rate Calculator
Paste cases and populations for 18 age groups and get the age-adjusted rate by direct standardization — to the US 2000 standard population, the WHO World Standard, the European Standard Population 2013, or one of your own — with a Fay–Feuer confidence interval, the crude rate for comparison, and the same data under every standard. The standard populations come from the NCI, WHO and Eurostat. Free, no sign-up, and nothing leaves your browser.
Paste from a spreadsheet, or separate values with tabs, commas or spaces. An age-group name in the first column is optional. Enter 0 cases for an age group with none. With your own standard, add a third column holding that standard's counts. The example uses made-up numbers.
| Weight | |
|---|---|
| 0–4 | 0.0691 |
| 5–9 | 0.0725 |
| 10–14 | 0.0730 |
| 15–19 | 0.0722 |
| 20–24 | 0.0665 |
| 25–29 | 0.0645 |
| 30–34 | 0.0710 |
| 35–39 | 0.0808 |
| 40–44 | 0.0819 |
| 45–49 | 0.0721 |
| 50–54 | 0.0627 |
| 55–59 | 0.0485 |
| 60–64 | 0.0388 |
| 65–69 | 0.0343 |
| 70–74 | 0.0318 |
| 75–79 | 0.0270 |
| 80–84 | 0.0178 |
| 85+ | 0.0155 |
Paste cases and populations by age group to get the age-adjusted rate with a confidence interval.
How to Use the Age-Adjusted Rate Calculator
- 1. Paste your data. Eighteen rows, youngest first — 0–4, 5–9, … 80–84, 85 and over — each with the number of cases (or deaths) and the population or person-years. An age-group name in the first column is optional; enter 0 for a group with no cases.
- 2. Choose the standard. US 2000 for most US health statistics, the WHO World Standard for international comparison, ESP 2013 for European comparison — or supply your own counts in a third column.
- 3. Set the rate base and confidence level. Rates per 100,000 are customary for mortality and cancer.
- 4. Compare. The adjusted rate and its interval appear beside the crude rate, with a table showing the same data under all three standards.
Why Rates Need Age Adjustment
Most health outcomes rise steeply with age, so a town with many older residents has a high crude death rate even if each age group fares well. Direct standardization answers the question "what would this population's rate be if it had a standard age structure?" Each age-specific rate rᵢ = casesᵢ / populationᵢ is weighted by the standard's share of that age group, wᵢ, and the weighted rates are added: age-adjusted rate = Σ wᵢ rᵢ. Two populations adjusted to the same standard can then be compared without age getting in the way.
Treating the counts as Poisson, the variance is Σ wᵢ² casesᵢ / populationᵢ². The confidence interval uses the gamma method of Fay and Feuer (1997), as NCI's SEER*Stat and R's epitools do, which stays positive and has good coverage even with few cases. The crude rate gets an exact Poisson interval.
The Three Standard Populations
| Standard | Source | Under 15 | 65 and over |
|---|---|---|---|
| US 2000 | NCI SEER / NCHS, from the Census Bureau's 2000 projection | 21.5% | 12.6% |
| WHO World 2000–2025 | Ahmad et al., WHO GPE Discussion Paper 31 (2001) | 26.1% | 8.2% |
| ESP 2013 | Eurostat task force report KS-RA-13-028 (2013) | 16.0% | 19.5% |
The standards differ most in how much weight they give the old and the young. Because death rates are far higher at older ages, a standard that weights the oldest groups heavily gives a larger adjusted mortality rate; a young standard gives a smaller one. The choice is a convention — WHO's own paper says there is no conceptual justification for one standard over another — so what matters is that every rate in a comparison uses the same one, and that you name it. The older US 1940 and Segi world standards are not included.
Worked Example: A Made-Up Town
Load example enters invented deaths and populations for 18 age groups — 1,515 deaths in a population of 97,100, a crude rate of 1,560.2 per 100,000. Adjusting that one set of age-specific rates to each standard gives very different numbers:
| Standard | Adjusted rate per 100,000 | 95% CI |
|---|---|---|
| US 2000 | 1,346.1 | 1,278.5 – 1,416.9 |
| WHO World 2000–2025 | 831.2 | 787.7 – 877.7 |
| ESP 2013 | 1,965.3 | 1,865.9 – 2,068.8 |
The town's age-specific rates are the same in every row; only the weights change. That is why an adjusted rate is a comparison tool, not a count of anything that happened, and why a rate adjusted to one standard cannot be set beside a rate adjusted to another.
- Comparing rates adjusted to different standards. Always adjust both to the same one.
- Mismatched age groups. The age bands of your cases and your population must be identical; combine finer groups into the 18 used here.
- Using it for tiny populations. With only a handful of cases in some age groups, the age-specific rates are unstable and the interval wide; indirect standardization (the SMR) is steadier.
- Leaving out the standard's name. An age-adjusted rate without its standard is uninterpretable; report it every time.
Frequently Asked Questions
- What is an age-adjusted rate?
- It is the rate a population would have if it had the age structure of a chosen standard population. It is calculated as a weighted average of the age-specific rates, using the standard's age distribution as the weights, and it removes the effect of differing age structures when comparing populations.
- Which standard population should I use?
- Use the one your comparison group or publication uses. US 2000 is the standard for US vital statistics and cancer data, the WHO World Standard is common in international and global-health work, and ESP 2013 is the European standard that replaced the 1976 version. The best choice is the one that matches the rates you are comparing against.
- How do I calculate an age-adjusted rate by hand?
- Divide each age group's cases by its population to get the age-specific rate; divide each standard-population count by the standard's total to get the weight; multiply each rate by its weight and add the products. Multiply by 100,000 if you want a rate per 100,000.
- What is the difference between direct and indirect standardization?
- Direct standardization applies your age-specific rates to a standard age structure and gives a rate. Indirect standardization applies standard age-specific rates to your population's age structure and gives a ratio of observed to expected events — the standardized mortality ratio. Direct needs reliable age-specific rates; indirect copes with small counts.
- Why does the adjusted rate change so much with the standard?
- Because the weights change. A standard with more elderly people puts more weight on the high rates at old ages, which raises the adjusted rate; a younger standard does the opposite. The differences are a property of the standards, not a sign of an error.
- How is the confidence interval calculated?
- The interval is the Fay–Feuer gamma interval, computed from the adjusted rate, its Poisson variance and the largest ratio of weight to population among the age groups. It is the method used in NCI's SEER*Stat. It never goes below zero and is accurate when some age groups have few or no cases.