Calculator
Enter outbreak totals and optional adjustments.
Example data and your history
The first rows are sample scenarios; your submissions append below.
| # | Time window | Cases | Deaths | Recovered | Primary CFR (%) | Closed-case CFR (%) | CI (primary) | Deaths / 100k |
|---|---|---|---|---|---|---|---|---|
| 1 | Example: Baseline | 1,000 | 25 | 400 | 2.50 | 5.88 | 1.69–3.67 | — |
| 2 | Example: Expanded testing | 2,500 | 30 | 1,200 | 1.20 | 2.44 | 0.84–1.71 | — |
| 3 | Example: Under-ascertainment | 700 | 28 | 260 | 4.00 | 9.72 | 2.75–5.78 | — |
Note: sample confidence intervals are illustrative, using Wilson.
Formula used
CFR (%) = (Deaths ÷ Cases) × 100
Optional inputs can add probable cases and exclude non-attributable deaths to test sensitivity.
Closed CFR (%) = Deaths ÷ (Deaths + Recovered) × 100
Useful when outcomes are mostly resolved, but can be biased early.
For p = Deaths/Cases and z from the selected confidence level: CI = (p + z²/(2n) ± z·sqrt((p(1−p)+z²/(4n))/n)) / (1 + z²/n)
Wilson is typically more stable than the simple normal approximation, especially with small counts.
How to use this calculator
- Enter confirmed cases and attributed deaths.
- Optionally add recovered to view closed-case CFR.
- Use probable cases and excluded deaths to run sensitivity checks.
- Select a confidence level to compute a Wilson interval.
- Press Submit to show results above the form.
- Use Download CSV or Download PDF to export the table.
Why CFR remains a core outbreak metric
Case fatality rate (CFR) summarizes the proportion of identified cases that result in death. It is widely used in field reports because it can be computed from routine surveillance counts. However, CFR is not a biological constant; it changes with testing intensity, access to care, variant or strain differences, and reporting delays. This calculator standardizes the arithmetic and lets you document assumptions consistently across scenarios.
Inputs and common data quality checks
The required inputs are total confirmed cases and deaths. Optional fields support practical audits: probable cases can be added to test sensitivity, while excluded deaths remove events later judged unrelated. A validation toggle prevents impossible totals where deaths exceed cases. When comparing regions, ensure identical case definitions, comparable time windows, and consistent attribution rules for deaths.
Closed-case CFR for resolved outcomes
Closed-case CFR uses deaths divided by deaths plus recovered. This can be informative when most outcomes are resolved, such as late in an epidemic wave or within a well-followed cohort. Early in outbreaks it may inflate severity because recoveries can be under- recorded or delayed. Treat it as a complementary indicator rather than a replacement for the primary CFR.
Uncertainty and the Wilson confidence interval
CFR estimates are noisy when counts are small. To show uncertainty, this calculator reports a Wilson score confidence interval for the underlying proportion (deaths/cases) and displays it in percent terms. Wilson intervals tend to behave better than simple normal approximations, especially when the proportion is near 0% or when case counts are limited.
Using exports for reporting and review
Each submission can be appended to the history table, enabling side-by-side comparisons across time windows or assumptions. Use the CSV export for spreadsheets, dashboards, or audit trails. Use the PDF export for quick sharing in briefings. If you include population, the table adds deaths per 100,000 to contextualize severity with scale.
FAQs
1) What is the difference between CFR and infection fatality rate?
CFR uses detected cases as the denominator. Infection fatality rate includes all infections, including undiagnosed ones, and usually requires serology or modeling. IFR is typically lower than CFR when testing is limited.
2) Why can CFR change over time during one outbreak?
CFR can shift with case ascertainment, healthcare capacity, age distribution, treatment availability, and reporting delays. Expanding testing often increases detected mild cases, which can reduce CFR even if risk is unchanged.
3) How should I use the reporting lag field?
Use lag as a scenario label when deaths are reported later than cases. It does not modify surveillance data; it helps you document assumptions and compare results under different delay expectations.
4) When is closed-case CFR most appropriate?
It is most useful when follow-up is strong and outcomes are largely resolved, such as late in a wave or within a closed cohort. Avoid relying on it early, when recoveries are incomplete or delayed.
5) What does the confidence interval tell me?
It shows a plausible range for the true CFR proportion given your counts and chosen confidence level. Wider intervals indicate more uncertainty, typically due to small case numbers or very low event counts.
6) How do probable cases and excluded deaths affect results?
Adding probable cases increases the denominator and usually lowers CFR. Excluding deaths decreases the numerator and also lowers CFR. Use these fields to test how sensitive your conclusions are to classification changes.