Case Fatality Rate
Exam scores look completely different depending on whether you divide by the test-takers or the whole class. This is all about what goes on the bottom.
Definition The case fatality rate is the proportion of confirmed patients who die from a disease. Because the bottom number counts only diagnosed cases, it looks very different from dividing by everyone infected or the entire population. It only makes sense when you know who was tested and when.
Why the Same Disease Can Have Different Numbers
Let's imagine a fictional town to see how the math works. In this town, 100 people are officially diagnosed with an illness, and 2 of them pass away. If you put confirmed cases on the bottom, the rate is 2 out of 100 (2%).
Now, imagine 900 other people caught the bug, had mild symptoms, and never got tested. That means 1,000 people were actually infected. The number of deaths is still 2, but if you put total infections on the bottom, it becomes 2 out of 1,000 (0.2%).
If 10,000 people live in this town, the death rate across the entire population during this outbreak is 2 out of 10,000 (0.02%). The number on top never changed, but the bottom number changed, shifting the final result three times.
These numbers are purely made up to show the math, not real data from an actual disease. Confirmed cases depend entirely on how many people you find. Expand testing and contact tracing, and the entire denominator shifts.
Why Numbers Fluctuate During an Outbreak
During the peak of an outbreak, reported rates wobble in two different directions. But these two shifts compare against different benchmarks. They do not simply bounce above and below one single true value.
One shift comes from who gets counted on the bottom. Early on, testing focuses on severely ill patients. People with mild cases never make it into the confirmed tally. This makes the rate look higher than the actual rate among all infected people.
The other shift comes from the top not being full yet. A newly diagnosed patient is already counted on the bottom. However, if some of them later die, they have not yet been added to the top. Disease outcomes take time. Because of this time lag, the interim rate can look lower than what you would get after every patient's case resolves.
That is why numbers reported in the heat of an outbreak swing back and forth. They only settle down once testing widens and patient outcomes are complete. When different regions report different rates for the same illness, a big reason is how far they reached when counting the bottom.
Getting More Precise
Distinguishing three terms prevents a lot of confusion. Putting confirmed cases on the bottom gives the case fatality rate (CFR). Putting all infected people on the bottom gives the infection fatality rate (IFR). Putting the whole population on the bottom over a given time gives the mortality rate. Before debating the name, you must look at what sits on the bottom.
Even comparing CFRs between two places is tricky. The number depends on the cutoff date and the specific demographics. If the age breakdown differs or if one area tested far more people, you are not measuring with the same ruler.
There is also something this number cannot tell you. It only measures how dangerous the illness was for confirmed patients, not how easily the pathogen spreads. Contagiousness is measured by a completely different metric—how many people one infected person passes it to on average (the reproduction number, R).
The test score analogy holds up well here. Both depend on what you choose as your denominator. But unlike an exam with an official roster of test-takers, many infected people go uncounted, meaning the true denominator must be estimated.
🤔 Common misconceptions
The case fatality rate is your personal chance of dying if you catch the disease.
The bottom number only counts officially confirmed patients. People who recover without being tested are left out of the denominator, so this number is different from the risk across all infections.
A higher case fatality rate means the disease spreads faster.
This metric only measures the proportion of confirmed cases that end in death. How fast it spreads is tracked by transmission metrics, like the reproduction number.
The rate announced early in an outbreak is the disease's true fatality rate.
Early numbers swing in two directions. Limited testing makes the rate look higher than the total infection rate, while the time lag before deaths occur can make it look lower than the resolved rate.
🧺 Where you meet it
Because it divides deaths by confirmed cases rather than all infected people or the total population, the case fatality rate reflects testing practices just as much as disease severity.