Case Fatality Rate

The case fatality rate (CFR) is the proportion of persons diagnosed with a particular disease or condition who die from that disease within a specified time period. It measures the severity or lethality of a disease among those who have it, distinguishing it from the mortality rate, which uses the total population as its denominator.

The case fatality rate is one of the most commonly cited measures during infectious disease outbreaks and epidemics because it communicates how dangerous a disease is to those who contract it. Public health officials use CFR to compare the severity of different pathogens, prioritize response efforts, guide clinical management protocols, and communicate risk to the public. During the COVID 19 pandemic, for example, CFR estimates varied widely across countries and time periods, influencing public health policy and resource allocation decisions. Researchers studying disease prognosis, evaluating treatment effectiveness, and modeling disease dynamics all rely on accurate CFR estimates.

CFR is calculated by dividing the number of deaths caused by a specific disease during a defined period by the total number of individuals diagnosed with (or confirmed to have) that disease during the same period, then multiplying by 100 to express it as a percentage.

The accuracy of CFR depends on both the numerator and the denominator. The numerator requires accurate determination of cause of death, which can be complicated when patients have comorbidities or when death certification practices vary. The denominator requires accurate case ascertainment, which depends on the availability of testing, the sensitivity of diagnostic criteria, and the completeness of surveillance systems.

CFR can vary widely depending on several factors: the virulence of the pathogen, the age and health status of the affected population, the quality and availability of medical care, and the case definition used. For example, CFR for Ebola virus disease has ranged from 25% to 90% across different outbreaks, reflecting differences in viral strains, healthcare infrastructure, and the timing of medical intervention.

It is important to distinguish CFR from the infection fatality rate (IFR), which includes all infected individuals in the denominator, whether or not they were diagnosed. The IFR is usually lower than the CFR because its denominator is larger (including asymptomatic and undiagnosed cases), although a CFR can fall below the IFR when deaths are under recorded or have not yet occurred. During the early phase of an outbreak, when testing is limited, the CFR may substantially overestimate the true lethality of a disease.

CFR is appropriate for measuring disease severity among diagnosed cases, comparing the lethality of different diseases or different outbreaks of the same disease, and evaluating the effectiveness of clinical management over time. It is most informative when the case definition is standardized, diagnostic testing is widely available, and the follow up period is long enough to capture deaths that occur after a delay. CFR is not appropriate as a measure of population risk (use mortality rate for that) or as a measure of risk among all infected individuals (use IFR for that).

A common error is interpreting CFR as the probability that any infected person will die, when in fact it only applies to diagnosed cases. In diseases where many infections are asymptomatic or mild and go undiagnosed, the CFR can dramatically overstate the actual risk of death among all infected individuals.

Another mistake is comparing CFRs across countries or time periods without accounting for differences in testing capacity, case definitions, healthcare quality, and population demographics. A country that tests only severely ill patients will have a higher CFR than one that tests broadly, even if the underlying disease severity is the same. Researchers should also be cautious about calculating CFR early in an outbreak, when the final outcome (recovery or death) is not yet known for many cases; premature calculation can bias the estimate in either direction, because some current cases will die later (which makes the simple ratio of deaths to cases too low), while deaths are often recorded sooner than recoveries (which makes a ratio of deaths to resolved cases too high). Finally, conflating CFR with mortality rate, which uses the total population rather than the diagnosed population as the denominator, is a common but consequential error.

During an outbreak of a novel respiratory virus, a country with extensive testing capacity identifies 85,000 confirmed cases and records 1,700 deaths over three months, yielding a CFR of 2.0%. A neighboring country with limited testing identifies only 12,000 confirmed cases (primarily severe and hospitalized patients) and records 960 deaths, yielding a CFR of 8.0%. Public health officials note that the difference in CFR likely reflects differences in testing coverage rather than true differences in disease severity. A seroprevalence survey later estimates that 350,000 people in the first country were infected, yielding an IFR of 0.49%.