0

I'm a non-medical data scientist needing to quickly familiarise myself with epidemiological methodologies. (Don't worry I'm not going anywhere near modelling..)

The definitions of case fatality rate and mortality rate are very clear. I have not managed to find any references to a standard epidemiological measure of the broader fatality rate of a disease - i.e. an estimate of how deadly it is when also accounting for estimates of non-diagnosed infections (probably asymptomatic) cases. Is there a standardised term for this that I can research?

  • Could it be Cause-specific mortality rate? Ref cdc.gov/csels/dsepd/ss1978/lesson3/section3.html – geotheory Mar 19 '20 at 23:34
  • If you want to make that an answer you can always answer your own question. But answering in comments is frowned upon because comments aren't vetted and are subject to deletion at any time. – Carey Gregory Aug 20 '20 at 0:30
  • Thanks Carey. It was a genuine question, not an answer. I'm not new to StackExchange :) – geotheory Aug 20 '20 at 9:49
2

There is a “standard” definition of the epidemiological measure of the case fatality rate (or case fatality ratio—preferred term) for a disease. This is the number of people diagnosed with a disease who died because of the disease in an interval of time after the diagnosis divided by the number of people who had been diagnosed with the disease in the same interval.

https://www.britannica.com/science/case-fatality-rate

There is also a standard definition for the mortality rate from a disease. This is the number of people who died from the disease in a specific period of time divided by the number of people who could have died from the disease in the same interval of time—the total population.

https://www.cdc.gov/csels/dsepd/ss1978/lesson3/section3.html

EXAMPLE FOR NON-INFECTIOUS DISEASE

For these two measures (the case fatality ratio and the mortality rate), an example for a non-infectious disease—pancreatic cancer in 2016--is useful.

The one-year case-fatality ratio for pancreatic cancer is the number of people diagnosed with pancreatic cancer in (for example) 2016 who died in the year after their diagnosis divided by the number of people diagnosed with pancreatic cancer in 2016.

In 2016, about 53,000 people were diagnosed with pancreatic cancer in the United States. In the year after diagnosis, about 38,000 of these people had died of pancreatic cancer. The one-year case fatality ratio for people with diagnosed with pancreatic cancer in 2016 was 0.72 or 72%.

The (crude) mortality rate from pancreatic cancer in 2016 in the United States is the number of people who died from pancreatic cancer in 2016 divided by the number of people in the United States in 2016.

In 2016, about 42,000 people died from pancreatic cancer. The population of the United States in 2016 was about 323,000,000. In 2016, the crude mortality rate for pancreatic cancer was 13 per 100,000 (it is common practice to state mortality rates for a large population per 100,000).

INFECTIOUS DISEASES

For many infectious diseases (e.g., polio, influenza, hepatitis C), there is a “pool” of people that have been infected and but who are not diagnosed as having the disease and do not know they have been infected. This is because they don’t have symptoms or because the symptoms are minimal and the infection is unrecognized.

For infectious diseases, a third measure of mortality is the “infection fatality ratio.” This is the number of people who died from the disease in a given time interval divided by the number of people who have been infected in the same time interval.

The number of people infected in a given time interval—the denominator--is usually difficult to measure directly. It requires information on point or period prevalence of infection.

The infection fatality ratio can be estimated if there is reliable information about the proportions of people infected who are asymptomatic, who are minimally symptomatic, who develop the potentially fatal disease/condition caused by the infection, and, the proportion of those with the potentially fatal disease/condition who actually die. The latter— the proportion of those with the potentially fatal disease/condition who actually die—is the case fatality ratio.

INFECTION FATALITY RATIO EXAMPLE FOR INFECTIOUS DISEASE

An example—poliomyelitis--helps.

Polio is a highly infectious disease caused by the polio virus. It mainly affects children under the age of 5. After decades of study, much is known about the course of polio infection in children. About 70% are asymptomatic, 25%of have people infected with the polio virus have minimal or non-specific symptoms. About 0.5-1% develop acute flaccid paralysis. Of children who develop acute flaccid paralysis, 2-10% die.

https://www.who.int/news-room/fact-sheets/detail/poliomyelitis

https://www.aappublications.org/news/2020/03/18/mmwr031720

With these known facts, the infection mortality ratio for polio in children less than age 5 years can be estimated.

In 100,000 children less age 5 years, infected with polio:

Estimated cases of acute flaccid paralysis (lower estimate) = 100,000 x 0.005 = 500

Estimated cases of acute flaccid paralysis (upper estimate) = 100,000 x 0.01 = 1,000


Deaths in children with acute flaccid paralysis (lower estimate) = 500 x 0.02 = 10

Deaths in children with acute flaccid paralysis (upper estimate) = 1,000 x 0.10 = 100


Infection fatality ratio (lower estimate) = 10 / 100,000 = 0.0001

Infection fatality ratio (upper estimate) = 100 / 100,000 = 0.001
  • Thanks Diana for an expansive answer. 'Infection fatality rate' is exactly what I was after. I note it is now detailed in the wikipedia page for CFR (though it was absent when I asked this question!) – geotheory Aug 20 '20 at 13:20
  • Glad that wikipedia now includes this. People have been very confused! – Diana Petitti Aug 20 '20 at 17:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.