Coronavirus fatality rate
 |
| Body wrapped in cardboard and plastic left outside for collection, Ecuador. |
The Case Fatality Rate (CFR) is the fraction of confirmed cases that result in death rather than recovery. It differs form the infection fatality rate (which includes all infections, not just those that have been detected - necessary for most forms of modelling) or the population fatality rate, which uses the population as the denominator, and is thus a better measure of epidemic severity).
A 27 March Lancet article found a wide disparity in case deaths, and this has continued to be observed as the pandemic progressed. The CFR on the Diamond Princess cruise ship where almost everyone was tested was about 1%. Very few countries have been able to meet this benchmark.
The expected progression is for deaths to start low, then as the cases stabilise, deaths begin to catch up so that the CFR rises, then as testing levels improve, the case fatality falls away again. When the medical system is overwhelmed and few test kits are available, almost no-one except severe cases is tested, and the CFR may go as high as 25% as has occurred in New York.
Unfortunately, CFRs do not fall much when there is limited testing at the height of the pandemic - because most of the low-severity cases recover before they are ever tested. Consequently, global CFRs have stayed high throughout the first stage of the pandemic
CFRs at 10 June were 5.7% globally, the same as the United States.
The rates in Europe have been the highest in the world, probably because of their older populations and the severe infections that occurred in aged care homes. CFRs have ranged from 19% in France, 16.7% in Belgium and 14.5% in Italy down to 4.7% in Germany, 4.0% in Austria and 2.8% in Norway.
In East Asia, the benchmark was set by the countries who first had the pandemic, China, Japan, Indonesia and Iran, all of whom had a CFR around 5%: . There have been no deaths at all anywhere in Indo-China, which largely managed to avoid the pandemic.
In Latin America the highest rates have been in Ecuador (8.4%) and Brazil (5.2%). The CFR in most other countries has been about 2.8%. The under-recording of deaths has probably been substantial.
(Source: Our World in Data)
Factors affecting the case fatality rate
The apparent variability of the coronavirus case fatality rate (CFR) has made it difficult to estimate the outcomes of the pandemic in advance. Some of the factors affecting the death rate are known to be- Age
- Presence of other serious conditions (heart, liver etc)
- Quality of medical care
- Sex
- Smoking, and other forms of lung damage
Age and sex
The case fatality rate varies a great deal across age
groups, roughly doubling per decade of age and trebling after the 50-59 age
group. . Males are almost twice as likely to die as women,according to one survey.
Table. Case fatality by age group, men and women, per cent.
30-39
|
40-49
|
50-59
|
60-69
|
70-79
|
80+
|
|
Males
|
0.43
|
0.91
|
2.05
|
6.67
|
19.71
|
30.6
|
Females
|
0.26
|
0.55
|
1.23
|
4.02
|
11.86
|
18.5
|
These substantial sex differences, established in February, are heavily loaded towards Chinese cases, and have been disputed because many more older Chinese men than women are smokrse.
Another estimate by Our World in Data shows CFRs by age and country, indicating encouragingly that the the CFRs by age are not actually so different in different countries.
Another estimate by Our World in Data shows CFRs by age and country, indicating encouragingly that the the CFRs by age are not actually so different in different countries.
More estimates from Verity et al in Lancet on 30 March adjust the raw China figures for censoring, demography and under-ascertainment to give CFRs of 1.25% for 50-59, 4% for 60-69, 8.6% for 70-79, and 13.4% for over 80. This is equivalent to 1.38% overall. They also estimate from British data the proportion of infected individuals likely to be hospitalised.
Confusion over what is a coronavirus death
System failure from coronavirus causes organs that are already damaged to shut down, so that many older people who die already have multiple co-morbidities that 'might have killed them'. As many countries have done (and have been criticised for so doing [3]) it is probably better to accept that a CV infection is a smoking gun, and actually responsible for the fatality.A number of countries, especially Russia and its surrounds, have only included deaths that can be directly attributed to COVID-19. These are probably less than half of the total as many patients die of system shocks affecting major organs. Consequently, CFRs in Russia, Ukraine and Belarus are considerably less than one might expect.
Effect of inadequate case data testing - infection fatality rates
The problems affecting case data can also make a very big difference to case death rates. If only severely ill patients have been tested, then the CFR is obviously going to be much higher then when people with mild infection are also identified. The more are tested, the more the case fatality rate approaches the true infection fatality rate. [1] For example, early CFRs in China were 17% in the first few weeks of January, but reduced to 0.7% in February before increasing again. Similarly, even though the case fatality rate in Italy is over 14%, one research team thinks the infection fatality rate could be as low as 0.1%. [2]
Even broader testing may not identify the correct CFR. If an
estimated 1% of the population are infected, about 3% show up as positive in
widespread testing in South Korea and Australia, and up to 40% in Italy, where only
the severely ill have been tested. Accordingly, the ratio of deaths to confirmed cases has ranged from 14% in
Italy down to less than 1% in some countries.[2]
The ratio of numbers dead to numbers recovered is steadier, reflecting only national differences in the risk factors above. People either recover or die at about the same time so this indicator is less affected by the progress of national epidemics..
A statistic that 21% of cases are Serious and 5% are Severe has been estimated from Chinese data. The exact meaning of these is probably non-standard in different countries.
The ratio of numbers dead to numbers recovered is steadier, reflecting only national differences in the risk factors above. People either recover or die at about the same time so this indicator is less affected by the progress of national epidemics..
Wrong data items are being collected
The sequence Infected=>Test positive => Hospitalised=>Severe/ICR=>Die/Recover is what we should like to follow. At the moment we have only the second and the last figures, which is hampering modelling.A statistic that 21% of cases are Serious and 5% are Severe has been estimated from Chinese data. The exact meaning of these is probably non-standard in different countries.
Comments
Post a Comment