Update 2021-08-21: The Delta variant has now become the dominant strain, but apparently does no more damage than earlier variants. It is at least a hundred times more infectious than earlier variants. Since even in vaccinated people it takes time for the immune system to react, so-called "break-through" infection in vaccinated people are fairly common. But unvaccinated people are about seven times more likely to be hospitalised, and to die.
Evolutionary theory predicts this: Any mutation that improves the odds of infecting a host will eventually dominate. The same theory predicts that more variants will arise, but if they are less infectious than Delta, they will not spread. The most worrisome possibility is that Delta could morph to be more damaging.
Since none of the vaccines prevent infection 100% of the time, the virus (SARS-COV-19) will continue to infect people, and mutated strains will continue to arise. Sooner or later, there will be a strain that the immune response to vaccines will not be able to fight efficiently, which means that the infected people will be shedding virus longer, which will hasten the spread of that strain. There is of course the likelihood that the immune response will also change, but that change will be slow, and so such new strains will at first infect a lot of people.
So we will have a virus that mutates regularly, and epidemics will come and go. Covid-19 will become an endemic disease like the flu. Both will cause epidemics at fairly regular intervals. The evidence so far indicates that the covid epidemics will be worse.
NB: I have corrected an omission in the discussion of false test results below. The actual effect of a test error rate depends on whether it's the same or different for positive and negative results.
Update 2021-01-25: There are two worrisome new variants of the virus. One identified in the UK, the other in South Africa. Both are more infectious, and both may be more lethal, but the vaccines apparently work as well as with the original variants. The vaccination programs have hit predictable snags in production and logistics, causing a great deal of pointless finger-pointing.
Today WHO announced that covid-19 "may" become endemic.
Huh?????
No "may" about it. It will become endemic. The reason? Several, actually.
a) The wide range of responses to exposure to the virus, which vary from zero symptoms to serious and lethal illness to serious and lethal complications, to long-lasting after-effects. People with zero to mild symptoms are infectious.
b) Mutation. Current research on the mutated strains suggests that the virus will mutate continually, which means that by the time a vaccine or drug is developed, it may not work. Mutation also implies that vaccines and drugs would have to be updated at intervals, and that some future strains may be even more lethal than the current ones.
c) Variable immunity to the virus, which means re-infection and recurrence of the illness is certain. In fact, it's already happening. Also, the presence of antibodies does not guarantee immunity.
d) Variable incubation stage. It ranges from about 5 days to about 14 days after exposure. Also, repeated exposure seems to be a factor.
e) Variable asymptomatic but infectious stage, which follows incubation. Current data suggest two to four days.
f) Error rate of antibody test. It's about 90% accurate. Suppose an actual infection rate of 5%, then the false positive rate will close to 70%. (See the Wiki article on the base rate fallacy and the Footnote below.) This means that the anti-body test will generate a false sense of security.
These factors add up to guaranteed infections in future. The only unknowns are the future rates of infection, what factors affect those rates, and how lethal future infections will be. Experience with the flu shows that every now and then a more lethal strain will emerge. Current data suggest that covid-19 will be somewhere between twice and five times as lethal as the flu.
We will have to learn to live with it. It's likely that covid-19 will always be worse than the flu.
Footnote: Why a 90% positive accuracy isn't very good.
Suppose a 5% infection rate. Suppose a test with 90% accuracy for both positives and negatives. Then, in a population of 100,00, we will have:
Actual positives: 5,000
Detected actual positives: 4,500
Actual negatives: 95,000
Negatives detected as false positives: 9,500
Total positives detected: 9,500 + 4,500 = 14,000
Percentage of false positives: (9,500/14,000) x 100 = 67.8%
So if you get a positive test, it's more than twice as likely that you're negative than you're positive.
No comments:
Post a Comment