Someone asked me a question as to what the actual Virginia flu data means since there have been fewer Covid-19 deaths in Virginia that from the seasonal flu in the 2018-2019 season. This person assumed that this implies that Covid-19 is no more dangerous than the seasonal flu. Here I answer that question by showing the Virginia data of flu and Covid actually demonstrate that Covid is many times more fatal than the flu.
Let me begin by giving the bottom line, and then you can work through my reasoning in detail. First, there have been only 2 months of Covid versus 8 months of 2018-2019 seasonal flu. Second, there have been only around 26,000 Covid cases in Virginia versus over 1 million seasonal flu cases in Virginia in 2018-2019. These numbers are consistent with Covid being at least 30 times more fatal in Virginia than the seasonal flu (the actual fatality rate varies by state and by country). These figures are consistent with the known worldwide data (which is less optimistic). This implies there would be well over 100,000 excess Covid deaths in Virginia (probably many more), if the disease were not mitigated (via shutdowns or vaccine), irrespective of the time period over which this happened (6 months or two years), IF 70 percent of the Virginia population were to become infected in order to reach “herd immunity,” and people die at the rate we currently have. These conclusions are all supported by hard data, consistent with data from other states and countries.
The detailed reasoning follows below, based on nothing but data-based simple arithmetic—no fancy modeling. My assumptions are laid out in detail so you can follow and check them, and you are welcome to challenge them. If anyone wishes to challenge my data, please do so, and I will give you my source and you can look at it yourself and make a judgment concerning it. You do not have to agree, but at least be fair. Please note that like most scientists, I will often round off to one or two digits, since magnitudes are what is important, not every last digit.
For the sake of simplicity, we need to know the fatality rate per unit number of cases. We could take cases per 100 people (percentage) or cases per 100,000 people, as it often is reported. Let me stick with the percentage. The problem with the data is that while it is easier to count the number of deaths, it is harder to know the actual number of cases, because of those with no or mild symptoms do not get counted. So the known number of cases is likely an underestimate. This is one place where we do not have good data yet. So I have to stick with what we know and go with the reported count. So far there are nearly 4 million cases worldwide and around 290,000 reported deaths (but not all who will die have yet died), or slightly more than 7 percent of cases, or to say it another way, 1 person dies for every 14 people who are reported to have the disease. This number varies with the country. The best case in Europe, Germany, has kept the fatality rate to around 4 percent (1 in 25), whereas in Italy, France, Spain, and the UK it has ranged between around 12 and 15 percent (1 in 6 or 1 in 7). Currently, in the USA (83000 deaths in 1.4 million cases), the fatality rate is around 6 percent (1 in every 17 cases). So far, we have been doing much better in Virginia, nearly 900 deaths out of 26000 cases, or a fatality rate around 3 percent, or 1 in 30, instead of the roughly 1 in 14 worldwide or 1 in 17 for the USA.
These numbers will be less severe, depending on the number of uncounted cases. For example, there are inconclusive data that suggests there may be between 2 and 10 times as many cases as are in the “official” count. That would drop the death rate from 1 in 14 (worldwide) to 1 in 30 to perhaps even 1 in 300, that is, 3 to 30 times that of the seasonal flu, which is 1 in 1000. But this more optimistic reading depends on very uncertain data that has not yet been verified or generally accepted. It is important to keep this uncertainty in mind since that could provide a basis for a more optimistic reading than I am giving you. I am going with what we know, not what we don’t know. You are free to be more optimistic, as long as you admit you might be wrong too. In the end, only the final data will tell.
As for the seasonal flu numbers from Virginia, since I did not find the total number of cases, let me first look at the national numbers and assume Virginia is similar. The CDC estimates that during the 2018-19 flu season in the US, 37.4 million to 42.9 million people got sick with the flu, 17.3 million to 20.1 million visited a doctor for their illness, 531,000 to 647,000 were hospitalized for the flu, and 36,400 to 61,200 died from the flu and its complications. Let’s take the mean of these figures, 49,000 deaths out of 40 million flu cases in the USA in the 2018-2019 season, giving a 0.12 percent fatality rate (1 in 820 cases); the 40 million cases corresponds to 12 percent of the total population of the USA in 2018. Virginia reported 1813 deaths from pneumonia and influenza in 2018-2019. If we take the fatality rate of 1 in 820 for the USA as a whole in 2018 to be the same as the rate for Virginia, the 1813 deaths would imply 820 times as many cases, or 1.5 million cases, out of a population of 8.5 million, which corresponds to 18 percent of the Virginia population (it was 12 percent of the US population, according to CDC). The 18 percent is a little higher than, but still similar to, the overall rate of 12 percent of the USA population having flu in the 2018-2019 season. This can be explained if not all of the 1813 deaths were due to the flu, say if 300 or so were due to other causes of pneumonia than the flu, or maybe Virginia had a seasonal anomaly.
These data then let us understand the meaning of the 2018-2019 Virginia flu data: 1813 deaths out of 1.5 million cases over a whole flu season of 8 months (October to May). This is to be compared to 891 deaths over 25800 cases occurring over essentially 2 months (March-April). If you scale the flu deaths to the same 2-month interval (don’t compare apples and oranges), it would drop from 1812 to a quarter that, or around 450 deaths as compared to nearly 900 from Covid. This implies 2 times more deaths from Covid relative to flu over a comparable time period. But the far more telling number is the number of cases, only 26000 Covid cases versus 1.5 million (estimated) flu cases, that is, there were over 50 times as many flu cases in 2018-2019 than we have had Covid cases so far (and I rounded down—it should be 58). Consequently, we need to scale up the number of Covid deaths that would occur if there were 1.5 million cases of Covid, as there were of the flu. Then, multiplying 50 times 891 deaths for 50 times more cases means over 44,000 Covid deaths would occur in Virginia if we had 1.5 million cases of Covid, similar to the number of cases of seasonal flu. This scaling of Covid to seasonal flu, or 44000 deaths out of 1.5 million cases, corresponds to a fatality rate of 2.9 percent, or 1 in 34, which is very close to the actual fatality rate in the current numbers reported by the Virginia Department of Health. Thus, the numbers I have extracted from the data are mutually consistent and all hang together.
IF 70 percent of the population of Virginia were to get Covid (and this is a big IF, but this is the number of cases assumed to give “herd immunity” to shut off disease transmission), the above numbers, based on actual real-world data (no lies here—you can look up all the numbers I have used above from available sources) taking the best estimate that 1 in 34 of the people who get infected in Virginia will die, tells us that out of a population of 8.5 million in Virginia, 175,000 Virginians will die of Covid-19, unless mitigation steps are taken to reduce the number of cases (maybe a vaccine before it runs it course). In the very best-case scenario, if the fatality rate were only 1 in 300 or 3 times that of the seasonal flu, around 20000 Virginians would die. I fervently hope the latter lower number will turn out to be the case once all the data are in. But I can’t make it be so by wishful thinking. I have to go with the data I have, not the data I don’t.
The last paragraph is assuming the low 1 in 34 mortality rate now current in Virginia (whether the correct value is 1 in 10—as in some European countries–or 1 in 100—being VERY optimistic–the number of deaths would still be enormous, and far worse than the seasonal flu). These numbers do not depend on some model of how fast infections occur (they depend on percentages and populations only). If cases occur slowly (as now under mitigation measures), the hospital system seems capable of handling it in Virginia. If infections occurred much more rapidly, with exponential growth with a short doubling time (2 days at the peak of the NY breakout, or a over a hundredfold increase in cases in a 2-week period), then the medical system would be threatened with a massive overload with untreated patients, deaths at home, and a death rate closer to 1 in 6, as happened in Northern Italy and Spain. Anyone who closes their eyes to this does so at their peril and to the peril of their family, friends, and loved ones.
I do not offer these numbers with any “political” agenda in mind or to cause fear. I simply speak as a scientist who has worked all my life with numbers and the modeling of complex systems, who wishes to save lives based on facts. I understand numbers and magnitudes, and I respect real and hard data. I respect above all the truth. We must all seek to get at the truth of reality, to be guided by what is real. That is at the core of my being as a scientist and as a follower of Jesus Christ. I want to save lives by encouraging wise actions and sensible mitigation measures based on what we know. I think much can be done to minimize unnecessary death and economic hardship. And we will be able to get our society and economy back to good health much more quickly and effectively, and save far more innocent lives that do not need to be lost, if we pay attention to reality, to what we know from real data-based knowledge, and approach these matters “scientifically.”
I offer this in the spirit of science so that anyone can dispute any of my numbers and offer a different conclusion. I mean that. Scientists want to be proven wrong—that is the only way we ultimately get to the truth. But if I am wrong, any justification of that needs to be based on facts, or at the very least, plausible assumptions that do not contradict any known facts. Remember, I have already granted that the fatality rate MAY be much lower if there is a large number of uncounted cases not included in the published data, based on known cases only. If one wishes to justify mitigation policy based on unknowns (which I skeptically tend to include in the category of wishful thinking) rather than knowns, one can try to make a rational case for that.
Finally, I as well as many others, recognize the enormous economic hardship imposed on so many by this dreadful pandemic. I do not minimize that, and I am just as eager as others to get back to “normal.” Both the economic and the medical/statistical aspects of this disease need to be given due consideration. I only ask that all parties keep to facts and sound reasoning, and not ignore inconvenient facts and data for “political” or “ideological” reasons, that no virus knows of or respects.
Word and Fire 