Get best of the week

Coronavirus: A fiasco of the century in the making? How we make decisions without reliable data

Translation of the article for the specialized medical publication " STAT " in the section " Opinions ". Posted by John P.A. Ioannidis ( the John PA Ioannidis ) - a professor of medicine, epidemiology and public health, biomedical science and statistics at Stanford University and co-director of the Stanford Innovation Center Meta-study .


COVID-19, is already called the "pandemic of the century." But also, perhaps, this is the largest “fiasco of the century”.

UFO Care Minute


The pandemic COVID-19, a potentially severe acute respiratory infection caused by the SARS-CoV-2 coronavirus (2019-nCoV), has officially been announced in the world. There is a lot of information on Habré on this topic - always remember that it can be both reliable / useful, and vice versa.

We urge you to be critical of any published information.


Official sources

, .

Wash your hands, take care of your loved ones, stay at home whenever possible and work remotely.

Read publications about: coronavirus | remote work

At a time when everyone needs reliable information - from scientists, doctors and governments to quarantined people or just “social distance” - we lack reliable data on how many people were infected with COVID-19 or who still getting infected. Reliable information is vital in order to have a direct impact on decisions and actions that are of fundamental importance and to monitor their impact.

In many countries, unprecedented draconian countermeasures have been adopted. If a pandemic dissipates — either on its own or through these measures — short-term social distance and isolation may be tolerable. However, how long should such measures continue if the pandemic continues around the world? howCan policy makers say they do more good than harm?

It takes many months (or even years) to develop and properly test vaccines or available treatments. Given these terms, the consequences of prolonged isolation are completely unknown.

The data collected so far about how many people are infected and how the epidemic is developing are completely unreliable. Due to the fact that the possibilities of full testing are limited today, some deaths and, probably, the vast majority of diseases caused by COVID-19 are not taken into account. We do not know if we can become infected 3 or 300 times. Three months after the outbreak, most countries, including the United States, are not able to test a large number of people, and no country has reliable data on the prevalence of the virus in a representative random sample in the general population.

This failure with evidence creates enormous uncertainty about the risk of death from COVID-19. Reported deaths, like the official 3.4% of WHO, are horrifying - but pointless. Patients who pass the COVID-19 test are disproportionately affected by severe symptoms and are fatal. As testing capabilities are limited in most global healthcare systems, selection bias may even worsen in the near future.

The only situation in which the closed population was fully tested was the quarantined passengers of the Diamond Princess cruise ship. The mortality rate there was 1.0%, but it was mainly elderly people, among whom the mortality rate from COVID-19 was significantly higher.

By projecting the Diamond Princess mortality rate onto the age structure of the US population, the mortality rate among people infected with COVID-19 will be 0.125%. But since this estimate is based on extremely small data — there were only 7 deaths among 700 infected passengers and crew — the actual mortality rate can be either five times lower (0.025%) or five times higher (0.625%). It is also possible that some of the infected passengers may die later, and that tourists may have a different frequency of chronic diseases - this is also a risk factor for adverse outcomes with COVID-19 disease - compared to the general population. Adding these additional sources of uncertainty, it can be assumed that reasonable estimates of the mortality rate in the event of illness among the US population range from 0.05% to 1%.

This huge range significantly affects the severity of the pandemic and what needs to be done. Mortality across the entire population at 0.05% is lower than from seasonal flu. If this is a real figure, then isolating the world with potentially enormous social and financial consequences can be completely irrational. It is as if a domestic cat attacked an elephant. And, scared and trying to avoid a cat, an elephant accidentally jumps from a cliff and dies.

Can the lethality of COVID-19 be so low? No, some say, indicating a high mortality rate among older people. However, some coronaviruses that cause symptoms like mild flu or common cold, and have been known for decades, can have a mortality rate of up to 8% when they infect older people in nursing homes. In fact, such “light” coronaviruses infect tens of millions of people every year, and make up from 3% to 11% of those who are hospitalized in the USA every winter with lower respiratory tract infections.

These “light” coronaviruses may be involved in several thousand deaths every year around the world, although the vast majority of them are not documented through accurate testing. Instead, they are lost in the form of noise among 60 million deaths from various causes each year.

Although successful influenza surveillance systems have been around for many years, this specific disease is laboratory-confirmed in a tiny minority of cases. In the United States, for example, 1,073,976 samples have been tested this season so far, and 222,552 (20.7%) have been tested positive for influenza. Over the same period, the estimated number of influenza-like diseases ranged from 36 million to 51 million, with an estimated mortality from influenza ranging from 22 thousand to 55 thousand people.

Note the uncertainty regarding mortality from "flu-like" diseases: a 2.5-fold range corresponding to tens of thousands of deaths. Each year, some of these deaths are caused by the flu, and some by other viruses, such as the coronavirus colds.

In a series of autopsies performed on the presence of respiratory viruses in samples from 57 elderly people who died in the 2016-2017 influenza season, influenza viruses were detected in 18% of the samples, while another type of respiratory virus was detected in 47% of the samples . In some people dying from viral respiratory pathogens, more than one virus is detected during autopsy and bacteria are often added. A positive coronavirus test does not necessarily mean that it is this virus that always bears the primary responsibility for the death of the patient.

Assuming that mortality among people infected with COVID-19 virus is 0.3% of the total population - the average guess from my Diamond Princess analysis is that 1% of the US population is infected (about 3.3 million people) , then this means about 10 thousand deaths. It sounds like a huge number, but it is buried in the noise of the assessment of deaths from "flu-like" diseases. If we did not know anything about the new virus and did not test people using PCR tests , then the number of deaths from all “flu-like” diseases this year would not look unusual. At best, we might accidentally note that the flu this season seems a little worse than average. Media coverage would be less than during the NBA game between the two most third-rate teams.

Some fear that 68 deaths from COVID-19 in the United States as of March 16 will grow exponentially to 680, 6 800, 68 000, 680 000 ... along with similar catastrophic pictures around the world. Is this a realistic scenario or bad science fiction? How can we say at what point such a curve can stop?

The most valuable information to answer these questions would be to know the current prevalence of infection in a random sample of the population and repeat measurements at regular intervals to estimate the frequency of new infections. Unfortunately, we do not have such information.

In the absence of data, “prepare for the worst” reasoning leads to extreme measures such as “social distance” and isolation. Unfortunately, we do not know if these measures work. School closures, for example, can slow down distribution. But such measures can also lead to the opposite result, if children still continue to communicate outside the school, that children spend more time with elderly family members susceptible to this disease, and much more. Closing schools can also reduce the chances of developing collective immunity in age groups, with minimal risk of serious damage to the disease.

It is this view that underlies the UK’s official position to keep schools open, at least as of this writing. In the absence of data on the actual course of the epidemic, we do not know whether this prospect was brilliant or catastrophic.

Smoothing the curve so as not to overload the healthcare system is conceptually justified - but only in theory. A visual image that has become viral in the media and social networks shows how flattening the curve reduces the size of the epidemic, which exceeds the threshold of what the healthcare system can handle at any time.

However, if the healthcare system is still overloaded, most of the additional deaths can no longer be caused by coronavirus, but by other common diseases and conditions, such as heart attacks, strokes, injuries, bleeding, etc., which do not receive adequate treatment. If the level of the epidemic still overloads the health care system, and extreme measures will have only modest effectiveness, then flattening the curve can aggravate the situation: instead of being overloaded for a short, acute phase, the health system will remain overloaded for a longer period of time . This is another reason we need data on the exact level of epidemic activity.

One of the final indicators is that we do not know how long social isolation and isolation measures can last without serious consequences for the economy, society and the mental health of citizens. Unpredictable consequences may follow, including the financial crisis, unrest, civil unrest, war and the breakdown of the social fabric of society. At a minimum, we need objective data on the prevalence and incidence of infections in order to drive decision making.

According to the most pessimistic scenario, which I do not support, if a new coronavirus infects 60% of the world population and 1% of those infected die, it will lead to more than 40 million deaths worldwide, which corresponds to the 1918 influenza pandemic.

But the vast majority in thishecatomb will be the elderly. This contrasts with 1918, when many young people died.

One can only hope that, as in 1918, life will continue. Conversely, with isolation for months, if not years, life will largely stop, and the short-term and long-term consequences are completely unpredictable, and millions, if not billions, of lives can ultimately be at stake.

If, however, we decide to jump off a cliff, then we need reliable data to inform everyone about the justification for such a step and about the chances of falling somewhere in a safe place.

March 17, 2020

John PA Ioannidis
jioannid@stanford.edu
@METRICStanford

PS


Answer Mark Lipshicha ( by Marc Lipsitch ), Ph.D., professor of epidemiology at the Harvard School of Public Health named TH Chan (TH Chan) and director of the Harvard Center for the Dynamics of Infectious Diseases . - article “Now we know enough to act decisively against COVID-19. Social distance is a good start. ” ( note: article in English )

March 18, 2020

More articles:


All Articles