Lies, damn lies, and covid-19

In the past two weeks, in Italy, we have been drowning in information about the novel coronavirus infection, but the statistics that have been circulating were lacking proper context and interpretation. Is covid-19 just a stronger form of the flu or is it a threat to the world economy? Yes.

Now that the first community transmissions are happening in my adopted home in the San Francisco Bay Area, I would like to relay to my American readers what I learned from the Italian experience.

The most quoted statistics concerned the mortality rate, which has been around 2-2.5% worldwide but only about 0.8% outside of Wuhan, while the flu has a mortality rate around 0.1% and measles around 0.2%. Furthermore, there are no reports of children and infants having died (source) or even having been in critical conditions. This means that it is an infection only somewhat more serious than the flu and that, in particular, parents should not be concerned about their small children.

The other important statistics, however, is the number of infected people that require intensive care. Some source give it at 5%, while the number circulating in Italy is 10%.

Now, developed countries have of the order of an ICU bed per 10,000 people: the European average is one bed per 9,000 people, Italy has one per 8,000, Germany has one per 3,400 (source) and the United States has one per 4,000 (source).

ICU beds tend to have high occupancy, and they are used for victims of stroke and heart attacks, patients recovering from difficult surgeries and so on.

This means that if a person in 1,000 is sick with covid-19 at a given time, the critical cases might overwhelm the capacity of intensive care units. So far, this has happened only in Wuhan, at which point the Chinese government started implementing increasingly strong measures, leading to the lockdown of Wuhan and other places in Hubei province. A ratio of 0.1% of infected people at a regional level has not occurred anywhere outside Hubei province. Even in Lombardy we have “only” about 500 cases out of 10 million people. Two nights ago, however, in the town of Lodi (which has about 50,000 residents) 17 people where brought to the hospital in critical condition from the nearby “red zone”, and many of them had to be routed to other cities because Lodi could not deal with them.

It has been estimated that a covid-19 infected person infects on average another 2-2.5 other people (the “R0” parameter of the disease), typically over a week or less, leading to a rather fast exponential growth. On the Diamond Princess, about 20% of passengers and crew were infected. The only way to reduce the R0 below 1, which would make the epidemic die down, or at least to a value not much bigger than 1 (which would make it grow more slowly) is to reduce person-to-person contact. This is why clusters are locked down, and in places too big to lock down there are measures to reduce such contact, such as closing schools, shutting down sports events, conferences, fairs, concerts etc., and encouraging people to work from home.

The latter “lockdown-lite” measures have stopped the growth of infections in cities like Shanghai, Hong Kong and Singapore, where there were relatively large clusters of cases (much more in Shanghai than in Hong Kong and Singapore). We haven’t seen the number of cases in Korea and Italy leveling off, but this is because testing tends to discover infections that happened one or two weeks prior to the test, so the effect of any measure is only seen in the test numbers a couple of weeks or more after they are implemented.

Meanwhile, Italy is building field hospitals in parking lots, using technology tested during earthquakes, to add hospital bed capacity to the system.

Finally, the measures introduced a week ago are starting to show dramatic economic consequences. The government first emphasized the gravity of the situation, in order to be seen as taking strong actions in a grave moment, and then realized it had caused a PR disaster and now is trying to sound a more optimistic note and to walk back some of the measures that are having the most negative economic consequences. Other governments are unlikely to repeat the same errors, so one should expect official government communication to be biased on the side of not creating alarm.

So here are the lessons for America, in brief:

  • Don’t worry about your young children
  • If the government follows public health best practices, prepare to see school closures and cancelations of big events wherever there are signs of community transmission
  • With containment efforts similar to Hong Kong’s and Singapore’s (and Italy’s, Japan’s and Korea’s), we might see as little as one infection per 10,000-100,000 people, meaning nearly zero risk on an individual basis and no major strain on hospitals, but with significant effects on daily life and on the economy
  • There will be political pressure to send out optimistic messages (see how the White House is asking to pre-approve all communication out of the CDC and the NIH) and to avoid measures that could hurt the economy
  • If the containment efforts are too light, and infections reach even a person in 1,000 on a regional level, the individual risk is still extremely small, but the health care system will be unable to deal with the critically ill. Note that this hasn’t happened anywhere outside of Wuhan and whatever directives come from the top, it’s hard to imagine that local officials would let it happen in their jurisdiction
  • Try not to have a stroke, a heart attack, or serious surgery in the next few months

3 thoughts on “Lies, damn lies, and covid-19

  1. I don’t see why you are so optimistic that “containment efforts similar to Hong Kong’s and Singapore’s (and Italy’s, Japan’s and Korea’s)” will suffice to keep R0 below 1. So far it seems that an exponential explosion with R0>>1 is still very much at play at least in S.Korea and Italy (and maybe also France+Spain, not to mention Iran). While the apparent slowdown in Singapore, Hong Kong, and Japan is indeed encouraging, it would seem way too early to draw such optimistic conclusions.

    If R0 can’t be reduced below 1 then we will have a very large fraction of the population infected and, with a mortality rate of ~1%, this translates to hundreds of thousands or maybe even millions of casualties.

  2. @abc Due to the average incubation period being around 7 days, containment efforts start to play a role approximately one week later.
    Taking Italy as an example, this means that the cases registered so far are almost “independent” of the actions taken by local authorities.

  3. “If the containment efforts are too light, and infections reach even a person in 1,000 on a regional level, the individual risk is still extremely small”

    A recent study indicates that the infection rate among New York City adults is 1 in 7, or at any rate much higher than 1 in 1,000. If that study is true, then it appears that your initial estimate in this post may have been too optimistic.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s