COVID19: The Math of Why 'Herd Immunity' Isn't Viable in 2020

2 April 2020—Like every human on this planet, I want to see an end to the pain COVID19 is causing. The human scale of lives and livelihoods lost is staggering. We are making progress on slowing the disease, as I recently posted on LinkedIn (US Turning A Corner). But it will get worse before it gets better. Many of us are going through the five stages of grief: Denial, anger, bargaining, depression and acceptance.

Claiming COVID19 is no worse than the flu is, perhaps, part of the denial stage. Anger at decision makers for letting it get this bad is a predictable stage (and probably warranted). Expecting we can get to herd immunity this year is, perhaps, part of the bargaining stage. "If we can just get to herd immunity, we'll be ok," I heard one person say.

In the interest of promoting a more constructive path, let's look at the math of herd immunity and recognize that herd immunity in 2020 is implausible in the US.

Herd immunity can be understood as part of a SIR model. SIR stands for Susceptible, Infected, Recovered. A virus stops with herd immunity when the percent of recovered is such that those infected are less likely to pass it on to those that remain susceptible because there are few of them, and those recovered can't get it a second time. The general belief is that herd immunity would slow the virus once we are somewhere around two-thirds or more recovered. The data from closed populations of the Texas college group headed to spring break together and the church chorus group in Washington suggest that even a couple of hours together singing, or cheering as college kids heading to spring break together are prone to do is enough for 65% to 75% to become infected. Based on the data available, few, if any, are likely to have a natural immunity. A high percent of susceptible means the bar for heard immunity may be even higher than 70%, but let's assume it is 70%. How fast are we building herd immunity in the US?

At the time of writing this article, on April 2, there are around 331 million people in the US of which, John Hopkins reports 8,861 of the 236,339 confirmed cases have recovered. About 34,500 have been hospitalized based on figures collected from states (at 15% hospitalization rate of those testing positive). The total cases likely undercounts that actual scope of infections, so some would argue that we are further down the path toward herd immunity than the case numbers show.

Based on my analysis, I would agree there are a lot more infections than reported. There are three reasons for the difference:

  1. Some are are asymptomatic. They don't even know they have it. We can calculate this number using the Italy Vu study and the Diamond Princess Cruise, where everyone was tested and asymptomatic cases were computed.
  2. Mild cases are unlikely to get tested due to the shortage of tests and the fact that the most prudent thing to do right now is stay out of a hospital unless you feel your life is in peril.
  3. Shortage of tests and backlog means some of those with symptoms either can't get tested or their results are not yet included in case counts.

I've built a model based on the infection fatality rate (IFR). It uses the Diamond Princess Cruise data and calculates the age of those that died from COVID19 in the US in mid March. It projects the total number of cases at that time. I then apply the rate of growth in the number of deaths to project total infections so we have a good estimate of how close we are to achieving herd immunity.

By my model estimates, there will be about 5 million people infected by the end of this week. Some are asymptomatic. Most are mild COVID19 infections. The vast majority will not require hospitalization. They will recover and develop immunity within a couple of weeks. By the end of April, we could have about 1.5% of the US population with immunity -- but that is nowhere near enough to give us any herd immunity.

The math: 70% of 331 million is 232 million people infected. Based on current data available from The COVID Tracking Project, over 10 million would need hospitalization (20 million if the current trend of hospitalization / confirmed cases holds). Half a million to a million would die.

To get to 70% recovered and immune (herd immunity.), we'd need to repeat all the death and pain of March and April 47 more times. If we flatten the curve and maintain this painful pace, it would take us into 2021 before we achieved sufficient herd immunity.

Alternatively, to shorten the time frame to 90 days, consider what it would look like to our medical system to have 47 times more cases over the next three months. Take the worst images from Wuhan, Italy, New York, Louisiana and multiply them because none of these areas got anywhere near this volume.

“Try running an economy with major hospitals overflowing, doctors and nurses forced to stop treating some because they can’t help all, and every moment of gut-wrenching medical chaos being played out in our living rooms, on TV, on social media, and shown all around the world,” Senator Lindsay Graham tweeted last week.

COVID-19 Cases Hospitalizations and Herd Immunity
We have a long way to go to get to Herd Immunity

We all want a solution to the trauma we are seeing unfold. Thinking we can get from where we are today to herd immunity in 2020 isn't a viable solution. The math doesn't work. Asking about herd immunity is a natural part of our collective grief. Call it denial or bargaining stages of grief. Like the comparisons of COVID19 to the flu, we've got to look at the numbers and realize what we are dealing with.

We need to keep working on solutions. We got beyond the mathematically naive "its just like the flu" and recognized we needed a much stronger response. Let's look at the numbers with herd immunity, recognize it isn't viable and move to the viable solutions. First, help others commit to #StayHome. Second, let's start working on phase 2, which is about testing, tracing and containment. Phase 3, which may come with better therapeutics and a vaccine can be a game changer. If people could be given a treatment that they can take at home to recover, and our medical systems isn't overwhelmed, that is viable. You can help by donating some of your computer cycles to folding proteins.

If you want some good insight as to what Phase 2 might look like, check out Taiwan. We can save a lot of lives and livelihoods if we are data driven in the moves we make. There is a way out of the hole we are in -- it just isn't herd immunity.

Recent Analysis

  • 2 April 2020:
    The Math Of Why Herd Immunity Isn't Viable in 2020
  • Forecasts (Jan to April) Vs. Actual :