Reminder: today's Covid counts tell you how bad it was a week or so ago.

(Today is probably worse.)

Originally published on Medium.com 30 November 2020

The headline says it all. One of the hardest things about pandemic decision making is that it exists within a time warp; today’s data is always telling you what happened in the past. That means that today’s situation is often much worse than it appears. It also means that measures you take to fix the problem today will not bear fruit for another couple of weeks.

Time Warp Math

You can understand the time warp phenomenon with a little bit of quick math, illustrated below.

Day “Zero” is the day a group of people get exposed to the virus. They are infected — the virus has entered some of their cells and begun to replicate —but they are not sick, in the sense of having symptoms. As you probably remember from high-school bio, when the virus penetrates a cell, it turns the cell into a virus factory. From the outside, though, the cell still appears normal-ish. You can only detect that you have a problem when the cell starts shedding virus, often by bursting open like a piñata. This is part of why it takes so long for symptoms to come on; the body only fires up a response (fever, runny nose, coughing, etc.) after enough of those viral time bombs have gone off to set off alarms in the immune system. And it helps to explain why the amount of exposure matters so much for how sick you get — it’s about the number of time bombs that get set in your body to begin with.

In the case of Covid-19, those symptoms appear in a range that spans from two to fourteen days, represented on the timeline above as a green block. That matters because most people do not get tested until they become symptomatic. And most people experiencing symptoms wait for at least a day before they decide to get tested. By the time you factor in the hassle of getting to a testing center and turning back the results, its pretty reasonable to guess that the first positive test coming out of a particular infection event comes at Day 4.

That, however, is only the first test. To guess how many people got infected on Day 0, you need a bunch of tests. And since different people had different levels of exposure and different levels of immune response, you cannot ascertain what happened on Day 0 by looking at test results on Day 4. The test results from Day 0 will continue to trickle in for the next two weeks. There is also a lag in a lot of places between testing and getting results, particularly as the number of tests goes up. If you get tested on Day 4, you might only get the results on Day 5 or 6.

This is the “time warp” effect: to know how bad the infection event on Day 0 actually was, you need to have test results from the next week to two weeks.

Testing can shorten the time warp but you cannot eliminate it.

To shorten the time warp, you need to bump the testing back a bit. You can do this through contact tracing. The basic idea is to talk to a person who got sick and find out who they might have infected. Then you can ask that person to go get tested even though they do not have symptoms.

But moving the timeline back is not the same as eliminating the time warp. None of the current tests — antigen, PCR, or antibody — works until the virus has gone through a cycle of replication in the body. The tests can’t detect the virus while it remains hidden inside your cells. Moreover, the PCR and antigen tests look for the virus (either its genetic material or bits of the viral capsule) after it has been released back out into the body. Early in the infection cycle, there is just less stuff to detect. Antibody tests have an even longer lag (but remain potentially viable for much longer), since they are looking not for the virus, but for the antibodies made by the body in response to the virus.

What the time warp means for decision making:

Once you know about the time warp, you can change the way you think about pandemic case counts. When you look at the numbers for a given place, you know that you are looking at the situation from a week or two ago. The useful question to ask is, “What has been happening in the last week?” If the answer is that nothing much changed about people’s behavior, then the situation today is much worse than the number you are looking at. If you look at the numbers and think, “I don’t like how this looks, but it’s not that bad yet,” look again. It wasn’t that bad A WEEK AGO. Chances are, you already crossed the “that bad” threshold; you just don’t know it yet. If you look at the numbers and think, “This is a disaster,” watch out. It was a disaster a week ago. You have been living in a disaster zone for a week.

It’s also important to recognize the time warp effect if your community has been taking steps to combat the virus. The time warp tells you that efforts made today will not show up for a week or two — don’t feel despondent when the numbers keep rising, despite lockdowns or service reductions. It’s not that mitigation measures do not work; it just takes time to see them in the data.