Regular followers of my cataloging the COVID-19 statistics for D.C. will know that I routinely argue that we still don’t understand the specifics of COVID-19 transmission in D.C.:

Despite being months into the pandemic, we still don’t understand how spread is happening in D.C. We can repeat certain generalizations and bromides about particular activities, but what we need are cases that have been examined in detail, followed by communication of those cases to the public. Unless we believe–with absolutely no evidence to do so–that the transmission of SARS-CoV2 is fundamentally different in D.C., we are missing superspreading events. It’s tempting to joke about the White House Covidpalooza, but, were it not for leaks to the press, it’s doubtful our current system would have identified those cases either, other than (possibly) as another unknown rise in cases.

What the city must do is refocus some of its contact tracers and rapid testing capability to dive in deep into a case. Find out where this person shops for groceries and other supplies–and test everyone there. Do they take the bus? Find everyone who was on the bus with them. Are there other stores, restaurants, and other indoor activities where this patient has spent time? Trace and test people there. Are people supposedly in quarantine actually in quarantine? Run one of these cases down until there are no more threads. Some will be busts, but it’s likely you’ll eventually stumble across a superspreading event.

Without details of the natural history of transmission, we simply do not know which restrictions can be loosened and which should be tightened.

That is, we need backwards contact tracing (boldface mine):

In COVID-19 and many infectious diseases, most people don’t infect anyone else. A small percentage of people cause most of the transmission….

Paying attention to superspreading events is also important for contact tracing…

If you find out someone has COVID-19, you can trace their contacts and test or quarantine them. But there is a high likelihood they didn’t infect a single other person.

However, there is a high likelihood that person became infected at a superspreading event. That’s why contact tracers in some countries now spend more time understanding where someone was infected and finding other people from the same cluster.

This is called backward contact tracing. It can help find more chains of transmission.

Backward contact tracing also helps scientists better understand where superspreading events happen. That may help prevent more such events in the future.

In other words, D.C.’s contact tracing isn’t finding these superspreader events. For example, consider this example of an adult who infected multiple family members at a child’s birthday party:

The real question isn’t what happened within the family–that’s obvious. These also are the cases that are easy for D.C. to detect because the focus is on close contacts, not deep investigation. What we need to ask is how the person who brought COVID-19 to the party (literally) contracted it in the first place. Was it work? Dining out with friends? Because the good news, if it can be called that, is that we typically don’t have to go too many links back to find the superspreading event. Once we understand those, we can make better policy.

But until we reallocate some of our resources to really understanding how spread occurs, we’re left with broad brush generalities–which means our responses have to be broad too, even if not truly necessary. That leads to a political problem, since the city (and in fairness, no other states too) doesn’t want to have broad rollbacks. But broad rollbacks are the only policy option right now–or letting people get seriously sickened or die, since we don’t have the evidence to put any other policy options* on the table.

We need to move backwards.

Related: Given yesterday’s disastrous new case numbers, combined with the steady rise, the question is will Mayor Bowser follow her own administration’s rollback guidelines and drop back to phase 1?

*After collecting some ‘anecdata’, I’m convinced D.C. really needs to get serious about wearing masks in apartment building common areas, but who knows if that’s a serious mode of spread.