Thread by @sunsopeningband

Another 🧵. I recall when, early in the #COVID19 pandemic, patients with #MECFS and clinicians/researchers who work with them, warned about #LongCOVID before there was time for cases to develop and thus for many folks to start to experience the syndrome. Early but prescient. 1/n

We’re now seeing hopeful reports of people who are recovering or have recovered from #LongCOVID. I want nothing more than for people to be able to walk away from the horrors of post-exertional symptoms, which I discussed in another thread. This one is about survivorship bias. 2/n

https://twitter.com/sunsopeningband/status/1353927774306619392

There’s a story so famous about survivorship bias that it’s on the esteemed academic website, Wikipedia. The U.S. military asked Abraham Wald to study airplanes returning from combat to determine how best to distribute armor to prevent the airplanes from being shot down. 3/n

You see, armor is heavy, so putting armor all over the plane would prevent it from being able to fly well. Initially the idea was to determine the areas of heaviness damage and to place the armor there. Makes sense, right, to cover the holes before they are holes. 4/n

The problem with that logic is the damage to the planes that returned to the base was not sufficient to destroy them. This means the holes in planes that returned actually were a map for where *not* to place the armor and the parts that *weren’t* damaged required the armor. 5/n

What does this have to do with #LongCOVID? It means we need to define “improvement” endpoints for clinical practice and research carefully so we can study and compare. It also means we can’t make assertions that we can only learn from people who improve. 6/n

Because, at the end of the day, even though we may be looking at the experiences of people who have improved as one set of data that informs our journey to understand #LongCOVID and #MECFS, we’re *always* learning from the people who didn’t or haven’t yet. 7/fin