Post-script. We can repeat the same calculations using prevalence estimates among COVID-infected people that are just out recently--36% (!)

As prevalence goes up, positive predictive value starts looking better. Problem: the 95% confidence interval includes chance (50%). Not statistically stable.

Look what happens to positive predictive value!

* sad trombone sound

Summary: 🚽. Adjusting for prevalence, the test now performs like a coin flip. The consolation prize is the negative predictive value went up. Now the test is only robust when it's negative.

But when it's positive? Who knows.

Let's plug the data from our 2x2 table into MedCalc and see what happens.

Voila! Here's the output. The sensitivity and specificity values are close to what was reported in the paper. Differences are from rounding to the nearest whole person and not accounting for the adjustments the authors made.

In case anybody cares--and why would you?--I got bored and reconstructed the 2x2 table from the calculations reported by the authors. I didn't see the data expressed this way in the paper including appendices. Maybe it's there and I missed it, doing all this for nothing. It would be on brand for me.

Anyway, here was my attempt at getting a decent point of view published on the topic in the @sfchronicle.com. I’ll
post it here. It’s just taking up space on my laptop. Maybe there were legitimate reasons for not running it but having a weird, poorly reasoned, self-indulgent take wasn’t one of them.

A little while back, the @sfchronicle.com decided not to run my op-ed on Long Covid and the chronic health concerns with letting infections run through our communities unchecked based on my expertise in the field and contemporary data. This is the kind of thing they chose to run instead. Gross. 🤢

And, of course, this account. 😆

Here’s where the numeric PEM data came from. Does it make sense to you? Because it doesn’t make sense to me.

This is taking me a little while because I am so not a visual artist lol. Totally worth it. What are your thoughts on this piece? I added a part about top symptoms to monitor and then the general plan for a green day. Tried to balance making the language too rosy and yet also not too restrictive.

So...what do you think about this for the header of an action plan form?

I've been working on a thing with a very talented group, many of whom with lived experience with Long Covid. I'm honored for the chance to help curate their work in a special place to share their findings and tell their stories. Coming in January 2025 to the Cardiopulmonary Physical Therapy Journal.

I mean, with ideological balance like this, who needs enemies?

Well, team. We had a good run. I made my first nuclear block today. 🤷🏻‍♂️