So! Pregnancy tests are a fine way to explain *some* things about covid testing or testing for infectious disease in general, but are a really-not-great way to talk about other aspects of tests and transmission. Quick thread.

Background: This morning I was listening to a health/science reporter whose work I hugely respect do an interview, and he mentioned a (risky) sleepover for some teenage girls whose parents allowed it bc all the girls had negative covid tests.

But then he said that if the danger was pregnancy, you wouldn’t give them all pregnancy tests and then send them to the party! Because “tests don’t prevent covid any more than tests prevent pregnancy.”

(Instead, he said, you’d hand out condoms or you’d lock the door or maybe sit on the porch and guard them with a shotgun, depending on the kind of parent you are.)

There’s…a lot going on there. But temporarily setting aside my *many* feelings, it made me look back at all the sometimes puzzling uses of pregnancy test metaphors I’ve seen floating around from both reporters and scientists.

So here’s some things that might be useful in deciding a.) yes, this is a great metaphor for my covid explanation or b.) maybe I should rethink this. Are pregnancy tests a good metaphor for:

1. When to take a test? Maybe! A thing about pregnancy tests is you can’t use them and expect an accurate result before X number of days from the start of your last period. They don’t then get LESS accurate as pregnancies advance, however, so the covid test comparison has limits.

2. Test sensitivity? Maybe! It’s a reasonably intuitive way to explain how sensitivity—and even pre-test probability—work in the real world. I've found this helpful, myself.

3. Confirmatory or orthogonal testing? Maybe! Test makers sell pregnancy tests in multi-packs in part bc a lot of people use more than one within a given month (or week, or day) to try to validate their results, and you generally follow up at-home positives with lab tests.

4. Misusing tests as a sole means of preventing transmission. Eghhhh. You wouldn't do this with a pregnancy test—but some covid tests CAN help identity if a person is infectious to others, while a pregnancy test CAN'T tell you whether you’re likely to get someone pregnant.

5. More generally, pregnancy as an infectious disease? NO, PLEASE, NO. You don’t want to go down this rabbit hole for a million reasons, including literally ALL the biology—plus, pregnancy is something many people deeply desire and pursue for years at great effort and expense.

Sometimes the comparison you want here is actually STI testing, which can help people reduce their likelihood of transmitting an infection. But thanks to decades of stigmatizing bullshit, this is also probably not a great option. Talk to people who understand harm reduction!

People like @JuliaLMarcus, for example https://twitter.com/JuliaLMarcus/status/1262448399888142337

Finally, if you could generally avoid public health and infectious disease comparisons that equate disease mitigation efforts with exerting control over the fertility of women and girls, that would be…better.

Huge thanks to @jessicamalaty for reality checks and advice on the above—and for patiently explaining public health things over and over all day every day at @COVID19Tracking .