Thread by @COREIMpodcast

1/ If a test to detect a disease whose prevalence is 1/1000 has a false positive rate of 5%, what is the chance that a person found to have a positive result actually has the disease?

Share your thoughts below; we’ll share the answer and more this afternoon!

2/ Did we hear you ask for a math consult...STAT? As you may have noticed, we’ve taken a break from our regularly scheduled cardio programming to bring you a little statistics lesson.

To start, we’ve provided another way of looking at this beyond the 2x2 box. Try it yourself!

3/ We are assuming this test is 100% sensitivite. That being said, it is worth underlining here that the prevelance of the disease in the tested population is of the utmost importance. Changing the prevalence would change the ratio of true positives to false positives.

4/ If you look at our pool, you’ll find it’s not too far off from previous versions of this study. Stats do hurt the brain, and we tend to be pretty bad at them!

5/ So that’s today’s lesson. For an excellent lesson in the same topic, shout out to @AdamRodmanMD and Bedside Rounds. Episode 15 - Innumeracy covers the topic beautifully.

Keep learning, friends! And keep going - it’s nearly Friday!

6/ And a huuuuge shoutout to @arjunmanrai and @JudithStrymish whose work we drew on for this mini tweetorial!

7/ Link to their paper in @JAMAInternalMed cited in slide 4 @sacjai @arjunmanrai @JudithStrymish

jamanetwork.com/journals/jamai…