Expected, but disappointing
-
That was a quick turnaround. I haven’t actually scheduled mine yet. I suppose I should. What’s the error rates on that test?
-
Yeah, Tuesday morning blood draw. Results mid day Thursday.
Pretty high quality:
-
There’s what? 2 Abbott antibody tests? 3?
-
So mine is a standing order that I'm eligible for weekly, though I only do it monthly.
If I showed a positive at some point, I would repeat it that next week before I started celebrating too much.
-
@LuFins-Dad said in Expected, but disappointing:
There’s what? 2 Abbott antibody tests? 3?
The Architect is a big floor model instrument. It's not something you find in small hospital labs.
-
I thought that I had read that the sensitivity or specificity of the Abbott Antibody Test was 85%, which made me think that one lab may run several different types of Antibody testing...
-
This data is about one test (though there are in fact two different abbott tests).
Sensitivity of 100% means there were no false negatives in their sample of 88. Specificity of 99.6% comes from 4 false positives in their 1070.
The PPV and NPV are positive and negative predictive values. These are a function both of test accuracy AND the prevalence of the virus in the population (in this case they're assuming ~5%)
THe logic there is as follows:
Suppose 1 in 100 was a false positive, and we knew that there were fewer than 1 in 1000 ACTUAL positives? In that case if you tested 1000 people you'd expect 10 positive tests even though only one person would expected to actually be infected. SO the predictive power would be 10%.
But what if half the population was infected, and you still had a 1 in 100 false positive rate?
If you tested 1000 people, you'd expect 500 to be real positives and 10 to be false positives. So the predictive power of a positive result would be 98%.
-
LD - see this for some other FDA results:
-
George - no 'data', but there are anecdotes. THey're mixed so far but not as bad as many of us expected they'd be.
Though FWIW I personally knew just one and she died, early - March 30.
I am tracking it pretty closely because at some point I'll need to make a call about whether/how to remain isolated when things open back up.
Like, do I keep the boy out of school in September? Do I take my chances? Or do I let him go to school and get an apartment nearby? Same w/ Rachel going back to work.
I will need to make a call on that by September at the latest, and will base it heavily on my own (imperfect) expectations as to my survival odds.
-
@jon-nyc said in Expected, but disappointing:
George - no 'data', but there are anecdotes. THey're mixed so far but not as bad as many of us expected they'd be.
Though FWIW I personally knew just one and she died, early - March 30.
I am tracking it pretty closely because at some point I'll need to make a call about whether/how to remain isolated when things open back up.
Like, do I keep the boy out of school in September? Do I take my chances? Or do I let him go to school and get an apartment nearby? Same w/ Rachel going back to work.
I will need to make a call on that by September at the latest, and will base it heavily on my own (imperfect) expectations as to my survival odds.
If I were in your situation, I'd be just as cautious as you. I totally understand your concern. It's tough to know what the right thing to do is...
