This post has been de-listed
It is no longer included in search results and normal feeds (front page, hot posts, subreddit posts, etc). It remains visible only via the author's post history.
We're using a diagnostic ELISA where we calculate antibodies against some a viral agent. We have the cut-off value where samples >1 are considered positive. There's also a grey zone >0.9
I know that if a sample with OD 0.045 is as negative as a sample with 0.81 and that is as negative as a sample with OD 0.10.
But my question is: If we have a given patient where once he has a value of 0.09 and then he's tested months later (where we have a different LOT and of course a different samples) would it be normal to have a higher value like 0.60? Or could it indicate that he should be checked sooner?
My professor didn't quite answer that, he just said that it's negative end of story but he didn't give it much tought I think, but I'm curious.
Bonus question: when someone is turning positive, if we took blood from him every let's say a week, would he be like 0.1, 0,3, 0,5, 0.9, 1.10, 1.7 etc? I don't believe it could be a linear thing, it doesn't work that way but on the other hand he couldn't jump from nothing to 2.0
Subreddit
Post Details
- Posted
- 4 months ago
- Reddit URL
- View post on reddit.com
- External URL
- reddit.com/r/labrats/com...