Method Correlation Total Allowable Error vs Slope and Correlation Coefficient

17 followers
0 Likes

I was reading James Westgard's book on Basic Method Validation and came across the passage below.

I have been to many institutions and before I came along, they tend to just look at the slope and R value to see if two methods correlate.

I was always use to just using total allowable error as a means to compare acceptability. I think this is actually the correct way to interpret acceptability and I don't know why others *just use slope and R values. 

Can a lab director define acceptability as >.9-1.1 & R>.95?

Can they pick and choose TAE *or slope/R when no TAE is defined for a regulated analyte?
what are other people's practices at their institution?
What is best practice?

Attachment.
image.png

1 Reply

Most use Demming regression for Accuracy/Bias comparison not just (r)

TEa targets for Precision are common but not for Accuracy/Bias

Reply
Subgroup Membership is required to post Replies
Join POCT Listserv now
Jo Win
about 20 hours ago
1
Reply
0
Likes
17
Followers
173
Views
Liked By:
Suggested Posts
TopicRepliesLikesViewsParticipantsLast Reply
HMS Hepline kit
Cassie Laranio
about 16 hours ago
1068
Miranda Cotter
about 13 hours ago
Critical Documentation on the GEM 5000
Samantha Draper
about 19 hours ago
80143
Miranda Cotter
about 13 hours ago
Method Correlation Total Allowable Error vs Slope and Correlation Coefficient
Jo Win
about 20 hours ago
10174
Carl Bayaca
about 15 hours ago