How to perform equivalence test for the slope of fit line?

[#1]

Hello,

I am performing validation the analytical procedure and checking the accuracy. Recommendation from USP<1225> is to built the regression line of observed concentration of tested substance vs expected concentration and have the confidence interval for the slope be contained in predefined interval around 1. This is sort of equivalence test (TOST). But I can only get estimate for the slope and std error. Is it possible to perform the TOST analysis with this data somehow?

Also, I see that the actual uncertainty of observed concentration is around 1%, however, uncertainty of the slope of the fit line is around 7% (this doesn't match). How to overcome this discrepancy?

[#2]

Method comparison often uses Deming regression to fit the line based on the matched pairs of observations. Try Analyze > Fit Y by X with test method in Y and reference method in X. Click OK and then click the red triangle and select Fit Orthogonal > Univariate Variances.

The confidence interval estimate of the slope includes 1 in this example, so there is insufficient evidence that the methods do not agree.

There will be more capability for method comparison in JMP 17.