I tried this in SAS (thanks to Helmet for filling in the missing lines from the Progesterone guidance), and my CV% from SAS matched Helmet's Phoenix workflow, 33.16. However, my critbound value was different than both of the previous values given here, -0.040468, although I'll admit I'm a poor SAS programmer and I could have errors in handling the datasets.
stats06, would it be possible to get the SAS code that your company used? I attached a file with my code.
Linda Hughes