Bioequivalence and Bioavailability Forum

Dear Ankita,

the answer to your question can be found in the references you gave (especially http://www.fda.gov/downloads/DrugsGuidanceComplianceRegulatoryInformation/Guidances/ucm070118.pdf page 7).
An example of the sentence "People able to read are clearly in advantage in reading" .

What the SAS code excerpt you cite does is the calculation of the critical value (V as a function of the two group variances and their degrees of freedom) and with that the confidence interval for the difference
     (µT-µR)
the very first term in the population BE criterion.

The test used for that end is a variant of a test for comparing two means without the assumption of equal variances.
This is commonly termed Behrens-Fisher problem.

The more commonly known (approximate) solution to the Behrens-Fisher problem is the Welch t-test.
Search the forum for that.

The guidance suggest to use the Lee-Gurland^1,2,3 variant instead of the Welch test because this is said to have better performance with small sample sizes.

¹Lee A.F.S., J. Gurland
"Size and power of test for equality of means of two normal populations with unequal variances"
J Amer Statist Assoc, 70, 933-941 (1975)

²Lee A.F.S., N.S. Fineberg
"A fitted test for the Behrens-Fisher problem"
Comm Statist- Theory Meth, 20, 653-666 (1991)

³Lee A.F.S.
"Coefficients of lee-gurland two-sample test on normal means"
Comm Statist- Theory Meth, 24, 1743-1768 (1995)

Astonishing enough I havn't found any software which delivers the Lee-Gurland test out of the box, even not in R.