Bioequivalence and Bioavailability Forum

❝ its for FDA submission.

• Do you control the transfer volume(s) in your method? If yes, you could construct an external calibration curve (peak area instead of analyte/IR-ratio vs. concentrations). If the suspect value fits now into the PK profile you have some additional evidence that IS addition was not correct. Don’t forget that an IS only protects against errors in sample transfer – relying on the assumption of error-free IS addition. It’s always a good idea to transfer volumes as accurate as possible (like in an ES method). An IS is no excuse for sloppy work.
  See also this old (and lengthy) thread.
  
• You said the suspect value is in the elimination phase. If you have to deal with an FDA assessor who likes EMA’s GLs (note: FDA’s guidance is currently under revision), keep the original value for the calculation of AUC_t, but exclude it from the estimation of λ_z (→ AUC_∞). In the worst case exclude the subject from the AUC-assessment but keep him/her for the comparison of C_max.