Questions About Sparse Sampling Designs and SE Estimation [NCA / SHAM]

posted by martin  – Austria, 2025-01-28 15:14 (14 d 13:05 ago) – Posting: # 24350
Views: 253

Dear Mittyri,

I wanted to share some insights on sparse sampling pharmacokinetic (PK) and non-compartmental analysis (NCA) that you might find useful:
  1. Standard Error in AUC Calculation: The R package ‘PK’ uses the linear trapezoidal rule based on arithmetic means per time point, properly accounting for standard error (SE). In this case, the area under the curve (AUC) is essentially a weighted mean, where weights are determined by the linear trapezoidal rule and time points. Therefore, the SE can be directly derived because of its linear relationship.
  2. Geometric Means Usage: When using geometric means instead of arithmetic ones, the standard error formula needs manual adjustment, as it is not supported in the R package 'PK’. I think you can likely use the formula for calculation of the SE for the AUC by replacing the SE for the arithmetic mean by that for the geometric mean and weight it appropriate as indicated by the linear trapezoidal rule and time points (see papers below which gives the formulas for the SE for AUC using arithmetic means and linear trapezoidal rule)
  3. Linear-Up Log-Linear Down Rule: The SE calculation becomes quite challenging due to the non-linear elements introduced by the log-linear down rule.
The primary challenge in NCA with sparse sampling arises not from the AUC calculation methods but from handling values below the limit of quantification (LLOQ). I recommend the following paper for insights on this:Additionally, here are some papers on NCA with sparse sampling that might interest you. If I recall correctly in the reference section in one of those works refers to a paper which addresses the log-linear rule, though I can’t remember the exact source

Best regards,

Martin

Complete thread:

UA Flag
Activity
 Admin contact
23,380 posts in 4,914 threads, 1,662 registered users;
22 visitors (0 registered, 22 guests [including 6 identified bots]).
Forum time: 04:19 CET (Europe/Vienna)

Skill is a function of chance.
It’s an intuitive best-use of chance situations.    Philip K. Dick

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5