0

I have repeated measures (at different time points) of drug levels (different variants of the same drug) within two different subjects. After the first round of measurements the drugs were reversed and administered to the same two subjects. Thus, overall I have repeated measurements of two different drugs from 2 subjects each. I am interested in figuring out whether the difference in drug levels is statistically significant. Any suggestions on how to do that? Paired t-test, wilcoxons or something else?

To Clarify, the two drugs are different variants of the same drug, and these were administered to both individuals in two rounds i.e. in first round individual 1 got variant A and individual 2 got variant B; then in round 2, Inidividual 1 got variant B and individual 2 got variant A.

Hence the data looks like: Individual 1 - Variant A: x1, x2 ..... Individual 1 - Variant B: y1, y2 ..... Individual 2 - Variant A: z1, z2 ..... Individual 2 - Variant B: w1, w2 .....

Edward
  • 1
  • 1
    Why do any test? Would you expect the time-course of drug concentration to be the same in two different individuals? You have not given enough context for the rationale of the experiment to be clear. – Michael Lew Apr 12 '18 at 20:43
  • The two drugs are different variants of the same drug, and these were administered to both individuals in two rounds i.e. in first round individual 1 got variant A and individual 2 got variant B; then in round 2, Inidividual 1 got variant B and individual 2 got variant A.

    Hence the data looks like: Individual 1 - Variant A: x1, x2 ..... Individual 1 - Variant B: y1, y2 ..... Individual 2 - Variant A: z1, z2 ..... Individual 2 - Variant B: w1, w2 .....

     – Edward Apr 12 '18 at 21:05
  • So you are interested in seeing if there are pharmacokinetic differences between the drugs? If so then consider extracting pharmacokinetic parameters such as bioavailability and half life. A test that just returns a result that there was, or was not, a statistically significant difference at one or more time points is not going to be particularly informative. (Plot the data.) – Michael Lew Apr 12 '18 at 23:17

0 Answers0