Bioequivalence and Bioavailability Forum

BE-proff ● 2015-08-17 13:59 (3969 d 00:11 ago) Posting: # 15279 Views: 5,968	Indications for different BE-designs [Design Issues] Post reply
	Hi All! Could anybody explain "indications", advantages and disadvantages of the following BE-designs: 3×3 Latin Squares 6 sequence Williams’ design 4×4 Latin Squares, Williams’ design 2×4×4 fully replicated design Thanks to everyone!

Helmut
★★★

Vienna, Austria,
2015-08-17 16:55
(3968 d 21:15 ago)

@ BE-proff
Posting: # 15283
Views: 4,782

Effects and adverse effects

BE-proff,

for effects and adverse effects ask your physician or pharmacist. See f.i. this presentation. Some more:

2×2×3 Full replicate design (RTR|TRT)
- Estimation of CV_wR and CV_wT.
  Suitable for both ABEL (EMA) / RSABE (FDA) and ABE.
  Less sampled blood volume compared to RTRT|TRTR and lower chance of dropouts.
- The EMA requires ≥12 evaluable subjects in sequence RTR. Practically not relevant (see this post).

2×3×3 Partial replicate design (RRT|RTR|TRR)
- Suitable for both ABEL (EMA) / RSABE (FDA) and ABE.
  Less sampled blood volume compared to RTRT|TRTR and lower chance of dropouts.
- Allows only estimation of CV_wR.
  Sometimes convergence issues if evaluated by FDA’s method for ABE.

2×4×2 Balaam’s design (RT|TR|RR|TT)
- Allows estimation of CV_wR and CV_wT.
  Suitable for both ABEL (EMA) / RSABE (FDA) and ABE.
  Less sampled blood volume compared to 3-period replicate designs and lower chance of dropouts.
- Very (!) low power compared to 3- and 4-period replicate designs. Only ½ of the subjects are used to assess BE; CV_wR and CV_wT are estimated from ¼ of the subjects.

Balanced Incomplete Block Designs (BIBD)
- Suitable for ABE.
  Generally only used in pilot or dose proportionality studies if >4 treatments are tested and >4 periods cannot be used due to blood volume restrictions and/or expected dropout-rate.
  Number of periods < number of treatments.
- Lower power compared to higher-order crossovers (Latin Squares or William’s designs, where number of periods = number of treatments).

Notes:

Balaam’s design and the 4-sequence 4-period replicate designs (2×4×2 RTTR|TRRT|RRTT|TTRR) are currently (v1.2-08 of 2015-07-10) not implemented in PowerTOST’s functions for ABEL and RSABE.
The completely randomized 4-period replicate (RTTR|TRRT|RRTT|TTRR|RTRT|TRTR) is not implemented at all. Doesn’t matter – haven’t seen one in ages.

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

BE-proff ● 2015-08-18 10:26 (3968 d 03:44 ago) @ Helmut Posting: # 15293 Views: 4,762	Effects and adverse effects Post reply
	Dear Helmut, Thank you very much for your help! I do appreciate it