Log in
Register|
Alpesh Ramani ☆ Ahmedabad, India, 2010-07-03 16:09 (5830 d 13:30 ago) Posting: # 5586 Views: 5,249 |
|
|
Dear Forum, For normal cases 80% or more power is required to calculate the sample size for BE studies. But, when it is mandatory to use more then 80 % (i.e., 90%) for calculation. Mean to say what can be the special situations where we have to go for power 90% instead of 80 %. Alpesh Ramani TRC, INDIA Edit: Category changed. [Helmut] — Alps |
|
Helmut ★★★   Vienna, Austria, 2010-07-03 16:25 (5830 d 13:14 ago) @ Alpesh Ramani Posting: # 5587 Views: 4,299 |
|
|
Dear Alpesh! ❝ For normal cases 80% or more power is required to calculate the sample ❝ size for BE studies. >80% is not required, but desired. Since the producer’s risk (β) = 1–power, it depends on the budget of the sponsor.  ❝ But, when it is mandatory to use more then 80 % (i.e., 90%) for ❝ calculation. ❝ ❝ Mean to say what can be the special situations where we have to go for ❝ power 90% instead of 80 %. You may want to aim for power >80% in order to adjust for the expected drop-out rate (AEs, more than two periods, etc.). ICH E9 (Statistical Principles for Clinical Trials, Section 3.5, 1998) calls for a sensitivity analysis in the planing stage. You should investigate violations of assumptions (larger deviation the T/R-ratio, higher CV, drop-out rate) on power (based on a given sample size). For details see one of my lectures. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
Ing. Helmut Schütz