Back to the Top
Dear all,
Please let me know significance of selecting 80 % power In BA/BE studies. Please also let me know guideline asking for 80 % power in BA/BE studies.
Please advice in case of BA/BE studies:
1. what is significance of 80% power if my final data showing result less then 80 % power at Cmax & AUC level simultaneously?
2. what is significance of 80% power if my final data showing result less then 80 % power at Cmax level?
3. What if I conclude data on 28 subjects (Considering Dropout & withdrawn subjects). But as per protocol I have to conduct study by using 30 subject (Considering 80 % power).
Regards,
Mukul
Back to the Top
Dear Mukul
1) 80 % power is the standard set in the FDA guidance "statistcal approaches to establish BA/BE" in fact it is stated to be 80 - 90 % power that is recommended to conclude BE. It shows that you have provided enough sample size to conduct your study and therefore it is strong enough for the conclusion to be accurate.
2) It does happen at times that the power calculation later in the study show not to be enough below 80 % it is left to the regulator to evaluate the quality of your work and assess if they will accept the study based on the power analysis result but for future studies it is better to increase the sample size.
3) even if you don't manage to have the complete sample size for the study if all your data and work is clear it is left to the regulator to judge to accept the study and usually they do if everything is appropriate and well reported
Back to the Top
The following message was posted to: PharmPK
Dear Mukul!
> Please let me know significance of selecting 80 % power In BA/BE
> studies.
Convention. Your study should be large enough to have a reasonably high chance (i.e., statistical power) to achieve its target. Most people plan for 90% power allowing for a safety margin (drop-outs, uncertainties in GMR and CVintra used in sample size calculation, etc.) a\0x00\0x00 so 80% is some kind of worst case. ICH E9 calls for a sensitivity analysis a priori. Realistically IECs will not accept a protocol planned with <70% power unless you have (very) good reasons.
> Please also let me know guideline asking for 80 % power in
> BA/BE studies.
> None.
> Please advice in case of BA/BE studies:
> 1. what is significance of 80% power if my final data showing
> result less then 80 % power at Cmax& AUC level simultaneously?
> A posteriori power is irrelevant in bioequivalence and should not be calculated at all. The outcome of a bioequivalence study is dichotomous: Either you have demonstrated BE (CI within the acceptance range) or not. Full stop.
> 2. what is significance of 80% power if my final data showing
> result less then 80 % power at Cmax level?
> See above.
> 3. What if I conclude data on 28 subjects (Considering Dropout&
> withdrawn subjects). But as per protocol I have to conduct study by > using 30 subject (Considering 80 % power).
See above.
Further reader:
JM Hoenig and DM Heisey
The Abuse of Power: The Pervasive Fallacy of Power Calculations for Data Analysis
The American Statistician 55/1, 19-24 (2001)
http://www.vims.edu/people/hoenig_jm/pubs/hoenig2.pdf
RV Lenth
Two Sample-Size Practices that I don't recommend
http://www.math.uiowa.edu/~rlenth/Power/2badHabits.pdf
.... and one of my presentations:
Sample Size Challenges in Bioequivalence Studies and the Myth of Power
http://bebac.at/lectures/MU2010-CD2.pdf
All the best,
Helmut
-- Ing. Helmut Schuetz
BEBAC - Consultancy Services for
Bioequivalence and Bioavailability Studies
Neubaugasse 36/11
1070 Vienna, Austria
Back to the Top
The following message was posted to: PharmPK
Dear Raja!
> 1) 80 % power is the standard set in the FDA guidance "statistcal
> approaches to establish BA/BE" in fact it is stated to be 80 - 90 %
> power that is recommended to conclude BE.
Sorry, you are mixing something up. You are quoting Appendix C, but the topic there is sample size determination: a priori, study planning. But let's continue:
> It shows that you have provided enough sample size to conduct your study
> and therefore it is strong enough for the conclusion to be accurate.
> Now you are jumping to a posteriori power, which is useless. A posteriori power of 90% does not support a study which has already shown bioequivalence (as does 80%). 70% or even 50% do not invalidate a study as well. Please have a look at power curves. At the acceptance limits (for simplicity 0.80 and 1.25) power is 5% a\0x00\0x00 and this exactly alpha. In other words, even at the limits of the acceptance range, there is a (rather small) chance to be able to demonstrate bioequivalence. Following your logic of 80% or 90% do support a study, what's the reverse? When would you (if you step in the boots of regulators) not accept a study (BE demonstrated by CI within 0.80a\0x00\0x001.25)? 70%, 60%, 50%,...? Such a requirement does not exist and reminds me on FDA's awful 80/80-rule (which was abandoned almost 30 years ago).
> 2) It does happen at times that the power calculation later in the study
> show not to be enough below 80 % it is left to the regulator to evaluate
> the quality of your work and assess if they will accept the study based
> on the power analysis result but for future studies it is better to
> increase the sample size.
> Again - such a 'requirement' is an urban myth. If a regulator would not accept a study based on some obscure a posteriori power calculations, I would suggest my clients to go to court.
But you are right, for the next study I would take this result into consideration (what was the cause: higher drop-out rate, higher CV, 'bad' GMR).
> 3) even if you don't manage to have the complete sample size for the
> study if all your data and work is clear it is left to the regulator to
> judge to accept the study and usually they do if everything is
> appropriate and well reported
> On the contrary: It is left to the regulator to accept the study if the CI is within the AR if the product is a generic. If the product is an innovator's the regulator will have a look at the safety/efficacy data and may accept the study even if slightly outside the conventional limits.
To quote RV Lenth:
There is simple intuition behind results like these: If my car made it to the top of the hill, then it is powerful enough to climb that hill; if it didn't, then it obviously isn't powerful enough. Retrospective power is an obvious answer to a rather uninteresting question. A more meaningful question is to ask whether the car is powerful enough to climb a particular hill never climbed before; or whether a different car can climb that new hill. Such questions are prospective, not retrospective.
The fact that retrospective power adds no new information is harmless in its own right. However, in typical practice, it is used to exaggerate the validity of a significant result ("not only is it significant, but the test is really powerful!"), or to make excuses for a nonsignificant one ("well, P is .38, but that's only because the test isn't very powerful"). The latter case is like blaming the messenger.
Best regards,
Helmut
-- Ing. Helmut Schuetz
BEBAC - Consultancy Services for
Bioequivalence and Bioavailability Studies
Neubaugasse 36/11
1070 Vienna, Austria
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Significance of 80 % power In BA/BE studies" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)