Page 1 of 1: Q: Biostatistics: Bioequivalence Studies repeat analysis

berthoun · Credits: 77713 My Scrapbook My Reading List 262569 Books

Supposedly, in drug-plasma concentartion profile one concentration has been found way out of the profile this concentration sample was reanalysed twice and new conncentartions were obtained. First of all is there a documented statistical method to refer to in order to determine whither to accept that initial odd concentration (do you consider it an outlier) or not thus justifying the repeat analysis and whither a single repeat is sufficient or a double repeat is a must? Second, between the three concentrations is there a certain rationale to follow to what concentration must be accepted or whither to reject the data obtained or not? or is it possible to justify taking one of the concentation not the other or perhaps taking the average not the median?
if i am asking in the wrong area please inform me. Cause this is important. if anyone can point me at least to possible references i'ld be grateful.

E-mail:

Only RxPG members can see links here! Register or Sign In!

Bruno · Credits: 44624 My Scrapbook My Reading List 50 Books

Please post your queries in a way even dull heads like me can understand........ It is too much for a person like me to comprehend these complex worded questions and think of ALPHA ERROR at the same time

I gave up plans of reading literature (and landed in Medical College) because I was afraid of the long sentences of Corbett and Doyle

berthoun · Credits: 77713 My Scrapbook My Reading List 262569 Books

Supposedly, in drug-plasma concentartion profile one concentration has been found way out of the profile this concentration sample was reanalysed twice and new conncentartions were obtained. First of all is there a documented statistical method to refer to in order to determine whither to accept that initial odd concentration (do you consider it an outlier) or not thus justifying the repeat analysis and whither a single repeat is sufficient or a double repeat is a must? Second, between the three concentrations is there a certain rationale to follow to what concentration must be accepted or whither to reject the data obtained or not? or is it possible to justify taking one of the concentation not the other or perhaps taking the average not the median?

MEANs
in drug-plasma concentartion profile:
Time (hrs): 0.25, 0.50, 0.75, 1.00, 1.50, 2.00, 4.00, 8.00, 10.00
corresponding Concentration(ug/ml) 1, 2, 3, 10, 4, 3.5, 2, 2.25, BLOQ
BLOQ below limit of quantification

1. is the concentration at 1.00 hrs (10ug/ml) an outlier?
2. A bioanalytical lab will normally repeat the analysis depending on the quantity of sample remaining Regardless which one would you prefer:
a) Do you do the repeat once?
b) Or Do you repeat twice?
3. Following repeat analysis:
Repeat 1 concn.= 8 ug/ml
Repeat 2 conc = 7 ug/ml
Which concentration are you going to report 8, 7, or 10?
Are you instead going to report the average 8.3 ug/ml?
Or are you going to take the median “8 ug/ml”?
Regardless I need the rationale behind the preference of one choice rather than the other.
Is there a universal method?
Can you apply it in cases where the concentration is way below?

sorry no can do, this is the farthest i can go with explainin thanx

Bruno · Credits: 44624 My Scrapbook My Reading List 50 Books

Why can't you OMIT that value...... ????

I am not sure that you can omit the value in this example, but (in normal occasions, when p=0.05) if you have 20 variables, and if one is "out of relation" that is omitted

berthoun · Credits: 77713 My Scrapbook My Reading List 262569 Books

Well, cause we're doing pharmacokinetic analysis for bioequivalence studies... The FDA bides you that no collected data what so ever is omitted unless you have what we call missing samples or a rationale scientific bases for doing so and you have to prove that the consequences laid on that value well not jeoprodize the outcome of the study in having a failed product out... kind of faking the story of events and its not quite easy especially in the first part of the drug-plasma profile cuase no one would believe you (Human subjects are confined through the first part mostly if not all through the profile). plus the location of the value has sometimes its consequence to your PK analysis
Way too long to explain bioequivalence stuff. We're trying to affoid illegal approaches and trying to find another way out. Now, not all things follow the given example i.e p value at times may be acceptable but from a PK view it affects the outcome of the studies we're doing. We tried approaches such as 15 % difference thing but we don't have a scientific bases for this...we made it up. So was only trying to get some help not to get anyone pissed off the subject. We need at least some references to dig in. PS our PK analyst was unable to provide assistance on the subject. Thanx Dude.

Bruno · Credits: 44624 My Scrapbook My Reading List 50 Books

berthoun · Credits: 77713 My Scrapbook My Reading List 262569 Books

shat · Credits: 77713 My Scrapbook My Reading List 262569 Books

shatbolemagne

Guest · Credits: 77713 My Scrapbook My Reading List 262569 Books

		Sign In New User? Sign Up Sign in to access your control panel and messenger!
				TechZone \| SpiderNevi \| HowTo? \| Scrapbook!