BE-proff
●    

2015-12-20 19:57
(3020 d 13:15 ago)

Posting: # 15749
Views: 6,686
 

 Stati­stical sig­ni­fi­cance of lab para­meters in BE-studies [General Sta­tis­tics]

Hi All,

In most Russian BE-studies lab parameters are assessed twice - at the screening and at the end of study.
Then obtained results are compared to each other statistically (depending on type of distribution) and conclusion about statistical significance of changes is done.

Is it worth doing? :confused:

I have some doubts because sample size is calculated to show BE of medications. :-|
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2015-12-20 20:16
(3020 d 12:56 ago)

@ BE-proff
Posting: # 15750
Views: 5,730
 

 Nil

Hi BE-proff,

❝ In most Russian BE-studies lab parameters are assessed twice - at the screening and at the end of study.


Not in all? ;-)

❝ Then obtained results are compared to each other statistically (depending on type of distribution) and conclusion about statistical significance of changes is done.


❝ Is it worth doing? :confused:


That’s crap. Imagine the number of tests performed. At which level of significance do these experts run the test? Multiplicity is a killer. Furthermore:
  • Some lab parameters are closely related (say GOT, GPT, γ-GT), others are not (say blood pressure with creatinine)…
  • Hematocrit (post/pre) will be different since we suck so much blood from the poor subjects. Now what?
See also this post.

❝ I have some doubts because sample size is calculated to show BE of medications. :-|


Yep. Is useless. Save the woods.


PS: I speak about studies in healthy volunteers. Studies in patients or on biosimilars (immunogenicity…) are another cup of tea.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

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

2015-12-20 21:16
(3020 d 11:57 ago)

@ Helmut
Posting: # 15751
Views: 5,695
 

 Nil

Hi Helmut,

I do appreciate your sense of humour )).

As far as I know tests to compare results are done at 5% significance level...

About all studies - there are some protocols where lab is assessed only at screening.
mittyri
★★  

Russia,
2015-12-21 01:24
(3020 d 07:49 ago)

@ BE-proff
Posting: # 15752
Views: 5,669
 

 lab tests at least twice

Hi BE-proff,

❝ As far as I know tests to compare results are done at 5% significance level...


Please pay your attention to this problem

❝ About all studies - there are some protocols where lab is assessed only at screening.


Yes, some time ago assessors connived at the protocols where lab tests were performed only during screening (and in case of any AEs during the study)
But now assessors are taking care of this issue, they recommend to specify in protocols that lab tests will be performed at least twice (during the screening and after last sample).

ICH E3 states (and the famous letter from Center of Expertise dated 11.12.13 repeated it):
12.4.2 Evaluation of Each Laboratory Parameter
The necessary evaluation of laboratory values must in part be determined by the results seen, but, in general, the following analyses should be provided. For each analysis, comparison of the treatment and control groups should be carried out, as appropriate, and as compatible with study size. In addition, normal laboratory ranges should be given for each analysis.


BTW I don't know what kind of analysis do the assessors like to see in CSR;-)
Even when I provided just graphs of changes without any analysis of changes I got no LoDs.

Kind regards,
Mittyri
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2015-12-21 02:09
(3020 d 07:04 ago)

@ mittyri
Posting: # 15753
Views: 5,663
 

 Be creative! (#2)

Hi Mittyri & BE-proff,

❝ ICH E3 states (and the famous letter from Center of Expertise dated 11.12.13 repeated it):

12.4.2 Evaluation of Each Laboratory Parameter

❝ The necessary evaluation of laboratory values must in part be determined by the results seen, but, in general, the following analyses should be provided. For each analysis, comparison of the treatment and control groups should be carried out, as appropriate, and as compatible with study size. In addition, normal laboratory ranges should be given for each analysis.


Should be. A comparison given the sample sizes in BE? Forget it. See again the Q&A(R1) to ICH E3 (remember this post?). It’s up to you to adapt the report according to the study’s needs.

❝ BTW I don't know what kind of analysis do the assessors like to see in CSR ;-)


May I ask: What’s the ominous “letter from Center of Expertise”?

❝ Even when I provided just graphs of changes without any analysis of changes I got no LoDs.


I always provide only the original pre- and post-study exams and summary statistics. No comparisons, no bloody shift tables, no graphs, nada.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

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

Russia,
2015-12-21 02:40
(3020 d 06:33 ago)

@ Helmut
Posting: # 15754
Views: 5,697
 

 Creative Russian experts

Hi Helmut & BE-proff,

❝ Should be. A comparison given the sample sizes in BE? Forget it. See again the Q&A(R1) to ICH E3 (remember this post?). It’s up to you to adapt the report according to the study’s needs.


You're absolutely right (as usual). But it's up to assessors opinion in Russia :-D

❝ May I ask: What’s the ominous “letter from Center of Expertise”?


It was like a bomb in BEQ environment :-D
You know that the center of expertise cannot discuss anything with applicants directly (to reduce the corruption :clap:). So this letter was from Center of Expertise to MoH who published it on official site.
I translated the text of this letter with some service. I can send you a copy in Russian in case of need. (I can also attach it here)
Please be informed that recently in the framework of the state registration of generic medicinal products for medical use, cases of incomplete audit reports from clinical studies, especially in terms of compliance with the inclusion criteria and inclusion and exclusion, as well as monitoring the safety of research subjects.
For example, in reports about the investigation of bioequivalence often provide aggregated data about all subjects of research and (or) Declaration of the researchers about whether the research subjects the inclusion criteria, inclusion and exclusion. Initial results laboratoryinstrumental studies, and physical findings are not always. In addition, in some cases, researchers deviation of laboratory parameters (in some cases significant) is interpreted as clinically insignificant, whereas according to the concept of bioequivalence in the study should participate, if possible, healthy volunteers without any abnormalities that could affect and (or) distort the results of pharmacokinetic studies. Sometimes such deviations are not discussed.
In connection with the foregoing we consider reasonable in the preparation of clinical study reports among other things to specify:
1. Aggregated data using tables and graphs presented in the body of the report.
2. The lists of data on individual patients.
3. Comments to the phenomena of interest.
4. The results of all relevant to the safety laboratory investigations should be available in the form of tabular lists in
the following format: each row represents a patient visit, during which he carried out laboratory research in which patients grouped by investigator (if more than one) and treatment group; each column includes important demographic data, information about the dose of the drug and the laboratory results. Since not all research can be placed into the same table, they need to be grouped logically (haematological tests, liver biochemistry, electrolytes, urine and so on). You must use underlining, bracketing or otherwise highlight all abnormal values.
5. The list of all laboratory deviations for each subject (using above format).
6. In respect of each of the laboratory studies need to be repeated, if applicable, and allows the size of the study, the comparison between study and control groups. In addition, the analysis must specify the range of normal values of each laboratory parameter.
7. For each parameter in each time period throughout the study (e.g., at each visit) must submit the following data: the values of group average, or median, range of values, the number of patients with abnormal values or with abnormal by a certain value (for example, greater than twice the upper limit of normal, five times higher limit of normal; you must justify this choice). Use graphics.
General requirements for the preparation of clinical study report, as well as particular requirements for preparation of a report about the investigation of bioequivalence are discussed in detail in chapters 6 and 7 of Volume I of the Guidelines for examination of medicines.
In connection with the foregoing in order to protect the health and welfare of research subjects, and the attainment of high quality research please note applicants and holders of registration certificates to the need to comply with the above requirements in the preparation of clinical study reports.

sorry for poor translation, I don't have enough time to translate it by myself:-(

Kind regards,
Mittyri
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,641 registered users;
45 visitors (0 registered, 45 guests [including 9 identified bots]).
Forum time: 09:13 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5