Dipesh Jayswal
●    

2007-05-10 15:35
(6168 d 01:47 ago)

Posting: # 719
Views: 7,012
 

 Outlier in QC samples during BE study analysis [Bioanalytics]

Dear freinds,

I would like to know that which method is to be used to find out the outlier from the QC set used in study sample analysis??

Is it acceptable to label any value as statistical outlying value and to be presented with and without that value in BioAnalytical Report?

Please suggest some way out.

Regds,
Dipesh
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2007-05-10 16:03
(6168 d 01:19 ago)

@ Dipesh Jayswal
Posting: # 721
Views: 5,864
 

 Outlier in QC samples during BE study analysis

Dear Dipesh!

❝ I would like to know that which method is to be used to find out the

❝ outlier from the QC set used in study sample analysis??


There is no need for a formal 'outlier test'. 33% of the QC samples may be outside ±15% (±20% for the lowest QC) of their respective nominal values.

Example:
QCs at three levels (one > LLOQ, one in midrange, one in high range), all in duplicate.
<3 values at different levels outside ±15% of nominal --> valid batch
>3 values outside ±15% of nominal --> invalid batch
2 values ±15% of nominal at the same level --> invalid batch

❝ Is it acceptable to label any value as statistical outlying value and to be

❝ presented with and without that value in BioAnalytical Report?


Yes, you have to report all measured QC samples; it's good practice to flag values outside the acceptance range.

Details in http://www.fda.gov/cder/guidance/4252fnl.pdf FDA's Guideline (page 15pp).


Edit: Link corrected for FDA’s new site. [Helmut]

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
Ohlbe
★★★

France,
2007-05-11 00:52
(6167 d 16:30 ago)

@ Helmut
Posting: # 722
Views: 6,069
 

 Outlier in QC samples during BE study analysis

Dear Dipesh,

If you are planning for submissions to USFDA may I suggest you to have a look at this http://www.fda.gov/foi/warning_letters/archive/m2483n.pdf warning letter (see bottom of page 1 and top of page 2). Particularly, "Contrary to your response, values outside acceptance limits do not automatically qualify as outliers for precision analysis, unless there is an identifiable cause. It is imperative that such unfavorable data be included when evaluating the performance of analytical methods".

It seems that inspectors from the French agency (Afssaps) have a similar attitude during their BE trials inspections. The reasonning is based on section 3.4 of the EU Note for Guidance, which states that the method validation also comprises the study phase itself, to confirm the method's precision and accuracy. QC samples are used to validate each analytical run, according to the usual rules (67 % of all QC results in the run within 15 % of nominal, including at least 50 % at each level of concentration). But they are also used to check the method's precision and accuracy. If you exclude from the calculations all failing results, or even results identified after an outlier test, your QC results are no longer representative of the P&A of your method as applied to the subjects samples. There is no outlier test for subject samples, the only thing is that you may decide to re-analyse some samples for PK reasons if you have an SOP for this. But PK repeats would only be for samples with a rather important deviation from the concentration expected from the PK curve (especially during the absorption phase), not just 15 %.

As already stressed upon by HS if you decide to exclude outliers you should present the results with and without the outliers.

Regards
Ohlbe


Edit: Link corrected for new FDA site. [Helmut]
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
102 visitors (0 registered, 102 guests [including 11 identified bots]).
Forum time: 16:23 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

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