Dipesh Jayswal ● 2007-05-10 15:35 (6168 d 01:47 ago) Posting: # 719 Views: 7,012 |
|
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 ★★★ Vienna, Austria, 2007-05-10 16:03 (6168 d 01:19 ago) @ Dipesh Jayswal Posting: # 721 Views: 5,864 |
|
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 Edit: Link corrected for FDA’s new site. [Helmut] — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz 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 |
|
Dear Dipesh, If you are planning for submissions to USFDA may I suggest you to have a look at this 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] |