Bioequivalence and Bioavailability Forum • Choose and Pick in method validation

Ken Peh
★

Malaysia,
2013-11-24 05:41
(4203 d 23:43 ago)

Posting: # 11950
Views: 4,374

Choose and Pick in method validation [Bioanalytics]

Post reply

Dear All,

Appreciate feedback on the following questions:

In bioanalytical method validation, if we need N=5, is it acceptable to inject 8 and select 5 samples ? What if all 8 QC samples pass the precision and accuracy ? Should present results of 8 or 5 ?
For method validation, does it make sense to prepare samples until we get the numbers we want ? For example, inject 5, 1 fails, inject 1 again, the process continues until we get the results of 5.
If we follow EMEA method for bioanalysis, does it mean that we have to follow exactly ? There are some differences in opinions betwee FDA and EMEA in bioanalytical analysis. Any possibility inspector or regulatory agency asking us to carry out tests that are mentioned in FDA guideline but not in EMEA.

Thank you.

Best Regards,
Ken

Ohlbe
★★★

France,
2013-11-24 23:45
(4203 d 05:38 ago)

@ Ken Peh
Posting: # 11953
Views: 3,622

Choose and Pick in method validation

Post reply

Dear Ken,

❝ I. In bioanalytical method validation, if we need N=5, is it acceptable to inject 8 and select 5 samples ?

No. This would be considered cherry picking. If the unreported values fail: this may be interpreted as data manipulation.

❝ What if all 8 QC samples pass the precision and accuracy ? Should present results of 8 or 5 ?

If you analyse 8, report 8 whether they pass or fail.

❝ II. For method validation, does it make sense to prepare samples until we get the numbers we want ? For example, inject 5, 1 fails, inject 1 again, the process continues until we get the results of 5.

No. Again, this would be considered as cherry picking or even data manipulation.

❝ III. If we follow EMEA method for bioanalysis, does it mean that we have to follow exactly ? There are some differences in opinions betwee FDA and EMEA in bioanalytical analysis. Any possibility inspector or regulatory agency asking us to carry out tests that are mentioned in FDA guideline but not in EMEA.

Follow the guideline of the region to which you intend to submit data. There are indeed some differences between the EMA, FDA, ANVISA etc. guidelines. If you want to have an SOP which covers all areas: for each parameter, use the most stringent requirement or recommendation.

—
Regards
Ohlbe

Ken Peh ★ Malaysia, 2013-11-26 14:37 (4201 d 14:47 ago) @ Ohlbe Posting: # 11971 Views: 3,539	Choose and Pick in method validation Post reply
	Dear Ohlbe, Thank you very much for your input. Can analyst repeat if one reading is outside the range from five injections for precision and accuracy ? Left only with 4 readings. Should repeat again with five injections ? Otherwise, can not move forward with the assay method. What should be the right approach to take ? Thank you. Regards, Ken

Ohlbe
★★★

France,
2013-11-26 19:45
(4201 d 09:39 ago)

@ Ken Peh
Posting: # 11978
Views: 3,599

Choose and Pick in method validation

Post reply

Dear Ken,

If your run still passes for precision and accuracy: just report it, everything is fine. Nobody ever said that each and every individual value in the run had to pass.

If the run fails for precision and/or accuracy: repeat it, but report the failed run in your validation report too. These will be considered as too independent runs (no pooling !). The FDA and EMA will ask you to calculate the global precision and accuracy with all runs, whether passing or failing.

If another run fails: revise your method.

—
Regards
Ohlbe