Repeat based on ISR data investigation [Bioanalytics]
❝ During ISR sometimes we found that all the selected samples from any specific subject get failed to meet the acceptance criteria and failing with the huge margin with some specific trend i.e. positive or negative.
Excellent: you are looking at individual data, not just "67 % of my samples are within 20 % so ISR pass, so why bother".
❝ In such cases, scientists use to go for the investigation and run same samples of ISR again to find out the reason for the failure.
OK so far.
❝ During investigation they get the same values which they had got at the time of ISR analysis and not matching with the initial analysis.
Yes, it can happen.
❝ Based on this they conclude error in the first/ initial analysis and they reject the original batch which was otherwise technically passing the batch acceptance criteria.
That's where I disagree. If your initial run was passing (CC and QC were fine) but the results are not reproducible, then you have a serious problem, and you can't just say there was an "error". Actually, you may precisely have the type of problems that the ISR experiment was implemented to detect.
You have identified a problem that seems to affect all samples from one or several subjects. You can re-analyse the samples of those subjects. But if you don't know what happened, you can't prove that individual samples of other subjects, not re-analysed for ISR, were not affected too. Actually you may have to re-analyse all samples in your study, possibly after modifying and re-validating your method !
❝ If the study gets failed then it would be a problem for a filer and if we ignore result of the ISR from which the investigation is triggered then expecting that some day agency may ask for root cause for the failure in ISR and doubt on the initial analysis.
Oh yes, they would not hesitate to challenge the whole study.
❝ please share with us your experience or any suggestion to handle such situations.
Bioanalysis published a special issue on ISR back in 2011. Some of the papers could help you. Have a look for instance at the review from Manish Yadav and Pranav Shrivastav and at the paper by Aimin Tan et al. They are providing some examples that may give you some ideas.