Question 6 of 6 Advanced Scenario
How do you distinguish laboratory method variation from batch variation?
Short answer
Use a study design that separates measurement-system effects from true between-batch differences using shared reference materials, replicates and controlled comparisons.
Why
If potency varies across batches, running each batch once on a different day makes batch effects inseparable from run effects. Instead test retained aliquots across days, instruments and analysts with a balanced or randomized design and include stable controls. Estimate repeatability, intermediate precision and between-batch components, using a statistical model that reflects the actual sampling design.
Check whether sample handling or matrix differences confound the comparison. Consistent control-chart drift suggests an analytical issue, while reproducible differences within a balanced design may point toward manufacturing. Do not interpret an observed difference smaller than the method variation as established batch nonconformity.
The conclusion requires a risk-based assessment of both evidence quality and the release specification.
Common mistakes
- Attributing all observed batch-to-batch scatter to manufacturing without quantifying assay variability
The interviewer might ask next
How would you change the experiment if each batch can be sampled only once?