Questions: Analytical Method Equivalence and Transfer

Question 1

A pharmaceutical company transfers a validated HPLC potency assay from its R&D lab to a contract manufacturer. Both labs analyze identical reference samples and the means differ by 1.5%. How should they determine whether the transfer is successful?

• A. The transfer is successful if the difference is less than 2%, because this is below the typical HPLC precision limit
• B. The receiving lab should run the method on actual production batches, not reference samples, to determine success
• C. The 1.5% difference must be evaluated against predefined statistical acceptance criteria — equivalence testing or tolerance intervals — not judged by subjective comparison
• D. If both labs achieve mean recovery above 95%, the transfer is successful regardless of interlaboratory differences

Answer: (C)

Method transfer decisions must rest on predefined statistical acceptance criteria, not arbitrary thresholds or post-hoc judgment. A 1.5% difference may be entirely acceptable for one method (say, a raw material identity test) or critically unacceptable for another (a drug product potency assay with a narrow acceptance range). The predefined criteria — established before the transfer begins — specify the maximum acceptable difference and the statistical procedure (equivalence testing or tolerance intervals) that controls both the risk of accepting a truly non-equivalent method and the risk of rejecting an adequate one. This is what separates a rigorous transfer from a subjective exercise.

Question 2

A receiving laboratory independently performs key validation experiments — accuracy, precision, and specificity studies — and demonstrates it meets the same performance specifications the originating lab established. Which transfer strategy is this?

• A. Comparative testing — direct side-by-side analysis of identical samples at both labs simultaneously
• B. Covalidation — the receiving lab independently demonstrates it meets the same validation acceptance criteria as the originating lab
• C. Bridging study — a third neutral laboratory validates both labs independently
• D. Waived transfer — the receiving lab skips formal transfer because the method is compendial and well-established

Answer: (B)

Covalidation is a transfer strategy in which the receiving lab independently reproduces key validation experiments and demonstrates it meets the same performance criteria originally established — without direct simultaneous comparison with the originating lab. Comparative testing, the other major strategy, has both labs analyze identical samples and compares their results directly. The choice between strategies depends on regulatory requirements, practical constraints (whether reference samples can be shipped), and method complexity. Both strategies require predefined statistical acceptance criteria.

Question 3

If a method has been rigorously validated at the originating laboratory, the receiving laboratory can assume it will perform equivalently without formal transfer studies.

• T. True
• F. False

Answer: False

Validation at the originating lab demonstrates performance under specific conditions: particular instruments, analyst training, reagent lots, column batches, ambient temperature and humidity, and water quality. All of these differ — even subtly — at the receiving lab. A 'validated' method is validated at that lab; it is not automatically portable. Method transfer is precisely the formal process of verifying that these real-world inter-laboratory differences do not prevent the method from performing equivalently at the receiving site. Without transfer studies, the receiving lab has no documented assurance that its results are scientifically or legally interchangeable with the originating lab's.

Question 4

Equivalence testing in a method transfer study controls both the risk of accepting a genuinely non-equivalent method and the risk of incorrectly rejecting an adequate method.

• T. True
• F. False

Answer: True

This dual control is the key statistical advantage of equivalence testing over simple comparison of means. A conventional t-test only controls the false-positive rate (the risk of concluding a difference exists when it doesn't). Equivalence testing sets a predefined acceptance interval (the maximum difference deemed acceptable for the method's purpose) and requires that the observed difference, with its confidence interval, falls within that interval. This simultaneously controls both error types: false acceptance of a non-equivalent method (false positive) and false rejection of an adequate method (false negative). The explicit, prospectively defined interval also makes the decision defensible to regulators.

Question 5

Why is a predefined statistical acceptance framework essential for method transfer decisions, rather than relying on expert judgment after comparing the data?

Model answer: Without predefined criteria, transfer decisions are vulnerable to both false acceptance and false rejection, and are not reproducible or defensible. Post-hoc expert judgment can miss systematic biases that fall within wide confidence intervals around the mean (accepting a non-equivalent method) or can flag normal inter-laboratory variability as a failure (rejecting an adequate method). Predefined equivalence testing or tolerance interval approaches establish before the transfer begins what level of difference is acceptable for the method's intended use. This makes the decision rule explicit and protects both labs: the originating lab is assured its results are not being corrupted in transfer, and the receiving lab is protected against failing for normal analytical variability. Regulatory agencies (ICH, FDA) require documented, statistically principled transfer decisions precisely because a method's regulatory standing depends on demonstrating that data from receiving labs are scientifically and legally interchangeable with originating lab data.