What accept/reject threshold would you set for dulaglutide before seeing the result?

The method section is present, which is unusual enough that I want to make use of it.

I would like to set this up properly once, rather than adjust it repeatedly.

My budget is real but not tight, and my tolerance for uncertainty is low.

How would you structure this, and what thresholds would you set in advance?

In practice, two vials tested from a lot of ten is very different from two vials tested from a lot of ten thousand, and most certificates do not state the lot size.

If you have reason to suspect inhomogeneity — different appearance in different vials, or a long or warm shipment — testing more vials is the diagnostic move.

What each test answers

It helps to be literal here: for a quantitative result like content, the acceptable range determines how many vials you need to test to establish the lot complies.

I would treat a "complies with" statement without sampling details as a claim rather than as evidence.

If testing multiple vials, state how many you tested and why you chose those vials.

Specifically, a certificate that reports one test result on one vial extrapolates to claim that all two hundred vials in the lot are identical, which is an assumption worth questioning.

Under AQL sampling plans, testing two vials from a fifty-vial lot gives you an operating characteristic curve that tells you what risks you are accepting.

If the entire lot failed qualification, a retest on a different vial is sometimes done, but reporting a retest result under the same lot number is misleading.

Published data on lot homogeneity from manufacturers who sample multiple vials consistently find variation below the published specifications, suggesting the sampling plans work.

The caveat is that sampling is a trade-off between cost and confidence, and neither test nor assumption is cost-free.

Assume segregation is possible, and design your sampling to catch it if it exists.

Thermal history during shipping is different for every vial, so a lot that experienced thermal abuse may have internal variation even if it was originally homogeneous.

Acceptance Sampling by Attributes defines the number of samples you need from a lot to claim a specified quality level at a specified risk — it is in ANSI standard Z1.4.

Specifically, stratified sampling — testing one vial from the top, one from the middle, and one from the bottom of a shipment — is cheap insurance against segregation.

The limitation is that you cannot know for certain without testing every vial, and you almost never can afford to do that.

The practical summary: a lot number without a sampling statement is a lot number without meaning.

edited 19 Aug 2024 by RP_C18 — added a caveat about sampling

Batch testing establishes what can be claimed about the lot as a whole, and the sample size determines how much you can actually claim.

The statistical foundation here is well-established, which is why sampling plans from decades ago are still valid.

Lyophilised peptide homogeneity studies show that vial-to-vial variation is usually small but occasionally large, depending on the distribution in the freeze-dryer.

Worth noting that thermal excursions during shipping affect different vials differently, so the lot may not be homogeneous even if it left the factory that way.

If testing multiple vials, state how many you tested and why you chose those vials.

The failure mode here is publishing a result that applies to the tested vial alone while implying it applies to the entire lot.

Published segregation failures show that even modern automated processes sometimes produce lots with measurable vial-to-vial variation.

Sampling plans for pharmaceutical manufacturing are defined in ISO 2859 and ANSI Z1.4, and they are based on statistical sampling theory.

One qualification: testing more vials gives better confidence, but at some point the cost outweighs the benefit.

Assume segregation is possible, and design your sampling to catch it if it exists.