Comparing two lots without running a full study

Sooner or later a study outlives a lot, and a replacement arrives with a different number on the vial. The sensible question is not whether the new lot is as good, which is not answerable at a bench, but whether it behaves like the old one under your conditions. That question is answerable in an afternoon, provided the comparison is planned before either vial is opened.

Begin with the paperwork

Put the two lot records side by side before touching anything. Compare the fill, the form, the test list and the dates. Note which tests were run on both lots and which appear on only one, because a comparison that ignores a test present only on the new lot can miss a real difference. If one record has a field the other lacks, write down what you did about it rather than silently skipping it.

A side-by-side, not a study

• Same day, same operator, same diluent batch, same instrument
• Same order of preparation for both lots, not first one then the other by habit
• The same measurements, chosen before the session, with the acceptance criterion written down
• One extra pair of vials held back unopened, in case the first attempt is inconclusive

The point of these constraints is not statistical rigour, which a single session cannot provide. It is that a difference you see on the day can be attributed to the lots rather than to the many small variations that would otherwise be present.

What counts as a difference

Appearance is the easiest thing to observe and the easiest to over-read. A slight colour difference between lots is common and often does not matter; a difference that appears only after preparation is more interesting. Dissolution time, the amount of material left undissolved and the way a solution clears are the practical observations, and each should be written with a number or a description, not an adjective.

Measurements that are part of your normal methods are fair game. Measurements you invent for the comparison are usually not worth the effort, because they will not be repeated and the result will sit in a notebook with no context. Choose what you can compare again.

Deciding what the result means

Three outcomes are possible. The lots behave the same within what you can measure, and the study continues with the new lot. The lots behave differently, and the difference is small enough to bridge with a short overlap period. The lots behave differently in a way that matters, and the comparison becomes a question for the supplier. All three are useful results; only an undocumented comparison is wasted.

The part that gets skipped

Write the comparison up on the day, before the arithmetic moves from memory to guesswork. Two paragraphs and a table are enough: what was compared, how, what was observed, and what was decided. If the answer to a later reviewer is a shrug, the work was done for nothing.

Keep the comparison small enough to repeat

The temptation is to make the comparison impressive: more measurements, more replicates, more instruments. Resist it. A comparison that takes two hours can be repeated next month by someone else; one that takes two days will not be repeated at all, and a result nobody can repeat is a story rather than a record.

Small also means fewer ways for the session itself to differ from the routine. If the comparison session does not look like a normal working day, the differences you observe may belong to the session rather than to either lot.

The note that closes the loop

Close the comparison with a decision sentence, not with data. State which lot the study will continue with, whether a bridging overlap is needed, and the date the next lot will be reviewed. That sentence is what a later reader needs in order to understand why the study changed course, and it takes a minute to write while the session is still fresh.

Keep the comparison note with the lot register rather than in a separate folder. A comparison that cannot be found when the third lot arrives has to be redone, and redoing it is the moment when the first comparison stops being trusted at all.

One more thing worth recording

Note the date each lot was received, not only the date it was first used. The interval between arrival and first use is often where storage questions actually arise, and it is the interval that a reviewer will ask about if a result from that lot ever looks unusual.