Small batches and consistency

A batch, or lot, is a quantity of material produced in one defined process run and intended to be uniform throughout. Everything the batch record and the certificate of analysis say applies to the whole quantity, because the whole quantity is taken to be the same. The size of that quantity changes how easy it is to support that assumption, and how easy it is to read the record that backs it up.

One record per batch

Each batch carries a record of how it was made: the raw materials used and their lot numbers, the equipment, the process steps and their settings, the in-process checks, any deviations, and the people who performed and verified each step. ICH Q7, the guideline on good manufacturing practice for active pharmaceutical ingredients, describes what such records contain and requires them to be reviewed before a batch is released.

The record for a short run covers fewer hours, fewer shifts, fewer equipment changes and usually fewer raw material lots. There are fewer places for something to go wrong and fewer pages in which a problem can sit unnoticed. A reviewer can read the whole record, follow each step into the next, and check that every entry makes sense against the entries around it.

Why uniformity is easier to show

A certificate of analysis rests on samples. For those samples to speak for the batch, the batch has to be uniform, or the sampling plan has to account for the ways it is not. Long runs can drift. A reaction temperature creeps upward, a raw material lot runs out halfway and a new container is opened, a filter loads up and flow changes, a different operator takes over at the shift change. Each of these can produce a batch that is uniform within its parts but not across them.

A small batch made in one continuous run, from one set of input lots, on one set of equipment, has fewer of these seams. That does not guarantee uniformity. It reduces the number of points where differences can enter, and it makes each of those points easier to find when someone reads the record.

What the trade-off costs

Small batches are not free. Every batch needs its own record, its own release testing and its own certificate. Making the same total quantity in small runs means more records to review and more tests to run overall, not fewer. The burden of review shifts from reading one long record closely to reading many short ones consistently.

Variation between batches also becomes the thing to watch. If each small batch is uniform within itself but each differs a little from the last, the user meets that difference every time a new lot arrives. This is why consistency is judged across batches as well as within them. A series of certificates for successive lots, tested by the same methods against the same specification, shows whether the process makes the same material each time. Plotting those results over time, and investigating a result that moves away from the others even while it stays inside the limits, is a standard part of quality review.

Reading a short record

• Check that the raw material lots listed in the record match their own certificates.
• Read the deviations section first. A short run with a deviation still needs the deviation explained and its effect on the batch assessed.
• Compare in-process results against their limits, not only the final results.
• Confirm that the batch size in the record matches the quantity the certificate claims to describe.
• Set this lot’s certificate beside the certificates for earlier lots. Consistency is visible across the series, not in any one document.

The size of a batch is a production choice with a documentary consequence. A smaller batch produces a shorter and more complete account of one run, which is easier to check line by line. Whether the account is trustworthy still depends on what is written in it, who wrote it, and whether the entries were made at the time the work was done.