TrialNerd_Beth said:

Do it in two steps and it stops being confusing.

I am going to disagree with reconstituting low as a general rule. More diluent, more punctures, more in-use days at room temperature, and the stability trade-off is real. Precision is not the only variable being optimised.

The figures, for anyone assembling their own picture. Practical: unopened, refrigerated at 2 to 8°C; in use, refrigerated and used within the preservative-limited window; never frozen; and a visual check every time — a haze that does not settle is a reason to stop, not to wonder.

DataDave said:

I am going to disagree with reconstituting low as a general rule.

There is a second half to this that has not been said yet. Freezing is the one to avoid, and freeze-thaw more so. Ice-crystal formation and the concentration changes at the phase boundary drive aggregation, and aggregated peptide does not recover on thawing. A vial that has been frozen and thawed is not rescued by returning it to the fridge.

A narrower follow-up, since the general answer is now clear:

How much material I am losing to dead space, and whether that explains why a 10mg vial gives me nine usable draws rather than ten?

Closing the loop on my own question.

Follow-up: it was dead space. Switching to a low-dead-space fixed-needle syringe gave me the tenth dose back.