Taking the question as asked, rather than the general version of it. Dead space is the answer to the missing dose. A fixed-needle insulin syringe holds a few microlitres in the hub and needle after the plunger bottoms out, and on small draws that is a measurable percentage of every dose. Across ten draws it adds up to most of an eleventh, which is exactly the "nine draws from a ten-dose vial" complaint. Luer-lock syringes are worse; low-dead-space fixed-needle designs are better.

LarryQC_SD said:

Dead space is the answer to the missing dose.

Agreeing with LarryQC_SD, and the qualification matters more than the agreement. The arithmetic is right and one caveat matters: a U-40 syringe changes the markings, not the volumes. If someone hands you U-40 and you read it as U-100 you will be 2.5 times out, and that error has a direction — it is always an overdose.

NauseaFreeNow said:

Two of us with identical vials and identical diluent arrived at different unit counts, which means one of us has a conceptual error rather than a…

Same position here, arrived at the long way round. Do it in two steps and it stops being confusing. First concentration: 10mg into 2ml is 5mg/ml. Then volume: a 0.5mg dose is 0.5 ÷ 5 = 0.1ml. Then units, and this is where people go wrong — a U-100 syringe is graduated in hundredths of a millilitre, so 0.1ml is 10 units. The word "units" has nothing to do with milligrams; it is a volume marking that exists because insulin happens to come at 100 units per ml.

From the other side of the consultation, briefly.

Bacteriostatic water quality matters for dosing arithmetic reconstitution. Not all BAC water is created equal:

• Should contain 0.9% benzyl alcohol as preservative
• Must be sterile and pyrogen-free
• Should come in a sealed, intact vial
• Check expiration date — yes, BAC water expires
• Store at room temperature, protected from light

I buy from a reputable medical supply company, not Amazon marketplace sellers. The $3 savings isn't worth the infection risk from unverified BAC water.