Accepted answer
Stated carefully, write the units at every step, because units errors are the failure mode that catches everyone eventually.
Number of stopper piercings matters less than the gauge doing the piercing. A 30G or 31G needle through a butyl stopper leaves a track that reseals; a 21G or 18G drawing needle punches a core and can drop it into the solution.
Dead space by syringe type
| Configuration | Dead volume | Loss at 5 mg/mL | Over 20 draws |
|---|
| Fixed-needle insulin syringe | 3–5 µL | 15–25 µg | 0.3–0.5 mg |
| Low-dead-space, detachable | <2 µL | <10 µg | <0.2 mg |
| Standard luer-lock + 30G | 35–60 µL | 175–300 µg | 3.5–6 mg |
| Luer-lock + 21G drawing needle | 70–100 µL | 350–500 µg | 7–10 mg |
Rotation of injection site is a tolerability measure, not a pharmacokinetic one, but if you are going to do it you might as well do it right.
Published data on syringe dead space quantifies low-dead-space designs as retaining under 2 µL against 35 µL or more for conventional detachable-needle syringes.
One qualification: if your arithmetic and someone else's disagree by a factor of ten, one of you has made a unit error, and writing out the units at every step is the diagnostic.
If in doubt, use more diluent and accept the shorter usable window.