Accepted answer
7.5 mg a week is 390 mg a year and 32.5 mg in an average month — put every route on that denominator before comparing anything. Cost per milligram is the only figure that survives the comparison, because the presentations differ: a licensed pen prices a dose, a compounding pharmacy prices a vial, and a research supplier prices a mass. Divide each one's twelve-month cost by 390 mg and the three become the same number in the same unit. Then add what the cheapest route does not include — independent purity and content testing, the vials you discard, and the postage — because a route that needs testing to be trustworthy has that testing in its cost per milligram whether you account for it or not.
Start by listing every cost in the chain, since carriage, testing and wastage frequently exceed the difference in headline price.
Dead-space loss is small with fixed-needle insulin syringes — a few microlitres per draw — and substantial with detachable-needle luer syringes at 35 to 100 microlitres. Across twenty draws that is up to two millilitres of solution.
Certificate red flags and what each implies
| Observation | Implication | How to check |
|---|
| Lot number not on the vial | Certificate cannot be tied to your material | Photograph vial and certificate together |
| No method section | The number is not reproducible | Request column, gradient, wavelength |
| Purity to two decimals, no chromatogram | False precision | Request the trace |
| Test date before manufacture date | Certificate belongs to a different lot | Compare dates |
| Identical figures across lots | One certificate reused | Compare two lots side by side |
| “Sterile filtered” with no sterility test | Process claim substituted for a result | Ask for the sterility report |
Worth being precise here: carriage amortises across the order. Twenty-five pounds of carriage on one vial is £2.50/mg on a 10 mg vial; on ten vials it is £0.25/mg. That single term explains most of the case for larger, less frequent orders.
Independent testing prices at the services this community uses are published and are stable enough to model.
Nothing here is medical advice, and research-use compounds are not approved for human use.
Decide whether you are optimising cost or confidence before you build the model.
4Thank you — this is the answer I was looking for. – plate_count_9k 2 months ago 5I would add a line about writing the accept threshold down first. It is the step everyone skips. – siobhan_deasy 4 months ago add a comment