EOQ for hospital pharmacy under a 99.5 percent service-level constraint
Hospital pharmacy EOQ runs against a 99.5 percent cycle-service-level constraint and a next-day wholesaler cadence, so Wilson Q* is almost always smaller than the safety-stock floor. The binding constraint is the ROP, not Q*. EOQ still matters for non-critical items where carrying cost actually drives the decision[1].
| Annual demand D (vials / year) | 3,650 |
| Order cost S ($ / order) | $8 |
| Unit cost C ($ / vial) | $95 |
| Carrying cost i (% / year) | 35 percent (refrigerated) |
| H = i * C ($ / vial / year) | $33.25 |
| Lead time L (days) | 1 |
| Daily demand sigma_D (vials) | 2.5 |
| Z for 99.5 percent CSL | 2.576 |
| Annual holding | $698 |
| Annual ordering | $695 |
| Annual total (ex. purchase) | $1,393 |
USP 797 and 800 storage rules
Compounded sterile preparations under USP 797 and hazardous drugs under USP 800 carry storage-area and segregation requirements that materially raise the fixed cost of carrying inventory. Translate those rules into a higher i (we used 35 percent above rather than the 25 percent default) rather than try to model them as separate constraints on Q.
Daily delivery cadence
A wholesaler that delivers every weekday makes the safety stock low (L is one day) and keeps Q* small. The optimisation question shifts to which items go on the daily JIT pull and which sit on the standing PAR. Standard ASHP guidance treats vital-rare-emergency items as JIT and everything else on PAR levels managed against Q* and ROP[2].
When Wilson is the wrong tool
Vital-rare-emergency items (snake antivenom, antitoxins, dialysis emergency stock). For those, the cost of a stockout is unbounded; the answer is two units on the shelf, period, regardless of Q*. Wilson is for the long tail of routine items where carrying cost is the real decision variable.