Mistakes

Costly Mistakes in Medical Device Manufacturing and How to Catch Them

The recurring errors that make medical device cost, cleanroom, sterilization, and validation numbers wrong, each paired with the symptom that flags it and a numeric fix.

The most common medical device costing error is loading gowning and airlock time into a flat overhead rate instead of tracking it per shift. Symptom: your Medical Device Unit Cost lands 15 to 25 percent under actual on labor. Root cause is treating a 6 to 10 minute full gown up plus degown as unproductive noise. Fix: run Cleanroom Gowning Time honestly. Two entries and exits per operator per 8 hour shift at 8 minutes each is 32 minutes, or 6.7 percent of paid time gone before a single part is touched. Bill that time into the loaded rate, not overhead.

Cleanroom cost per part goes wrong when people spread facility cost over installed capacity rather than actual output. Symptom: Cleanroom Cost Per Part looks great on paper but the P&L bleeds. A Class 7 room running at 45 percent utilization carries the same 500 to 900 dollars per square meter per year in HVAC, HEPA changes, and monitoring whether it makes 100,000 or 220,000 parts. Root cause is using nameplate volume as the denominator. Fix: pull real numbers from Cleanroom Utilization first, then divide fixed room cost by parts actually produced. At 45 versus 90 percent, your per part burden doubles.

Sterilization cost per unit is routinely underquoted because the batch is assumed full. Symptom: Sterilization Cost Per Unit quoted at a nickel comes in at 12 to 15 cents. Root cause: an EO or gamma cycle costs the same 1,200 to 3,500 dollars whether the chamber is 60 percent or 100 percent packed, and half loads are common during launch. Fix: check Sterilization Batch Capacity against your real pallet pattern. A chamber rated for 40,000 units that ships at 26,000 raises unit sterilization cost by 54 percent. Never divide cycle cost by theoretical capacity.

A unit error that hides for months: mixing dose units in radiation sterilization. Symptom: validation dosimetry fails or overkill margin looks impossible. Root cause is conflating kGy with krad, where 25 kGy equals 2.5 Mrad, a factor of 100 that people fat finger. A target of 25 kGy entered as 25 krad is 100 times low and will not achieve a 10 to the minus 6 sterility assurance level. Fix: lock the validated dose in kGy across the router, the certificate, and any Sterilization Cycle Time model, and reconcile against the VDmax25 method dose before release.

Bioburden sampling gets mis budgeted when frequency is set once and never scaled to volume. Symptom: Bioburden Sampling Cost per unit spikes on short runs and quality flags out of trend results. Root cause is a fixed 3 samples per lot applied to a 200 unit lot and a 20,000 unit lot alike. At 40 to 120 dollars per plate count assay, three samples on 200 units adds up to 1.80 dollars each, versus 0.018 dollars on 20,000. Fix: tie sampling plan to lot size per your SIP, and treat recovery efficiency below 50 percent as a data validity failure, not a low reading.

Validation cost blows up when engineering and conformance batches are counted as saleable. Symptom: Validation Batch Cost per unit looks trivial because three PPQ lots are amortized across a full year of demand that has not been ordered yet. Root cause is spreading a fixed 150,000 to 400,000 dollar validation over projected volume instead of committed volume. Fix: amortize validation over the firm purchase order quantity only. Spread 250,000 dollars across a committed 100,000 units and it is 2.50 per unit; assume a hoped for 500,000 and you have silently understated cost by 2.00 per unit.

Cleanroom labor gets double optimistic when both utilization and yield are assumed high at once. Symptom: Cleanroom Labor Cost per good part is 20 to 30 percent light. Root cause: staffing built for 92 percent utilization and 99 percent first pass yield, when the room actually runs 70 percent utilized at 95 percent yield. Those two gaps compound. A part that costs 4.00 in direct clean labor at plan becomes 4.00 divided by 0.95 yield divided by (0.70 over 0.92) utilization, roughly 5.53. Fix: pull both numbers from live Cleanroom Utilization and yield reports, never from the capacity plan.

The quiet killer is forgetting scrap sits behind an expensive process wall. Symptom: unit cost is right at machining but wrong after sterilization. Root cause: a part scrapped after gowning, cleanroom assembly, and an EO cycle carries all that sunk cost, not just material. A 12 dollar molded component scrapped post sterilization has absorbed maybe 9 dollars of clean labor and sterilization, so a 2 percent post sterilization reject rate on 50,000 units is roughly 21,000 dollars, not the 12,000 a material only view shows. Fix: apply scrap cost at the operation where the loss occurs.

Published 2026-07-01.