An inventory-strategy guide for biomedical and asset managers, current as of September 2026. It gives no vendor pricing, makes no guarantee of uptime, and offers no clinical risk guidance.
Most equipment downtime is not caused by a difficult fault. It is caused by a known fault on a device whose replacement part is three weeks away. That is an inventory problem, and unlike many operational problems it is solvable with a method: decide which devices matter most, decide which parts protect them, and decide how much of each to hold.
Downtime is an inventory problem
Stock decisions feel like purchasing decisions, so they get made on price. The cost of the part is usually the smallest number in the calculation. The larger number is what happens when it is unavailable: a cancelled procedure, an idle room, a rented replacement, or a service call that costs more than the part because it was urgent.
Framing the stock plan around consequence rather than price produces a different list, and usually a shorter one. Few facilities need to hold many spares; most need to hold the right ones.
Ranking device criticality
Start by ranking the fleet, not the parts. Three questions sort most equipment quickly: what stops if this device is unavailable; how long can the service function without it; and how quickly can a replacement be obtained from outside? Devices that score high on all three justify their own analysis; devices that score low can wait for the next review.
| Criticality tier | Typical characteristics | Stocking approach |
|---|---|---|
| Critical | Failure stops service; no realistic substitute; long external lead time | Hold the parts that prevent extended downtime, plus a standby device where feasible |
| Important | Failure disrupts but does not stop service; alternatives exist | Hold the parts with the highest failure rate and shortest time-to-fail |
| Standard | Failure can be absorbed; replacement or rental is available | Hold consumables and fix on failure |
Selecting parts worth stocking
For each critical device, the parts worth holding are the ones that fail most often, the ones whose absence extends downtime most, and the ones that are hard to obtain at short notice. The three criteria frequently point at the same items: filters, seals, cables, batteries, sensors and interface components.
- Parts with a documented failure history on your own fleet, taken from maintenance records.
- Parts whose external lead time exceeds the downtime the service can absorb.
- Parts that a qualified technician can fit in a routine visit, so that holding them actually shortens the repair.
- Parts with a shelf life that matches the time they will sit on the shelf.
Where your own records are thin, the fleet’s manufacturer documentation and the service provider’s experience are the next best sources. Either is better than a stock list inherited from another department with a different case mix.
Stock classes and levels
Once the parts are chosen, assign each a stock class that describes how it is managed rather than what it costs. A workable three-class structure distinguishes parts held for critical devices, parts held because they fail frequently, and parts held opportunistically because they are cheap and slow to obtain. The class determines the review frequency, the reorder trigger and who owns the decision.
Set levels from failure rate and lead time rather than from intuition, then revisit them annually. A stock level that was right when lead times were six weeks is wrong when they are six months, and the correction cost is small compared with the interruption it prevents.
Budgeting and sequencing
Spares compete for the same budget as equipment, and they lose to it because equipment purchases are visible. Two arguments help. First, the stock plan is what protects the equipment already bought; a capital item with no spares provision is not fully funded. Second, the comparison is not part versus part but part versus downtime, and downtime has a number in most service budgets even when it is not labeled as such.
Sequence the build over more than one budget cycle if necessary: start with the critical tier and the highest-frequency parts, and expand as the records show which failures actually occur.
A sequence that works for many departments starts with the parts that protect the two or three most critical devices, adds the highest-frequency failure items across the fleet, and stops there for the first cycle. The second cycle, informed by a year of actual failures, is usually more accurate than any plan built from a catalogue ¡ª which is an argument for reviewing the plan annually rather than trying to design it perfectly the first time.
Reviewing the stock plan
- Review failure records annually and after any significant downtime event.
- Review lead times for the parts held on critical devices.
- Check expiry dates on consumable and shelf-limited parts.
- Confirm that the parts held can still be fitted by the people who perform repairs.
- Retire parts for equipment that has left the fleet.
Questions managers ask
How much stock should we hold?
Enough to cover the failure rate over the lead time for the parts on critical devices, and less for everything else. The number comes from your own records rather than from a general rule.
Should we hold a spare device instead of parts?
For the highest-consequence equipment, a standby unit often protects service better than a parts shelf, because it removes the repair time as well as the lead time. The two approaches can also be combined.
What if we do not have failure records?
Start with the manufacturer’s maintenance documentation and the service provider’s experience, then build the record as you go. The first review usually reveals that a small number of parts cause most of the interruptions.
Rank the fleet, choose parts by consequence, and set levels from lead time. If you are sourcing parts or equipment and want the documentation that travels with a unit, ask HHG for the records on a specific item. Related reading: how lifecycle management reduces equipment risk and the hospital and clinic lifecycle playbook.