A method for deciding what to stock that starts from what the failure stops rather than what the part costs, with the four categories worth holding and the ones that quietly expire in the store.
Most hospital spares stores are built the same way. Somebody once waited three months for a part, so the next order included two of everything that supplier sold. Five years later the shelves hold a great deal of stock, and the ventilator is still down waiting for a filter that nobody thought to buy.
The way out is to stop ranking parts by price and start ranking them by what their absence stops.
Take the equipment inventory and give every item two scores. The first is what happens clinically when it fails: does a service stop, does it degrade, or is there an identical machine next to it? A single steriliser in a hospital with one theatre is a different risk from one of four identical infusion pumps.
The second is realistic lead time from your actual supplier, not the manufacturer's catalogue figure. In Bangladesh a part that is stocked in country arrives in days. The same part ordered abroad, cleared and delivered, is a matter of weeks at best, and longer if it needs an import permit or if it arrives at the wrong time in the financial year for a payment to be released. Ask the supplier which parts they hold locally and get it in writing, because that answer is what tells you which parts you do not need to stock yourself.
Anything that scores high on both, meaning a service stops and the part cannot be here quickly, goes on the shelf. That list is usually much shorter than people expect, and it is worth the money in a way that a general stock of assorted parts is not.
Do not stock anything with a shelf life shorter than the interval at which you will use it. Reagents, sealed lead acid batteries, rubber seals and adhesives all age on the shelf, and a battery bought as an emergency spare three years ago is not an emergency spare, it is a paperweight. Where a battery has to be held, rotate it into service and hold the newer one.
Do not stock what the local agent genuinely holds. Verify the claim once by ordering something, then rely on it. And do not stock parts for equipment approaching the end of support, because those are the parts you will be left holding when the machine is condemned. Ask suppliers, at purchase, how long parts will be available after the model is discontinued, and put the answer in the contract.
Storage conditions are the part that gets ignored. Electronic boards in a humid store corrode on the shelf, so keep them in sealed antistatic bags with desiccant and replace the desiccant. Keep seals and hoses out of direct sunlight and away from heat. Keep the store dry, above flood level, and locked, with a card or a spreadsheet per item recording issue and balance. A spares store with no issue record becomes a store where nobody knows what is there, which is functionally the same as an empty store.
Then attach the whole thing to procurement, because the cheapest time to buy spares is at tender. Require a spares price list valid for a stated period, require the supplier to declare which items they hold in country, and require a recommended two year spares list from the manufacturer as part of the technical submission. Prices offered when a supplier wants the order are not the prices offered when your machine is already down.
Review the shelf list after every breakdown. If a failure stopped a service and the part was not on the list, it goes on. If a part has sat unused for two review cycles and the equipment it serves is still healthy, take it off. Have the list signed by the biomedical engineer and the head of the department, so that stocking decisions are clinical decisions with a name attached, not a store keeper's guess.