AMC versus CMC: the difference is spare parts
The distinction between an annual maintenance contract and a comprehensive maintenance contract is simple to state and expensive to get wrong. An AMC typically covers scheduled preventive maintenance visits and the engineer's labour for breakdown calls, but spare parts are billed separately as they are consumed. A CMC covers the same service plus the replacement parts, usually with a defined list of exclusions. The AMC is cheaper on paper and carries the parts risk; the CMC costs more and transfers most of that risk to the vendor.
Which is right depends on the failure profile of the equipment, not on a general policy. For equipment where the expensive consumable component is the thing that fails — the x-ray tube, the ultrasound probe, the detector — an AMC can produce an annual bill several times the contract value in a bad year, because the part that fails is precisely the part excluded. For robust equipment with cheap parts and predictable wear, an AMC is usually the better economics.
Read the exclusions before comparing prices, because that is where the two contracts actually differ. A CMC that excludes the tube, the probe, the battery, the detector, and anything classed as a consumable is an AMC with a comprehensive label and a comprehensive price. Ask the vendor directly which components are excluded and what each costs to replace, and put the answer in the file with the contract.

Contract questions to settle before signing
- Exactly which parts are included and which are excluded by name
- Number of preventive maintenance visits per year and what each covers
- Response time and resolution time commitments, and the penalty if missed
- Whether standby or loaner equipment is provided during extended downtime
- Whether calibration, software updates, and applications training are included
- Uptime guarantee, how uptime is calculated, and what remedy applies
Capture the warranty at commissioning or lose it
The most common and most avoidable loss in biomedical asset management is a warranty that expired without anyone noticing, followed by a maintenance contract purchased months late or a repair paid for that was covered. The root cause is always the same: the warranty terms were never recorded as structured data at commissioning. They exist in a purchase order and an installation report in a file, and no system knows the date.
Commissioning is the only moment when all the information is present and someone is paying attention. Capture, at that moment and into the asset register, the asset identifier, make, model, serial number, department and physical location, supplier, purchase and installation dates, warranty start and end dates, warranty scope, included preventive maintenance visits during warranty, and the vendor's service contact. Doing this at commissioning takes fifteen minutes and doing it eighteen months later takes a day of archaeology.
Then set the alert well ahead of warranty expiry — enough time to evaluate the AMC or CMC quote, negotiate, obtain approval, and execute, which for most hospitals means several months rather than several weeks. Equipment falling into an uncovered gap between warranty expiry and contract start is a routine and entirely preventable exposure. Where the asset register lives in a platform such as HealUDoc alongside department and location data, the expiry calendar and the responsible department are already linked, and the alert reaches the person who owns the equipment rather than a general inbox.
Preventive maintenance schedules that actually get performed
Preventive maintenance schedules fail for scheduling reasons far more often than technical ones. The visit is due, the equipment is in use, the department cannot release it, the visit is deferred, and the deferral becomes permanent. The fix is to schedule preventive maintenance against the clinical calendar in advance rather than treating it as an interruption — a booked slot in the theatre or radiology schedule, agreed with the department, in the same way a case is booked.
Distinguish vendor preventive maintenance from in-house checks, because both are needed and they operate on different cycles. The vendor visit is periodic and detailed. In-house checks are frequent, simple, and performed by the biomedical team or trained users: battery condition, alarm function, physical damage, accessory availability, and cleanliness. Many failures that present as breakdowns are actually an uncharged battery, a missing accessory, or a damaged cable, and a weekly in-house check catches those before they become an emergency call.
Every visit must produce a record: what was checked, what was found, what was replaced, what remains outstanding, and the engineer's name and signature. Unsigned or missing service reports are among the most frequent accreditation findings in this area, and they also destroy your negotiating position at renewal, because you cannot demonstrate what the vendor did or did not deliver.

Measuring uptime and downtime in a way vendors will accept
Uptime is only a contractual lever if both parties compute it the same way. Define, in writing, when downtime starts — the moment the fault is logged, or the moment the equipment becomes unusable — and when it ends: the moment the engineer declares it fixed, or the moment the department confirms it is back in clinical use. Those two definitions can differ by hours or days, and vendors and hospitals will naturally choose the one that favours them.
Define also whether downtime is measured against calendar hours or scheduled operating hours. A CT scanner in a hospital that runs it around the clock and one that runs a twelve-hour service have very different exposure to the same eight-hour repair. Agree this at contracting, because arguing about it during an outage is a poor use of everyone's attention.
Then log every fault call with the timestamp it was raised, the timestamp the vendor was notified, the timestamp the engineer arrived, the timestamp of resolution, and the timestamp of return to clinical service. That log is the single most valuable document you will hold at renewal, and keeping it in the same asset register as the contract — a platform such as HealUDoc can hold both against the equipment record — means the renewal conversation starts with evidence rather than recollection. Two intervals matter separately: response time, which is how long until someone arrives, and resolution time, which is how long until the equipment works. Vendors often meet the first and miss the second, and only a log distinguishes them.
Downtime log fields that make the record usable
- Fault reported timestamp and who reported it
- Vendor notified timestamp and ticket reference
- Engineer arrival timestamp
- Resolution timestamp and the cause identified
- Return to clinical service timestamp confirmed by the department
- Parts replaced, whether covered under contract, and cost if not
Spares lead time is the number nobody tracks
The most damaging downtime is rarely a difficult repair; it is a simple repair waiting on a part. Imported components, customs clearance, and vendor stocking policy can turn a two-hour fix into a three-week outage. Yet spares lead time is almost never tracked as a metric, which means hospitals negotiate contracts on price and response time while the actual driver of unavailability goes unmeasured.
Start recording, for every part replaced, the date ordered and the date received. Within a year you have a lead-time profile per vendor and per part category, and it will change your procurement decisions. It also gives you the basis for a specific and winnable negotiation: rather than asking for a better price, ask for a committed parts availability window, local stocking of the critical components, or a loaner unit if a part exceeds a defined lead time.
For critical equipment where the hospital cannot tolerate an outage, consider holding your own buffer stock of the components with the longest lead times and the highest failure frequency. This is a real inventory cost against a real availability benefit, and the lead-time data is what lets you make that trade on evidence. Identify your critical equipment explicitly — the items where an outage stops a service rather than inconveniencing it — and apply this analysis to that list first.

“Our average repair took two hours of engineer time and nine days of waiting for a part. Once we could show the vendor that number, the conversation about local stocking became much shorter.”
The renewal decision: contract, self-maintain, or replace
At renewal there are more than two options, and hospitals frequently consider only the first. You can renew the contract as offered, renegotiate scope or price using your service history, downgrade from CMC to AMC or upgrade the other way based on actual parts consumption, move to a pay-per-call arrangement, take maintenance in-house where your biomedical team has the capability and the vendor permits it, or replace the equipment. The evidence for choosing between them is the service history you have been keeping.
Work the arithmetic as an illustrative comparison. Consider a machine on a CMC at a given annual cost, whose service history over three years shows a certain number of breakdown calls, a certain set of parts replaced, and a certain amount of downtime. Add up what an AMC plus those actual parts would have cost. If the parts consumption is well below the CMC premium, you are paying an insurance premium against a risk that has not materialised — though for equipment with one catastrophically expensive component, that premium may still be rational.
Bring the replacement option into the same analysis rather than treating it as a separate capital conversation. Ageing equipment shows a signature: rising call frequency, rising parts cost, lengthening downtime, and a maintenance quote that climbs each year as the vendor prices its own risk. When annual maintenance cost plus the cost of downtime approaches the annualised cost of replacement, the contract renewal is no longer the right question. That is a conclusion you can only reach with a maintained asset history, which is the real argument for keeping one.



