PM Fundamentals

Preventive maintenance (PM) exists to catch predictable wear before it causes an unplanned failure or a patient-safety event - it is not a paperwork exercise, and treating it as one is the single most common way a PM program quietly stops doing its job while still technically running. A PM program is built from three inputs working together: the manufacturer's recommended service interval and procedure, the equipment's real failure history at your facility, and applicable regulatory requirements - for example, Joint Commission accreditation standards in the US require a documented, risk-based inclusion criteria for the medical equipment management program, not a blanket "service everything annually" policy.

Inclusion criteria: deciding what gets a PM at all

Not every piece of equipment in a facility needs a formal PM program. The standard inclusion-criteria model scores each device on function (life-support vs. diagnostic vs. accessory), physical risk (electrical, mechanical, chemical), and maintenance requirements (manufacturer-mandated vs. none). Equipment scoring above a facility-set threshold goes into the formal PM schedule; equipment below it can be managed on a run-to-failure or simplified inspection basis. This isn't a shortcut - it's how a biomed department concentrates limited PM hours on the equipment where PM actually reduces risk, instead of spreading effort evenly across everything regardless of real consequence.

Interval logic

Critical/life-support equipment - ventilators, defibrillators, infusion pumps, anesthesia machines - is typically serviced every 6–12 months regardless of usage. Lower-risk equipment can often move to usage-based or extended intervals once real failure data supports it: most accreditation frameworks explicitly allow an evidence-based deviation from the manufacturer's default interval when your own data justifies it - this is sometimes called an Alternative Equipment Maintenance (AEM) program, and it requires documented rationale (failure history, risk classification, and a qualified individual signing off), not just an informal decision to service something less often.

What a complete PM record actually contains

Every PM should leave a record of:

• What was inspected - the specific checklist items, not a generic "PM performed" note.
• What was found - including negative findings ("no wear observed"), which matter just as much as positive ones for trending.
• What was replaced, with part number and lot number where relevant.
• The as-found and as-left functional test results - both numbers, not just the final passing one, since the delta between them is what reveals drift.
• The technician's identity and the date.

An undocumented PM is functionally the same as a skipped one during an audit or an incident investigation - if it isn't written down, it didn't happen, and that standard applies regardless of how confident you are in your own memory of the work.

PM vs. corrective maintenance vs. calibration

These three terms get used loosely but mean different things. PM is scheduled, preventive, and performed whether or not a fault has occurred. Corrective maintenance is unscheduled, performed in response to a reported fault. Calibration is the specific subset of work that adjusts or verifies a device's measurement accuracy against a traceable reference - see Chapter 14. A single PM visit often includes elements of all three (a scheduled inspection that also happens to catch and fix a developing fault, plus a calibration check), but they should be logged as what they actually were, not lumped into one generic "service" entry - the distinction matters for both regulatory reporting and for trending which category of work is actually consuming your team's time.