Clinical operations note: a-medical-device-quality-inspectors-checklist-what-i-actually-check-before-it-119

I'm a quality/compliance manager at Globus Medical. I review every product before it reaches customers—roughly 1,200 line items a year; maybe 1,150, I'd have to pull the exact log. In 2024, I rejected 3.8% of first deliveries for spec mismatches, labeling errors, or incomplete sterilization documentation. This is the checklist I actually run, not the one from the quality manual.

Who should use this checklist: If you're involved in supplier quality, hospital purchasing, operating room supply, or even just selecting an autoclave machine for a new surgical center, these are the steps I think matter. There are six of them. Most are quick once you build them into your workflow.

Step 1: Verify the label against the device, not the packaging

The first thing I do is scan the UDI-DI and compare it to the product code on the purchase order. If you've ever had a box opened in the OR and found a different implant model inside, you know this is not a formality.

I also check the legal manufacturer address. People often search 'Globus Medical Limerick PA address'—the QA receiving location in the Limerick area is 2560 General Armistead Avenue, Audubon, PA 19403. Technically it's Audubon, about 10 minutes from Limerick Township. If you're a supplier, that's the address to put in your system; using the wrong ZIP code can delay traceability paperwork.

Step 2: Match sterilization claims to the actual autoclave machine

This is where I see the most expensive mistakes. A device can be labeled 'sterile' and still be a problem if the sterilization method doesn't match what the hospital uses.

For reusable surgical instruments, I require a sterilization cycle that the customer can reproduce. The ANSI/AAMI ST79 guideline for steam sterilization distinguishes between gravity-displacement cycles at 121°C and pre-vacuum cycles at 132–134°C. If someone tells me their tray set works in 'a standard autoclave,' that's not a specification—it's a guess.

We once had a vendor send a set of instruments with labels printed on uncoated paper. After one pass through an autoclave machine, the lot number peeled off. The device itself was fine, but the traceability was gone. That incident cost us a full re-labeling project and delayed the launch by three weeks. (That's the kind of cost people forget to put in a TCO calculation.)

The idea that autoclave selection is only a facilities decision comes from an era when sterilization was isolated in central supply. In practice, the quality team needs to be involved from the spec stage, because packaging and labeling behavior changes under steam and vacuum.

Step 3: Function-test a sample from every lot

I don't trust certificates of conformance alone. For a surgical stapler, that means testing a sample from each lot for firing force, staple formation, and reload lock-out. A stapler might look identical to the previous lot but have a jaw width difference that doesn't show up on a visual inspection.

We also check the anti-firing mechanism. A surgical stapler should not close and fire if the reload is absent or already spent. That's a patient-safety check, not a convenience feature.

One thing that took me years to learn: function testing is not just about pass/fail. It's about consistency across operators. In one audit, I ran a blind test with two OR resource teams using the same stapler model. One team rated it as more difficult to compress, even though the firing force was within specification. We ended up updating the user instructions. The device didn't change; the usability data did.

Step 4: Close the loop for pathology and histology

Here's the step most checklist templates leave out. The device doesn't stop mattering when the surgery is over. If it creates specimens—like a biopsy punch or a tissue resection stapler—the instructions need to be aligned with histology workflows.

What is histology? In simple terms, it's the microscopic study of tissue structure. It matters because a surgical specimen has to be labeled, fixed, and processed correctly. If the device's instructions for use say the specimen should be sent in a specific way, that statement needs to match what the pathology laboratory is equipped to do.

I know, histology sounds far away from a manufacturing floor. But a mismatched instruction can lead to a mislabeled specimen, and a mislabeled specimen can burn through hours of a pathologist's time. That's a cost that shows up on nobody's invoice.

Step 5: Audit traceability and post-market data together

Traceability is not just lot numbers on the box. I check that every device has the right UDI-DI in the accessdata.fda.gov database, then I check our internal system for the same number. If the numbers don't match, the device doesn't ship.

The Globus Medical NuVasive merger closing date—September 1, 2023, if you're tracking the timeline—changed a lot of our internal paperwork. Combining two product matrices from two companies creates duplicate catalog numbers and 'close but not identical' descriptions. The post-market surveillance files had to be merged line by line. Since then, I'm even more skeptical of anyone who says 'the system will catch it.' A system only catches what people entered correctly in the first place.

My experience here is based on about 1,200 line items per year, mostly implantables, instruments, and consumables. If you're doing software-based capital equipment like navigation systems—and Globus does have ExcelsiusGPS—the traceability checklist gets even more complex because the software version becomes a device attribute.

Step 6: Calculate total cost of ownership before you approve a vendor

I don't make buying decisions, but I do approve specifications, and the two look different. People think higher-priced vendors deliver higher quality. Actually, vendors who consistently deliver quality can charge more because the hidden costs are lower. The causation runs the other way.

When we were choosing a new autoclave machine for a customer validation project, the cheapest quote came in about 18% below the second vendor. That same vendor fell apart on documentation, calibration certificates, and delivery time. The $500 'savings' turned into a $1,600 project overrun after expedited shipping, a service visit, and two days of my team's time. TCO—total cost of ownership—includes unit price, freight, installation, validation, downtime, and the cost of an unhappy customer. It's never just the line item.

For hospitals, add reimbursement research to that list. If the vendor's regulatory documentation makes the device hard to code, your revenue cycle team pays the price in phone calls. That's why I pay attention to reimbursement support as part of the quality approval process.

Common mistakes I tell everyone to avoid

Test only one sample. A lot can have 200 pieces. Test three units minimum, ideally from different points in the run.

Assume new packaging works on old heat sealers. The sealer doesn't care what the packaging specification says. Run a seal strength test.

Rely on verbal assurance from a distributor. Ask for the sterilization validation report and the UDI registration confirmation before the PO is signed.

Ignore time in your TCO. A device that arrives with every certificate but is three weeks late costs more than a device that arrives at 5% above budget on time. In an OR, time is clinical risk, not just a delay.

That's the checklist. This was accurate as of early 2025, and device regulations don't sit still, so verify current UDI and sterilization standards before you use it as a formal procedure. If you're a supplier sending us a new item, bring the product code and the intended use statement. We'll match it to the paperwork, put it through sample testing, and if everything holds up, we'll clear it for the next step.

Oh, and I should add: the quality review doesn't end at the loading dock. The next time someone asks you what histology means or why a surgical stapler needs a function test, it's because a device is only as good as the information around it.