I've been handling equipment procurement for our health system for about 7 years now (since 2018, when our biomed team was literally two people and a shared toolbox). In that time, I've personally made—and documented—three significant procurement mistakes, totaling roughly $18,000 in wasted budget. Now I maintain our team's pre-purchase checklist to prevent others from repeating those errors. This is that checklist. Six steps, about 30 minutes per device. It would have caught every mistake except one.

Step 1: Walk to the actual installation site before you order

In 2019, I approved a Fujifilm portable ultrasound for our emergency department. The specs were perfect. The budget was fine. The delivery date worked. What I didn't check was the ED storage room where the machine would actually live.

On paper, the site was fine—"standard 120V outlet." In reality, that room had one shared outlet powering a mini-fridge, a nurse's station computer, and an existing ultrasound unit. Every time the fridge compressor kicked in, the new ultrasound flickered and re-initialized. We paid an electrician $2,300 to run a dedicated circuit. It would've cost nothing if I'd just walked 100 feet to look at the room.

I've made this mistake twice. Yes, twice. The second time was a rack of cardiac monitors in a step-down unit. Same issue: power load with no margin.

Checkpoint: Before you finalize the order, physically walk to the installation site with your phone. Take a photo of the outlet panel. Find the equipment's nameplate specs—volts, amps, frequency, which the brochure won't tell you—in the manual PDF, and compare them against the site. If the site shares a circuit with other devices, loop in your facilities person.

One more thing: if the manual PDF has unreadable nameplate specs, ask the vendor for a higher-resolution version. It sounds picky, but 300 DPI is the standard for printed documentation. If you ever need to print that page for a reference binder (you will), you'll want it legible. And if they can't deliver a legible PDF, that tells you something.

Step 2: Ask "what's NOT included" before you ask "what's the price"

I learned this lesson on that same portable ultrasound.

I went back and forth between two vendors for a week. One quote came in at $28,000. The other at $31,000, but with four pages listing exactly what was included. I chose the $28,000 quote because the number was smaller. I approved the PO and immediately second-guessed myself. Didn't relax until the machine arrived. Then the extra invoices started coming.

The cart was $1,050 extra. The training session: $1,800. The extended warranty continuity package: $1,300. Ultrasound gel, probe disinfection wipes, and a probe case: about $600. All things the $31,000 quote had listed as either included or optional line items.

None of it was a trap. I just never asked.

Here's my rule now: before comparing prices, I email every vendor the same question—"would you please separately list everything that's NOT included in your quote?" If a vendor can't answer that in one straightforward email, that's an answer too. The vendor who lists all fees upfront, even if the total looks higher, usually costs less in the end. No surprises. Simple.

Checkpoint: Create two columns: "included" and "not included." If the second column is empty, you're not done asking.

Step 3: Test the device against your real workload, not the demo setup

A cardiology nurse once asked me, with complete sincerity, "how does a cardiac monitor actually work?" Fair question. The short answer: basically, electrodes on the chest pick up the heart's electrical signal, the device amplifies it, filters out muscle and motion artifact, and displays a rhythm trace. The long answer is in the manual. The practical answer is what you need to test.

The sales demo for our monitors was clean. Simulator waveforms. Perfect. The filtering was impressive—on the bench.

Then the monitors went to the step-down unit, where real patients move, cough, and scratch in their sleep. The same "artifact-resistant" system flagged normal movement as arrhythmia. False alarms. Staff stopped trusting the alarms. And I should've predicted it—I knew the patient population.

This is the exact same mental shortcut that had me buying a fujifilm instax mini 12 for a patient activity group last year. The fujifilm instax mini 12 instant camera reviews all said the same thing: fun camera, good for social settings, terrible in low light. I thought, "our activity group is basically a social setting." First 10 photos? Dark. Our activity room doesn't have great overhead light. The reviews were right. I just cherry-picked the part that fit my assumption.

The camera was an $85 lesson. The cardiac monitors were not.

Checkpoint: For any device that touches a patient, ask the vendor for a demo with your actual patient population. Then ask directly: "what causes false alarms or errors with this model?" If they can't answer, that's your answer.

Step 4: Verify the output actually fits your existing systems

Our plastic surgery department asked me to find a camera for procedure documentation and patient education videos. I searched "best fujifilm camera for video," because their team likes Fujifilm color science. The results (circa early 2025, at least) kept praising the X-S20. Great specs. Good price. The issue wasn't the camera. It was my failure to check the workflow.

Would its video files import into our media archiving system without a conversion step? Did the HDMI output work with the recording setup we already owned? The answer: not natively. We needed to convert every clip. Doable, but annoying. The clinical team deserves better than "annoying."

Same trap with the ultrasound. The images were DICOM, which is standard for medical imaging. But the network export setup needed our IT security team to approve a specific configuration, and that took real time. We now transfer files by USB. Works fine. It's also a daily workaround that I never saw in the product demo.

Before approving the ultrasound, I should also have looked up its FDA 510(k) clearance more carefully. The public database tells you exactly what the device was cleared to do. It's not a quality rating, but the scope can be narrower than the brochure implies.

Checkpoint: Ask the vendor for a sample output file—image, video, report—and test it through your actual system. Five-minute test. Saves you from a five-year workaround.

Step 5: Check the consumables before you commit

The device isn't the product. The consumable is the product.

For that machine, the transducer is the expensive consumable. It wears out eventually, and replacing it costs a big fraction of the purchase price. The gel, the probe wipes, the carrying case—they're all ongoing costs that don't show up in the headline quote.

Same story in our wound care clinic, where they order ostomy supplies constantly. Back in 2022, the team tried switching to a cheaper brand of ostomy supplies to cut costs. On paper, the savings were real. In practice, the pouches didn't fit the skin barrier as well, leaks were more frequent, and patients had to do complete changes more often. The "cheaper" products ended up costing more in supplies and patient comfort. The team switched back.

So now, before signing any contract, I ask three questions:

  • What consumables does this device require?
  • Which consumables are exclusive to the manufacturer?
  • If we substitute a "compatible" consumable, who takes responsibility if something fails?

If a vendor says "no guarantee" on substitution, that's not necessarily a no. It's a risk. And you need to price that risk into the decision.

Checkpoint: Calculate the total cost of consumables over a realistic device lifespan, then compare that total—not the initial quote—when evaluating vendors.

Step 6: Book the 90-day review before you install anything

This step sounds obvious. Almost nobody does it.

At 90 days post-install, sit down with the clinical team and ask three questions:

  1. What stopped working as advertised?
  2. What's the weekly maintenance burden?
  3. What would you change if you could start over?

Our cardiac monitors taught me this the hard way. At the 90-day review, I found out the false alarm issue had been bothering the charge nurses for weeks. Nobody had told me. Why would they? No one at the bedside wakes up thinking, "I should email procurement about this." I had to go ask.

And honestly? I've never fully understood why some devices that perform beautifully in a demo behave differently in the real world. My best guess: it's the difference between a demo engineer and a testing engineer. One works in a perfectly controlled room; the other has watched real clinicians at 2 a.m. If you're a vendor reading this and you have insight, I'd genuinely love to hear it.

Checkpoint: Put the 90-day review on the calendar the same week you submit the PO. Before the device even ships.

Common errors I still see

First: skipping Step 2 because you're afraid of looking difficult. The $1,800 training charge seems trivial now. But this same failure mode is how a teammate of mine signed a service contract where travel time was billed at the same rate as repair time. Ask early. Ask everything.

Second: treating Step 6 as optional. It's not. The feedback loop is the entire point of a checklist like this. Otherwise it's just a list of wishes.

Third: trusting the spec sheet. Specs describe what a device does in a lab. They don't describe what happens at 2 a.m. when a patient is coughing, or when the substitute gel is all the supply room has, or when a busy resident tries to export the images. The checklist exists for those moments.

This worked for us, but our context matters: we're a mid-size health system with in-house biomed support and a predictable procurement flow. If you're a solo clinic with no biomed team, or a giant IDN with a centralized purchasing office, the calculus might be different.

But the core is the same. Walk to the room. Ask what's not included. Test the real case. Check the consumables. Verify the output. Follow up at day 90.

Since we started using this checklist in mid-2023, our team has caught 47 potential errors before purchase. Power problems. Missing accessories. Incompatible consumables. Each would have been a headache. Instead, they're lines in a spreadsheet.

Take it from someone with $18,000 in documented lessons.