Nine of My Eleven Ordering Mistakes Were Spec Errors. Here's the 9-Point Check That Catches Them.

If you order medical supplies for a living, the most expensive word in your purchase history is "assumed."

Nine of the eleven significant ordering mistakes I've made in eight years — roughly $3,900 of about $4,300 in total waste — were specification errors, not price errors. Not one of them happened because I chose the more expensive vendor. Every single one happened because I verified the price and assumed the spec.

For context: I run operations for a mid-sized outpatient wound and ostomy service. Three treatment rooms, 28 supported step-down beds, and a small specimen bench. My purchase orders cover ostomy and wound care products, a centrifuge machine, and a steam sterilizer — three categories that have no business sharing a checklist, except they kept failing in exactly the same way.

So I wrote the checklist down. Nine items. Six minutes. We've used it on roughly 70 orders since, and it has flagged 47 potential errors — most of them trivial, two of them not.

The four mistakes that built it

Short version, because the lessons matter more than the war stories.

2017 — the "standard size" that wasn't. I reordered our standard foam dressing from memory. Twenty-four boxes, 10 × 10 cm. Our treatment chairs had been standardized on 7.5 × 7.5 cm nine months earlier. That's $610 of dressings that sat in the store room until we'd used them up on the wrong wounds. No patient harm. Plenty of unhappy nurses.

September 2021 — the centrifuge that couldn't reach the g-force our protocol needed. Our written protocol was in RCF (relative centrifugal force, in ×g). The quote was in RPM. I compared them as if they were the same number. They are emphatically not. The conversion is RCF = 1.118 × 10⁻⁵ × radius (cm) × RPM², which means a machine spinning at 4,000 RPM at a 10 cm radius delivers about 1,790 ×g — not the 3,000 ×g in our procedure. The instrument was fine. The rotor was wrong for what we were doing, and the correct one added $1,150 to the project.

2019 — the Coloplast paste. This one still gets discussed at new-staff orientation, so let me be specific, because the "Coloplast paste — how to use it" question comes up in every single onboarding and I got the answer wrong the first time.

We brought in Brava Paste for a 30-bed unit dealing with a lot of peristomal skin creasing. I trained the night staff myself. Ten minutes, in a corridor, and I described paste as "the glue." It is not glue. Paste is a gap filler — it goes into the creases and dips around the stoma so the baseplate has an even surface to seal against. Applied like an adhesive, in a thick ring, it does the opposite of what you want: over-application can work its way under the baseplate and break the seal that was already holding. Thin, not thick. And the tube is not the whole procedure — the sequence around it (clean, dry, measure, fill, then apply) is the part people skip.

Cost: about $740 in extra pouches and an estimated nine hours of nursing time over three weeks, plus the retraining session I had to run in front of the entire unit to correct it. The product did nothing wrong. My ten-minute corridor training did.

Q1 2023 — the sterilizer bought by chamber size. I sized the autoclave by how much would fit inside it, and picked a gravity-displacement unit because it was cheaper. Our load mix included wrapped hollow instruments and porous packs. That's not a gravity-displacement conversation. When I asked "how does an autoclave work" I should have been asking "which cycle type do my loads actually require," and those are different questions. Restocking cost $680 and we lost six weeks waiting for the right unit.

The checklist

Nine items. The order matters less than doing all nine before the purchase order goes out.

Products (items 1–3)

1. Confirm the item code, not the description. "Foam dressing" is a category, not a product. So is "barrier ring." Descriptions are how you end up with a case of something adjacent to what you meant.

2. Confirm the size against a measurement taken this month. Wounds change. So do stoma dimensions after surgery. A dressing sized to last quarter's wound is how a store room ends up holding two sizes and a nurse ends up cutting one of them.

3. Confirm the function, not the name. This is the paste lesson generalized. Paste, powder, barrier rings, skin barriers, and adhesive removers are five different jobs. If you cannot state in one sentence what a product fills, flattens, protects, or removes, don't put it on the order yet — ask the clinical specialist. Coloplast's clinical support people will answer this for free, and in my experience faster than the sales contact will.

Equipment (items 4–6)

4. For a centrifuge machine, confirm four things together: the unit your protocol is written in (RPM or RCF), the rotor radius, the tube and adapter sizes you actually use, and whether the rotor is sold separately. I got two of those four right.

5. For sterilization, confirm the load type first, then the chamber size. Unwrapped solid instruments, wrapped instruments, hollow loads, porous packs, and liquids all behave differently, and small steam sterilizers are classified for exactly this reason — Type N, Type S, Type B under EN 13060, with Type B being the class that covers wrapped and hollow loads. Chamber volume is the last question, not the first.

6. Confirm consumables and validation are in the same budget line. Pouches, biological indicators, and — for vacuum-assisted units — Bowie-Dick tests are not optional extras. If they aren't on the quote, they'll appear three months later as an unbudgeted surprise.

Process (items 7–9)

7. Put training on the purchase order as a line item, not a favor. Two hours of clinical inservice with the right specialist is the cheapest thing on the whole order and the one that decides whether the order works. More on that below.

8. Confirm restocking terms and lead time before you need them. Ask what happens if the spec is wrong. The answer tells you how much the vendor expects this to happen.

9. Get a second pair of eyes from someone who doesn't report to you. Mine is the CSSD lead. She catches about one thing in eight. That one thing is usually the expensive one.

The part that surprised me

I expected risk to scale with price. It doesn't.

My log counts two things: dollars and staff hours. The equipment errors — the rotor, the sterilizer — account for the biggest dollar figures. But the four lowest-value lines in the whole log (paste, powder, adhesive remover, dressing size) account for roughly 60% of the hours we lost, against about 25% of the dollars. The reason is boring: consumables get used by everyone, every shift, in volume. A wrong equipment spec is one expensive conversation. A wrong consumable spec is forty small conversations a week until somebody notices.

Here's the other thing nobody told me. The 8% volume discount I negotiated in 2022 saved us about $1,900 that year. The two hours of clinical training that came with the same account — free, unglamorous, and originally scheduled as an afterthought — eliminated roughly $2,300 of rework in the fifteen months after. I have never once negotiated my way out of a spec error.

Where this stops working

Three honest limits.

It's for planned orders. In an emergency you don't have six minutes. Keep a pre-approved alternate spec on file for your top twenty lines so the urgent order is a lookup, not a decision.

It doesn't fix a wrong clinical judgment. If the wound assessment or the stoma assessment is off, the correct product applied correctly is still the wrong answer. The checklist catches procurement errors. Clinical reasoning belongs to the wound care nurse and the stoma nurse, and this is not a substitute for them — it's a net that catches the things they'd catch if they had time.

Your IFU and your facility's policy override everything here. My cycle times, temperatures, radii, and g-forces are illustrative of how the parameters relate, not instructions. AAMI ST79, EN 13060, the manufacturer's instructions for use, and your infection-control lead govern what's actually correct in your building. Where mine and theirs disagree, theirs wins, and you should assume I made a mistake somewhere — my track record is right there.

And the fair objection: six minutes times forty orders a year is four hours. If you've had zero spec errors in two years, you don't need this. Keep the four hours. I needed it because for eight years I was a guy who confirmed the price and assumed the spec, and the only thing that changed was writing the assumptions down.