The Most Dangerous Sentence in Medical Device Sourcing
"We can do that."
I've spent four years as a quality compliance manager at a medical device manufacturer, reviewing roughly 250 components and finished products a year. I've rejected around 6% of first deliveries in 2024, mostly due to spec drift and incomplete documentation. The clearest predictor of a quality problem isn't price point or factory age. It's how willing a vendor is to say yes to work outside their lane.
Here's my opinion, and it's not neutral: in medical devices, professional expertise has boundaries—and the companies that admit theirs are the ones you can trust. Manufacturers who claim they can make your infusion pump, your vascular access sets, and your surgical robots with equal mastery are almost always overpromising. In this industry, overpromising gets people hurt.
I understand why suppliers do it. Saying "that's outside our specialty" feels like losing a bid. But I've watched too many partnerships collapse because a vendor accepted work they weren't set up to execute well.
To be fair, scope creep isn't always malicious. Some of it is optimism—engineers genuinely believe they can learn adjacent domains fast. But belief doesn't change device performance or regulatory paperwork.
Specialization Is a Proxy for Quality
Take our syringe pumps. Smiths Medical has built infusion and syringe pumps for decades, and that history shows up in details you don't get from a generalist: consistent flow-rate accuracy as the battery drains, occlusion alarm thresholds that don't false-trigger, a clamp that still holds after 10,000 cycles. These aren't features you can add to a purchase order. They're accumulated engineering knowledge.
Contrast that with a large vendor I evaluated last year. Their catalog included fundus cameras, surgical robot components, dialysis disposables, and a line of anesthesia machines. Impressive breadth, on paper. But when we asked for a customized fluid path on a simple infusion set prototype, their team couldn't answer basic questions about luer connector tolerances. They eventually delivered samples with a 12% leak rate—and admitted they were "newer to infusion" only after we'd burned six weeks.
That isn't a bad company. They're probably excellent at what they specialize in. But their attention is split across product domains that share almost nothing beyond "uses plastic and electronics." You don't want your Smiths Medical syringe pump OEM partner to have a fundus camera project on their mind when they're validating your flow-rate accuracy claims.
The Most Honest Phrase in OEM Sourcing
I have mixed feelings about vendor specialization. On one hand, managing a network of focused suppliers is administratively exhausting—more contracts, more audits, more qualification protocols. On the other, it's saved us more times than I can count.
The best OEM conversation I've ever had went like this: "This molding geometry is outside our core competency. Here are three companies who do it better, and here's how we'd hand off the tooling specs." That vendor earned every subsequent contract we had. They knew their limit, told us clearly, and protected their reputation by not attempting work they'd likely fumble.
I've also seen the rookie version. In my first year reviewing supplier qualifications, I made a classic error: I assumed "medical grade" meant the same thing to every vendor. It doesn't. One well-known manufacturer assured us their injection-molding resin would meet our biocompatibility requirements across their full production run. They'd validated the spec on their high-volume line, but not on smaller batches. We discovered the discrepancy during incoming inspection—8,000 units in storage that didn't meet the USP Class VI test our label claimed. That's a $22,000 redo and a delayed launch that no procurement convenience justifies.
That failure wasn't a lie. It was overconfidence, which is worse—because a liar knows they're guessing, while an overconfident engineer believes they know.
Regulatory Realities Make "Doing Everything" Nearly Impossible
The regulatory environment for medical devices is unforgiving, and it's the strongest case for specialization. A single product category requires ISO 13485 quality management certification, FDA QSR compliance for U.S. shipments, CE marking under MDR for Europe, plus product-specific performance testing that can easily run into six figures.
Each of those is a discipline in itself. Nobody has the bandwidth to be world-class at syringe pumps, dialysis machinery, and surgical robots at the same time. They might share a building, but they don't share a skill set.
Look at "how does hemodialysis work." Sounds simple—it isn't. Blood is pumped through a dialyzer where a semipermeable membrane filters waste products and excess fluid, while dialysate flows counter-current to maintain concentration gradients. Building a hemodialysis machine requires expertise in anticoagulation control, fluid balancing at small volumes, and air embolism prevention. That engineering domain shares almost nothing with syringe pump design: different failure modes, different clinical environments, different regulatory submissions.
A generalist might say they can do both. I'd rather trust a company that says "that's not our space—here's a dialysis specialist." That admission isn't a lost sale; it's a trust-building moment.
This principle applies to the physical details, too. Color-coding on our airway products must be Pantone-matched across production batches—Delta E < 2 for brand-critical colors, per Pantone Color Matching System guidelines. Clinicians memorize color conventions for cuff inflation lines and oxygen ports, and a subtle mismatch in a darkened ICU can create hesitation at the worst moment. Print quality on our instructional booklets likewise requires 300 DPI minimum for graphics and text (standard commercial print resolution). These details are patient safety, not pedantry.
What About One-Stop Procurement?
"You're overcomplicating this," someone told me at a trade show. "We consolidate vendors so purchasing doesn't have to manage twenty relationships." I get that. Centralization saves time, and supplier rationalization is a legitimate strategy. Budgets are real.
But even diversified manufacturers keep separate business units with separate quality systems and separate validation protocols. The brand name on the door doesn't mean expertise transfers from the ventilator division to the endoscopy line. When you buy from a specialist, you're getting a quality management system designed around your product category's specific failure modes.
Granted, some companies do manage broad portfolios successfully. They tend to be very large, and even then, their divisions rarely share engineering talent or quality systems. The exceptions prove the rule: they succeed because each division behaves like a specialist inside a shared brand.
A company that offers one comprehensive quote for a fundus camera and a syringe pump is a generalist. If they're honest, they'll tell you which is their core business. If they're not, you're the one accepting the risk.
Professionalism Means Knowing What Not to Take On
A quick caveat: my experience is based on reviewing about 250 quality-related items a year, mostly in critical care product lines. If you're sourcing for ophthalmology or surgical robotics, your supply chain dynamics might differ. But the principle holds: specialization and reliability tend to move together.
In manufacturing, saying "we don't do that" can feel like a failure. I'd argue it's the opposite. It's what keeps your existing products strong for decades, because you're not draining resources into problems you haven't earned the right to solve.
Clinicians trust Smiths Medical for infusion pumps, vascular access, and airway management because that's what we've focused on for over a century. We don't build fundus cameras. We don't manufacture surgical robots. That's fine—our customers know exactly what they're getting, and so do we.
Professionalism isn't saying yes to everything. It's knowing what you're the best at—and being honest about the rest.
