Medical Device Quality Has Changed. Hospital Procurement Hasn't.

After five years as a quality compliance manager at Coloplast—reviewing roughly 40 medical device categories a year before they reach customers—I've reached a conclusion that doesn't always make me popular in procurement meetings: the way hospitals evaluate device quality is about eight years behind where the industry actually is.

That's not because procurement teams are careless or uninformed. It's because the mental model most of us inherited—device quality equals spec compliance—was designed for an era when devices were simpler. And it's still hanging around.

I should know. I used to review devices that way myself.

In 2020, when I first stepped into this role, I approached every review the same way. Does it meet the dimensional tolerances? Does the sterility documentation line up? Does the test data match the declared specifications? I treated quality like a fixed property—something you could confirm with a ruler and a checklist.

It took me a few years and a handful of uncomfortable conversations with clinicians to understand that I was solving the wrong problem.

What Changed? Just About Everything

Think about an ICD device for a moment. In 2015, a quality implantable cardioverter-defibrillator was defined by a few measurable things: shock delivery reliability, battery longevity, lead performance. We reviewed those specs thoroughly. We were right to.

But look at what the device has become. Modern ICD devices communicate with remote monitoring platforms. They upload diagnostics automatically (meaning the device sends data to a clinic-accessible portal without the patient having to do a thing). They run algorithms designed to reduce inappropriate shocks. Some can be updated wirelessly—a feature that didn't even exist in the spec sheets of the previous generation.

So what does quality mean for a device like that? It means interoperability with the hospital's existing infrastructure. It means cybersecurity is part of the review, not an IT afterthought. It means data accuracy matters beyond the sensor. A device that generates false alerts or unreliable trend data isn't high-quality, even if every physical component meets specification.

This was true 10 years ago when device ecosystems were simpler. Today, it isn't even debatable.

The frustrating part is that the old specs didn't become irrelevant. They became insufficient. That's a much harder problem to fix in a procurement process, because it means adding complexity instead of swapping out old requirements.

The Hospital Bed Problem

Try a simpler example: the hospital bed. Ask a procurement committee what makes a quality hospital bed and you'll get a familiar list: weight capacity, side rail strength, pressure redistribution, ease of cleaning. All legitimate criteria. I've written them into specification documents myself.

(I've also watched a hospital spend significant money on beds that met every one of those criteria, then discover the beds didn't integrate with their nurse call system. That was an expensive lesson—and it wasn't the hospital's fault. It was the evaluation criteria's fault.)

Today's hospital beds are data hubs. They send alerts, track patient position, adjust pressure support automatically, and feed information into the electronic health record. So when I review a hospital bed in 2025, I don't just ask what it can hold. I ask what it connects to. A bed that can't speak to the systems around it isn't high-quality. It's a well-built island.

The same logic applies to patient monitoring, and it's a question that comes up constantly: what does a patient monitor measure?

Heart rate, blood pressure, SpO2, respiratory rate. That's the textbook answer, and it's technically correct.

But spend a shift on a busy hospital floor and you'll find that monitors do more than measure. They generate alarms. Some are actionable. A lot aren't. So a monitor with great sensors but poor alarm management can actually be worse than a "less capable" monitor that gives clinicians information they can confidently act on. That's a quality consideration that has nothing to do with component accuracy—and everything to do with how the device functions in a real clinical environment.

The surprise wasn't the technical differences between monitoring systems. It was how much those differences affect what care teams can actually do with the data.

The Chronic Care Shift

Where I changed my mind most is in chronic care, not high-acuity devices.

My company, Coloplast, develops ostomy, continence, wound, and skin care products. As the person who reviews Coloplast products when they come through our quality gate, I used to evaluate them the same way I'd evaluate any manufactured medical device. Sterility, adhesion strength, dimensional consistency, packaging integrity. The whole checklist.

All of that does matter.

But here's what took me years to understand: a wound dressing or an ostomy pouch isn't used once by a clinician and then discarded. It's worn by a patient—sometimes for days, often repeatedly, always while that patient is trying to live a normal life. So the real quality questions are entirely different:

• How long does the product stay secure before it needs replacing?
• Does it damage or irritate the skin when removed?
• Does it allow the person to shower, sleep, and move without fear of leakage?
• Does it support enough confidence that the person keeps using it consistently?

People assume better materials and more engineering drive better patient outcomes, and that's what justifies a higher price. In my experience, the causation runs the other way. Products that improve outcomes demonstrate their value in use, and that demonstrated value is what supports investment in better materials and design. The price follows the outcomes, not the other way around.

In other words, the quality of a chronic care product can't be fully measured in a lab. It has to be measured in a person's life.

It took me about three years and dozens of hospital visits to reach that conclusion. Partly because it's genuinely complicated. And partly because the old measuring habits are kind of comfortable.

Regulations Raised the Floor, Not the Ceiling

Some of my quality colleagues would say regulatory compliance already covers this ground. To be fair, the framework has tightened considerably.

The EU Medical Device Regulation (MDR 2017/745) raised the bar for clinical evaluation and post-market surveillance, and it's still reshaping the industry as its deadlines continue to phase in. ISO 13485:2016 strengthened quality management systems in ways the old 2003 version didn't imagine. These standards genuinely improved medical device safety.

But regulatory compliance tells you a device is safe to bring to market. It doesn't tell a hospital whether a device works well in its specific ecosystem—with its specific EHR, its specific staffing model, its specific patient population. That's a different bar entirely, and I think it's the one that matters most to outcomes.

What was best practice in 2020 may not apply in 2025. That's not a criticism of the old practices. It's just a description of progress.

Bringing Quality Into the Present

I'm not 100% sure my conclusions apply to every hospital in every country. The ones I've visited have their own pressures, budgets, and legacy systems. But I've seen enough to believe the direction is consistent: devices are becoming more integrated, and quality evaluation needs to integrate with that reality.

If you're in procurement, or you're a clinician involved in device selection, ask yourself this: when was the last time you updated the criteria themselves—not the vendors, not the contract terms, but what you actually consider quality?

A quality process that hasn't been revisited in five years is probably measuring a 2020 model of the world.

The fundamentals haven't changed. Safety, sterility, accuracy, durability—all as important as they ever were. But the execution has transformed. A quality process that doesn't account for integration, interoperability, and real-world usability will approve devices that meet every checklist item while failing the test that matters most: whether they help patients get better.

Device quality isn't just what the device is. It's what the device does for the patient using it.

That was always worth more than a spec sheet. It just took me five years and about 200 reviews to understand it.