Most people picture a hospital’s environmental impact as bins of yellow clinical waste and a lot of electricity. That’s the visible part. The bigger part sits somewhere you can’t see: the supply chain.

In England, the NHS produced 25 megatonnes of CO2 equivalent in 2019. That was around 4 to 5% of the country’s total emissions, and roughly 62% of it came from the supply chain rather than from the buildings themselves. Medicines, equipment, and the devices clinicians use every day. Direct delivery of care (heating, lighting, fuel) accounted for just 24%.

So if you want to move the number, you don’t start with the boiler room. You start with what gets bought, used once, and thrown away.

Where the carbon actually hides

Medical devices alone are estimated to account for about 10% of NHS greenhouse gas emissions and roughly 15% of hospital emissions in the UK. In France, the figure for device supply is even higher at 25% of healthcare emissions. Operating theatres are the worst offenders. They generate up to a third of all hospital waste, mostly from disposable instruments and packaging.

The default for the last 30 years has been single-use. It feels cleaner, it’s cheaper to buy, and nobody has to think about reprocessing. The problem is that the cheap acquisition price hides a much larger lifecycle cost, both in money and in carbon.

A systematic review of 61 devices found reusable versions consistently had lower lifecycle footprints. For single-use devices, the carbon hotspot is production and manufacturing. For reusables, it’s the reprocessing. And here’s the thing worth sitting with: research cited by UK health policy work found reusable devices cut carbon emissions by 38 to 56% across the product lifecycle, while often costing less over time.

One study on electrophysiology catheters put it bluntly. Switching from original single-use catheters to regulated, reprocessed ones cut the global warming impact roughly in half, and beat the original in 13 of 16 environmental categories.

That gap doesn’t close on its own. It closes through better device design, smarter procurement, and connected systems that let hospitals track what they actually use.

Connected care moves the patient home

The other lever is keeping people out of hospital beds in the first place. An inpatient bed is expensive in every sense, including carbon.

A study published in BMJ Innovations in 2025 measured this directly. Researchers at a large acute hospital trust compared traditional inpatient care with “virtual ward” care, where patients get hospital-level monitoring at home through connected devices. They tracked 1,260 patients, including frail individuals and people with acute respiratory infections.

The result: an average inpatient bed day produced 37.9 kg of CO2e. A virtual ward bed day produced 8.8 kg. That’s a 75% reduction, roughly four times less carbon for the same level of care.

Across the study period, that added up to about 285 tonnes of CO2e avoided.

There’s an honest caveat the researchers themselves raised, and I’ll repeat it because it matters. Sending one patient home doesn’t shrink the hospital’s total footprint, because the freed-up bed just gets filled by someone else. What it does is let a hospital treat far more patients without building a new, high-carbon hospital wing. As the researchers put it, virtual wards expand capacity and enable low-carbon care without the construction of new high-carbon hospitals.

That’s the quieter win. Avoided buildings. Concrete and steel never poured.

The device is where it starts

All of this depends on the hardware in the patient’s home actually working. A virtual ward is only as good as the connected pulse oximeter, the blood pressure cuff, the wearable feeding data back to a clinical dashboard. If the device drops connection, misreads, or burns through batteries every two days, the clinician loses trust and the patient ends up back in A&E. Which puts the carbon right back on the books.

This is where sustainability and engineering meet. A well-designed connected device is built to last across many patients, sips power instead of draining it, uses recyclable or reprocessable components, and reports its own status so it gets serviced rather than binned. Building that kind of medical device takes embedded firmware, secure data pipelines, and a hardware design that treats longevity as a feature, not an afterthought. The companies that get this right are the ones thinking about the full lifecycle from the first schematic, not bolting on a recycling label at the end.

Software does heavy lifting too. Remote monitoring platforms that flag deterioration early stop a manageable condition from becoming an emergency admission, and emergency admissions are the single biggest chunk of admitted-patient carbon. One analysis of secondary care found emergency admissions made up 45% of measured emissions, dwarfing planned elective work at 7%.

Catch the problem at home, and you avoid the ambulance, the bed, the cascade of single-use devices that come with an acute stay.

Where the savings get real

Three places, ranked by how much they move the needle.

Procurement comes first. The supply chain is 62% of the footprint, so a hospital that shifts its purchasing toward reusable and reprocessed devices, and demands lifecycle carbon data from suppliers, changes the biggest number on the page. Reusables won’t always win, the energy mix used for sterilisation matters, but the direction of travel is clear.

The catch is that procurement teams have historically judged devices on sticker price. A reusable scope costs more up front and gets more scrutiny for it, while a box of disposables slips through because it looks cheap per unit. That math ignores the resterilisation savings, the avoided waste-disposal fees, and the carbon. A Lancet Planetary Health review of UK practice flagged exactly this problem: hospitals evaluate environmental impact inconsistently, so the same item can look fine in one trust and get rejected in another. Standardising how devices get assessed, with real lifecycle data attached, is half the battle.

Remote monitoring comes second. Every avoided admission is carbon that never gets emitted, plus a bed freed for someone who genuinely needs it.

Device design comes third, and it underpins both of the above. You can’t run a credible virtual ward on flaky hardware, and you can’t shift to reusables if the devices weren’t engineered to survive reprocessing.

The trade-off nobody fully escapes

I want to be straight about the limits. Reprocessing has its own carbon cost. Home monitoring shifts some energy use onto the patient’s electricity bill (government figures put average home emissions at about 7.4 kg CO2 per day). And measuring all of this is genuinely hard. The BMJ study found that automated hospital data systems undercounted emissions compared to manual audits, because they missed community and out-of-hospital activity.

So the numbers are directional, not gospel. But the direction is consistent across every study I’ve looked at. Reusable beats single-use on lifecycle carbon. Monitored-at-home beats admitted. Connected, well-built devices make both of those possible at scale.

The NHS has already cut its carbon footprint emissions by 68% against 1990 levels, with 14% of that coming in the last five years. The next stretch is harder, because the easy wins (efficient lighting, decommissioning the worst anaesthetic gases) are mostly done. The remaining gains live in the supply chain and in how care gets delivered.

That’s a hardware and software problem as much as a policy one. The hospitals that hit net zero won’t be the ones with the greenest car parks. They’ll be the ones who rethought what a medical device is supposed to do, and how long it’s supposed to last.