The commonest thing we push into a patient is a short tube of plastic over a steel needle. Here is where it came from, why it is meant to soften, and a question worth asking about the device itself.

A thing we trust and rarely examine

A peripheral cannula is one of the most familiar objects in a hospital, and one of the least examined. Millions go in every year. Very few of us stop to ask what the part left sitting in the vein is actually made of, or when anyone last looked hard at its design rather than at how we use it. This is a short, curious history of that small tube, and it ends with a question.

Steel, where it all starts

The needle that pierces the skin is stainless steel, and it always has been. It is coated so that it slides cleanly out of the plastic, and the shape of its tip is engineered to reduce the pain of insertion. For the first half of the twentieth century the steel simply stayed in the vein.

It worked, but a rigid metal tube in a moving vessel is unforgiving. Retained steel devices are associated with more infiltration and tissue damage than modern flexible catheters, which is why the literature moved away from leaving steel in the vein for routine use. That history matters, because "we could always go back to steel" is a tempting thought in a shortage. The honest answer is that steel is still with us as the insertion needle, but as the part that stays behind it was set aside for good clinical reasons, not forgotten by accident.

Plastic arrives, and then gets better

The part we call the cannula, the bit that stays in the vein, is where almost all the change has happened, and it followed the plastics industry.

The first catheters were made from polyvinyl chloride. PVC was stiff, did not slide easily off the needle, and was associated with a high rate of infusion phlebitis and bacterial adherence. Then, in 1969, Teflon arrived. Teflon catheters were more flexible, more resistant to bacterial colonisation, self-lubricating, and non-toxic and tissue-compatible, though still, in the word the textbooks use, tough.

Tough is the clue. A catheter needs to be firm enough to thread into a vein, then kind enough to live there for days. That tension, firm on the way in and soft once settled, is the whole design problem, and it is what the next material set out to solve.

Polyurethane, and the softening story

The first polyurethane over-the-needle catheter was introduced in 1983. Polyurethane proved less traumatic to veins. Studies found it roughly halved severe phlebitis compared with Teflon, and because it was more flexible it kinked less and may have carried a lower infection and thrombosis risk.

One particular polyurethane, marketed as Vialon, became the name people remember. It is worth being clear that Vialon is not a different material from polyurethane. It is a branded, patented formulation of it, in the way that a named stainless steel is still stainless steel. Nipro, MedSource and others make their own polyurethane catheters under their own names; they are different formulations, not different materials.

And polyurethane came with a lovely idea: rigid at room temperature, so it inserts well, then softer once it is inside you, so it irritates the vein less. Firm on the way in, kind once settled. Exactly the thing Teflon could not quite manage.

Here is the curious part. The softening is real, but almost nobody independent has ever measured it. Every clinical study from the 1980s and 1990s, the ones cannulation courses still cite, compared polyurethane against Teflon on phlebitis rates. None of them measured how much the catheter actually softens at body temperature. Softening is stated as the reason, never quantified.

The only hard numbers sit inside the manufacturer's own patents. A 1992 bench test, soaking the tube in saline at body temperature and measuring the force needed to bend it, put the softening of existing polyurethane catheters at about 60 per cent, and newer chemistry at 90 to 95 per cent, with most of the change happening in the first few minutes. The figure often seen on product materials, that a catheter softens up to 70 per cent in the vein, sits neatly inside that patent range. The one peer-reviewed paper that repeats the 70 per cent attributes it to a famous 1991 trial whose published abstract contains no such measurement at all.

So the softening is genuine and the citations are real, but where the 70 per cent actually comes from is, put politely, a manufacturer's bench figure that has been repeated until it sounds like an independent finding. After four decades we are still largely describing the softening rather than measuring it, and that is a strange place to be for the commonest invasive device in medicine.

An aside: what would we do if we ran out?

During the pandemic, a few ordinary things started to run short. Medical glass was one of them: the small vials that hold a vaccine are made from glass, and glass is made from sand, and suddenly the world needed billions of them at once. Cannulas were another, simply because so many were needed so quickly that supply could not keep pace.

In our communities of practice at the time, the question came up plainly: what would we do if we ran out? We were short of oxygen and of plastic; what if the materials could not be shipped at all? One half-serious answer was that at least we still had plenty of steel needles and could go back to those. But the real instinct was the opposite, to bring in something new rather than retreat to something we had set aside, and to be certain it was safe, known and tested before it went anywhere near a patient.

That question has not gone away. We are watching disruption to shipping and to oil again now, and oil is where these plastics come from, because polyurethane and Teflon are both petrochemical. A cannula is, in that small sense, a piece of the global supply chain pushed into a vein. When the chain shakes, it is worth knowing how exposed we are and what the honest alternatives are.

So what is next?

There is some genuinely new thinking. Researchers have built an experimental needle with a core that is rigid enough to insert and then turns flexible at body temperature, and in laboratory and animal work it caused less inflammation than conventional devices. Others are developing self-healing polyurethanes that mend themselves in blood near body temperature. Both are laboratory science, not approved products, and the road from a promising material to a device safe for millions of patients is long. But the direction, rigid when you need it and genuinely soft afterwards, is exactly the old unsolved problem looked at freshly.

A question worth asking

Here is what we keep noticing. Almost everything we can find on cannulas is about the human side: technique, dwell time, dressings, securement, how to reduce failure through how we use the device. That work is valuable, and much of it is ours. But the design of the cannula itself, the material, the structure, how it behaves in the vein, seems to have been left largely to the people who make it.

When did anyone last independently analyse the device structure, rather than take the maker's word for it? If you have found research on cannula design, on materials, or on anything emerging inside the device rather than around it, we would genuinely like to see it. This is education and discussion, not clinical guidance, and nothing here recommends a change to practice without device-specific evidence and proper governance review.

References

Phillips, Dougherty & Collins (eds), Venepuncture and Cannulation, Wiley-Blackwell 2011, chapter 4; Maki & Ringer 1991 (Annals of Internal Medicine 114:845); Gaukroger, Roberts & Manners 1988 (Anaesthesia and Intensive Care 16:265); McKee et al 1989 (Journal of Intravenous Nursing 12:288); Jacquot et al 1989 (Annales Françaises d'Anesthésie et de Réanimation 8); Stanley, Meister & Fuschuber 1992 (Southern Medical Journal 85:883); the softening figures from Becton Dickinson's patents (US 5,266,669 and related); and, for the emerging materials, work published in Nature Biomedical Engineering (2023) and Chemical Engineering Journal (2024).