Schrödinger's Kitten

Irreverent Science for Everyone

Tuesday 24 March 2020

Can we 3D print medical equipment to help the COVID-19 response?

  • policy
  • practical
  • medical

Yes, but maybe we shouldn't

DISCLAIMER: I am not a lawyer. I am not a medic. I am not a virologist. I am a doctor of research, with 4 years of scientific activity focusing on 3D printing — though not with plastics — and I have spent 10 years working with and supporting researchers in medtech and engineering. This is intended to raise questions that should be thought about, and connect up people to relevant existing work. I am not offering solutions to a complex public health problem.


3D printing can be quickly mobilised and modified, and there are printers around the world available for use. However the concept of 3D printed medical equipment still has key problems, which include guaranteeing sterilisation, mechanical strength, and issues of scaling. Depending on the needs of medical professionals, there could be a role for 3D printing, but experts in virology, medical care, 3D manufacturing, and supply chains would need to agree standards and acceptable levels of risk in order to produce something useful to the medical community. Testing and legal liability would also be key areas to get agreement and clarification. Makers can best help by contributing to organised initiatives, and targeting help where there is an explicit, stated need.

Background: 3D printing in a pandemic

As of March 2020 there is a global pandemic caused by COVID-19, a novel coronavirus which causes respiratory illness. The countries currently most severely infected have overloaded medical systems, and there are supply shortages of some medical equipment such as respirators, to help severely sick people breathe, and facemasks, which allow healthcare workers to minimize their risk of exposure while caring for the sick. Some of these shortages are caused by unexpected demand, such as members of the public stockpiling masks, and some by limitations on supply — as most facemasks are manufactured in China, the first country hit. Multiple suggestions have been made that 3D printing, a rapid prototyping technology for which many machines and users exist and are sitting idle, could be used to help alleviate shortages. There’s a few things that need to be considered before doing that. Providing unhelpful things or materials that don't do what people are trusting them to do is worse than useless.

Motivation: The potential of 3D printing

There is some evidence that 3D printing can be useful in this situation. The valves needed for crucial ventilators were reverse-engineered and printed by Italian engineers. Josef Prusa and the Prusa team in Czechia have modified an existing one-use 3D printable/laser cut facemask design and are producing 800 a day. Even if large scale factories would be a better solution, if the factories aren’t online there’s a possibility that 3D printing could fill some kind of 'better-than-nothing' gap. There are thousands of 3D printers in the UK alone, and many people who would like to help by running things off. Key questions here are:

  • Quality and fit for purpose of what can be produced — are 3D printable things useful here?
  • Expectations and quality assurance. Short notice devices will not be the same as the regular PPE; irregular volunteers cannot be scrutinised and guaranteed like a factory geared up to produce medical devices in sterile conditions.
  • Legal liability: if the prints are not sterile, if they break, if they don’t provide the protection the wearer thought they would, what is the situation? No-one wants to cause damage but accidents happen.

Specifications: Needs and abilities

Giving a worker a false sense of trust in their equipment can be worse than no equipment at all. It’s important that medical organisations have a clear understanding of what 3D printed equipment can and cannot do; I’ve outlined my thoughts below. (When I say 3D printing below, I’m referring to plastic, fused deposition modelling — squeezing out molten plastic — as that’s the most common and cheapest kind. I am aware there are others. I have a PhD in one.)

Equally, 3D print manufacturers need to understand what the needs and requirements of medical workers are — this is a question I am less qualified to talk about, but it would absolutely need to be done. They also need some sense-checking from virologists — for instance, the ‘3D printed mask containing copper’ will do nothing much to kill airborne viruses and is based on a faulty understanding of the science. I’ve roughed out some considerations here too.

What 3D printing can and cannot do

3D printing strengths:

  • No set-up costs — costs virtually nothing to change the design, so easy to personalise or improve
  • No set-up time — other than doing the design, you can print things off straight away without needing to prepare equipment
  • Many machines all around the world — could manufacture simultaneously near to point of use, no single point of failure if a group of people get sick
  • Can do small batches — if you only need 4, you only print 4. Less waste.

3D printing limitations:

  • Mechanically weak — they’re built of layers, which introduce potential cracks and weaknesses in the material
  • Can't be reliably sterilized — this is the biggie. 3D prints are porous and have a rough surface at a scale of fractions of millimeters — plenty of areas for viruses and bacteria to hide. They are also mechanically weaker than cast pieces. Sterilization processes usually involve:
    • High pressure and high temperature — this warps, weakens and damages the parts
    • Radiation sterilization (UV or gamma radiation) — can also damage prints
    • Chemical treatment (again, bad for parts, also could absorb chemicals in pores and cause damage to wearer. See introduction of paper on 3D printing surgical parts). Some work suggests sterilisation issues can be overcome (See Stratasys/University of Texas paper on sterilisation but concerns remain. Testing required!)
  • No economies of scale — this is another big one. 3D printing 1000 things takes 1000 times as long as one thing. It’s not the most efficient way to do things.

This is a very boring point, but: 3D printing is sexy and a lot in the media. Often there are other, older technologies more suitable to churning products out cheaply at scale — injection moulding and good old-fashioned sewing spring to mind. I’m guessing there’s already talks with appropriate factories to do this, but it will take a while to whip up the metal moulds and set up required to make a new product, a problem that 3D printing doesn’t have, of course.

On the other hand, 3D printing sucks at scale (see above). The only situation it can work is by using the novel properties of distributed manufacturing — having many small local facilities making small runs as needed, saving time on manufacturing batches and shipping. But in this situation that would need oversight, organisation and sterilisation procedures to guarantee standards and get things to where they are needed.

What medical experts do and do not need

I am not that kind of doctor and I don’t have a list, but a key question here is what degree of assurance medical workers need and want from novel equipment. Obviously they should have access to the safest equipment possible. So:

  • What are the ideal standards of safety? Can they be met?
  • Okay, what would be acceptable?
  • If that’s not possible, what's the closest: reusing equipment, sterilising as best they can with spray between each use, or using freshly 3D printed equipment which is newer, made to different standards and less extensively tested? Is there agreement here?
  • Which components (facemasks, valves or tongue depressors) would be most appropriate and useful to 3D print?
  • If printing is suitable, are there existing designs?
  • How many would be needed?
  • What timescale are they needed in and for?

This requires medics, virologists, and biosafety folks to have a chat with 3D printing folk, manufacturers and supply chain people about what is possible and doable. If the products that can be made aren’t as good as current options, it shouldn’t be done, no matter how cool it would be. If there is something to be done that meets their needs, the group would need to agree testing standards, volumes, specifications, and think about how to get it manufactured in this challenging situation.


Standard testing for medical equipment is long, thorough, and to national and international standards (as it should be). Clearly there isn’t time for the full standard certification now. Equally clearly, you don’t want to roll something out in the hundreds or thousands and discover a flaw you could have fixed if you’d tested ten first. So a balance needs to be found between those things to determine what level of protection, with what certainty, is suitable given the urgency of the situation. I’d think this would depend on the medics’ requirements, the availability of alternatives, the demands and the risk levels involved, but I am not an expert here.

Manufacturing options that might be worth exploring

3D printing in a sterile environment: the heat coming from a 3D printer head is sufficient to sterilise it. HOWEVER you would need to guarantee the part was handled with clean, sterile gloves, placed into sterile packaging, and transferred to the end user without any interruptions. Guaranteeing this sterile treatment in transitioning from printer to final user would be difficult to certify for manufacturers not used to this. Those who run medical printers in sterile environments (university and start-up businesses) and have trained users could produce a few thousands of masks to the Prusa specs above. Very dedicated users with no experience might also be able to get the same results, but verification would be important.

Distributed 3D printing, parcelled out to individuals and non-specialists, could be run off at volume, then assembled and sterilised centrally. This assumes that:

  • 3D printable designs are fit for purpose — the work of the Prusa team with the Czech health ministry is hopeful here. Ideally, a reputable organisation (oh Prusa, hi!) could look after a repository of useful tech, ensuring that the models are up-to-date and genuinely helpful to the medical teams who need them. (Idea from Charles Yarnold, thanks!)
  • Sterilisation in some of the ways mentioned above could be ‘good enough’ to remove viruses while not doing too much mechanical damage to the equipment, but would need testing. There are some studies (on gamma radiation, low temp sterilisation) that suggest that these techniques can be used, but if they were reliable they’d already be in manufacture).
  • Distributed manufacturing and delivery to where things are needed could be managed — this is quite a logistics challenge! E-nable, who print prostheses, might be a good group to learn from. Or retailers might be prepared to lend tech.

Other ideas: varnishing or post-treatment could improve resilience and reduce porosity; could other industries, such as garment manufacture, be re-purposed to make a better substitute (admittedly less cool than additive manufacturing)?

In summary, there might be something to be done, but we really need to communicate clearly and honestly between experts in different disciplines in order to make sure what is offered is actually what is provided and what is needed. It’s important everyone has a clear understanding of the trust they can place in their equipment, and if the most radical possibility — of distributed 3D printed manufacturing — is useful, then someone reputable and with the right skills needs to guarantee the correct models are used and sent to the right places.

But I still want to be helpful!

Hopefully you can! Remember to check if your idea is truly helpful before you do it. Don’t just assume your concept is ‘helping’ — our medical services and civil organisations have a lot on their plates without you adding to it.

Here’s some ideas:

  • You can donate computing power to analysing the virus with folding@home
  • You can register at Crowdfight COVID to offer your science and technical skills to COVID researchers, also sign up to a mailing list where they send out their problems for volunteers to solve.
  • Check in with your neighbours (from a safe distance) to see if there’s anything they need help with. Search ‘covid mutual aid’ to find groups organising support, or read more about support groups here.

If you specifically want to use your 3D printer:

  • There’s a hackathon to design a mechanical ventilator: CoVent Challenge / OSVentilator on Twitter.
  • If noone has already done it — look first, they don’t have lots of spare time to deal with enquiries! — contact local hospitals, care homes and doctors surgeries (think carefully about how to contact, don't go in person and don't take up medical time) and see if there is anything with less stringent requirements than full medical technologies — eg. door openers you can use without hands — that they need, or that you can provide them with.

Other ideas? Other people collecting opportunities? (Send email) or tweet me and I’ll add them.

And remember: stay inside, wash your damn hands.

Many thanks to Charles Yarnold for the suggestion of a repository with oversight, Prof Candice Majewski for the suggestion of a section with ideas for what people can do, and Dr Anna Ploszajski and Rikard Anglerud for reading.

PS. I'm aware I haven't updated for a while; I got a PhD, it was a lot of work. Will write more because I'm going to be inside a lot.

Content: Scary Boots — Design: Canis Lupus