3D printing at Point-of-Care... with CARE!
3D technology has found it's way to the hospital and is significantly improving patient care.
Point-of-care as a term may be misleading; the term is normally used as "at the patient's bedside", but with respect to 3D (printing) technology, it means "at the Healthcare Facility". (Probably coined by the marketing department of a 3D print supplier and it stuck).
More and more surgeons are looking at 3D visualizations of CT and MRI data instead of 2D grayscale slices. Reading and understanding 2D grayscale images from CT is a skill that needs to be learned, like reading X-ray photographs, whereas looking at and interpreting 3D models is an innate capacity.
So the use of 3D visualization of anatomy or pathology is increasing. 3D visualization offers detailed insight in complex cases. Also, surgeons really like tangible 3D printed models and see the added value of planning a surgery based on a virtual or physical 3D printed model.
Many companies offer image processing (segmentation), surgical planning and 3D printing services to hospitals; For surgeons to provide their input, they have to collaborate with engineers at those companies, often via online meetings or online platforms.
BUT:
As a result many hospitals are bringing 3D technology in-house, employing biomedical engineers, adopting specialized 3D software and using mainly small desktop printers to print anatomical models cheaply. Surgeons seem to prefer to have such a 3D (printing) facility with engineers inside the hospital (at PoC) and walk in whenever they have some time to spare to discuss their case(s). Once a hospital has an PoC 3D Printing facility, the adoption of 3D technology by various departments speeds up, thanks to the (physical) proximity and accessibility of the technology.
This is an ongoing trend and I fully understand why this is happening. From a hospital perspective it makes sense and I believe a 3D lab inside a hospital is becoming a kind of (supporting) department in its own right, similar to the medical imaging department that supports other departments with MR and CT imaging. And why shouldn't hospitals take advantage of modern 3D technology to improve patient care?
However, I also feel there's need for caution!
Hospitals have no manufacturing experience.
Manufacturing of medical devices happens primarily at medical device companies, who operate under strict regulations (MDR in Europe, FDA in USA,…) and quality management (ISO 13485 or FDA 21-CFR-820) and the ones who employ Additive Manufacturing use professional 3D printers.
Hospitals typically do not produce their own medical devices and as such they have no experience in manufacturing [Ref 1]. On top of that, there is no dedicated education for biomedical engineers that teaches anatomy, image segmentation, design and manufacturing. Specifically additive manufacturing is not yet a fixed part of any (biomedical) curriculum. As a result, many hospital engineers need to reinvent this wheel by trial and error. Sharing Best-Practices (for example in a PoC LinkedIn group) might be part of the solution.
In terms of dimensional accuracy there might not be a significant difference between parts printed on a desktop printer and parts from a professional 3D printer, which is confirmed by many recent publications. However "quality" of 3D printed parts is more than meets the eye, more then dimensional accuracy (see also Quality of 3D printed titanium cranial plates). "Quality" means having control over your processes, so you consistently produce high quality medical devices. I'll get back to this.
3D printing immaturity
3D printing is a relatively new production technology and there are practically no (medical) standards (ASTM/ISO) available. Desktop 3D printers are not developed to be (serial) production machines (nor are they regularly maintained/serviced).
An important distinction between Additive Manufacturing and (the more mature) Subtractive manufacturing (CNC milling) is that in AM you are both processing the material (from raw to final material) and the part simultaneously. Even professional manufacturers struggle with control and validation of the 3D printing process.
The FDA guidance document “Technical Considerations for Additive Manufactured Medical Devices” provides an excellent and thoughtful read of things to consider when 3D printing medical devices [Ref. 2]. I highly recommend reading it!
The “Certificate” pitfall
Biocompatibility certificates
Manufacturers of 3D printers that aim to sell their devices in the medical market offer their customers tables of favourable biocompatibility results (ISO 10993) of parts printed:
So they provide the Raw material with a promise that if you do it right, the resulting part and material after manufacturing could be biocompatible (to a certain degree).
Note: Biocompatible ≠ implantable!
People trust the machine manufacturing or material supplier to provide the settings needed for printing and post-processing, but rarely are the resulting parts validated. Personally I am quite surprised that raw materials (like a toxic resin) can be sold with final device classifications (e.g. a class IIa resin), because in a (PoC) 3D printing facility, many things can go wrong, for example when printing surgical guides on a resin printer (SLA or DLP) [Ref. 3]:
领英推荐
So using a class IIa resin cannot ensure that your final part is a class IIa medical device, right?! Similarly in powder 3D printing (typically PA12), users may not achieve similar printed part quality as the manufacturer.
Software certificates
Various software packages are used to get from DICOM data to a 3D printed anatomical model or surgical guide. On the one hand, you have many biomedical engineers who like to use free (open-source) software, and on the other hand there are hospitals who are risk-averse and gladly pay more for a piece of certified software. Regardless of the software used, you should validate your workflow. Indeed, using certified software does not absolve you from this obligation! Similar to the errors that can occur in 3D printing, many errors can be introduced in the software workflow (e.g. mirroring, scaling, user error, poor design, file format conversions,…).
Medical Device Production System
A relatively new phenomenon is the Medical Device Production System (MDPS); a manufacturer provides the entire workflow from scan to 3D printed device. If you use their software, their raw materials, their 3D printer, their post-processing, their instructions and have your designers and operators trained by them and they maintain everything, then you should be guaranteed to produce one(!) type of medical device safe and effectively. This can obviously be valuable for PoC 3D printing, but regulators are still struggling with this, liability is still a question mark and to make this commercially interesting for a hospital, I guess said hospital would need to be making many of this type of devices. Let’s keep an eye on these developments.
Certificates in short
Obviously, certificates do have value. The manufacturer of the material or software has spent effort, time and money to get this certificate. They provide proof and a sense of trust that you too can achieve good results, provided you fully control the process now, next month and in a few years still. When working with an uncertified supplier, a supplier audit is advised, whereas a certified supplier has already been audited by an independent 3rd party. Anyway, you should always validate your process yourself.
Risks
Printing anatomical models is obviously a low risk application. At the most, there's an inconvenience during surgery if a fixation plate that was pre-bent onto the model doesn't fit on the anatomy of the patient. However, it already becomes riskier if an entire surgery was planned on a mistakenly mirrored anatomical model.
Printing surgical guides is medium risk; the patient can be exposed to potentially toxic material debris (as a result of sawing or drilling through the guide).
And obviously printing long term implants is high risk! My main concern is that when a hospital would be producing patient matched implants and they would structurally start to fail… the volumes are probably too low for any red flags to go up anywhere. These cases might all be treated as individual fails, whereas it might be caused by one underlying error in the process. For this reason, I am in favour of implants being produced by traditional medical device manufacturers (with proper traceability, Post Market Surveillance and recall procedures) and not at PoC. (Although I recognize that for personalized devices it will always be difficult to conclude if a fail is a result of the individual design or caused by a structural error).
Regulatory aspects of PoC 3D printing
Regulatory bodies are struggling to keep up with the technological advancements in general and PoC 3D printing in particular. For example, the FDA recently requested stakeholder feedback on how to regulate 3D printing in healthcare facilities (HCF's) [Ref. 4]. The MDR in Europe now requires a QMS for in-hospital manufacturing, the MDCG recently released specific guidance about the health institution exemption [Ref. 5] and a dedicated ISO standard for medical Additive Manufacturing is under development [Ref. 6]. The Australian TGA seems to be a forerunner in regulating Personalised Medical Devices with well written regulation of personalized medical devices [Ref. 7].
PoC 3D printing facilities may be taking on the role, and thus the liability, of manufacturer in the eyes of regulators and the law. This means they could be subject to product liability if their product causes harm.
Quality Management
“Quality” of a part is often thought to be synonymous with dimensional accuracy. In terms of dimensional accuracy there might not be a significant difference between parts printed on a desktop printer and parts from a professional 3D printer, which is confirmed by many recent publications [Ref. 8]. However "quality" of 3D printed parts is more than meets the eye, much more then dimensional accuracy (see also Quality of 3D printed titanium cranial plates). "Quality" means having control over your processes, so you consistently produce high quality medical devices. And processes means àll processes in a facility, not only the manufacturing process. For example your Human Resources processes; is there proper training and knowledge transfer when the main specialist in a PoC facility leaves? Is everything properly documented and up-to-date?
With conventional manufacturing techniques, implants are mass-produced and every production batch is tested (product validation). When 3D printing personalized devices, you can't test every device, instead you need to rely on process validation, but if you don't control your process, you don't know when it fails (and as said before, the failure might not be immediately visible in the produced part).
A Quality Management System (QMS) is often confused to be a system for Case Management, tracking cases from start (DICOM data) to finish (3D printed medical device) and recording who designed, produced, approved and when. This kind of “traceability” is definitely a requirement in any QMS, but there’s more. Much more! A proper QMS makes your entire facility much more efficient. With the introduction of the MDR, an appropriate QMS is a prerequisite for 3D printing at PoC [MDR Article 5.5b]. What is appropriate? Ideally, this QMS is based on ISO 13485, but this might be overkill for PoC facilities. The MDCG Guidance document lists areas covered by such an appropriate QMS for PoC facilities [Ref. 4], which may be considered ISO 13485 Lite.
Implementation of a QMS seems daunting and expensive. It doesn't need to be! Contact me for help in building an affordable QMS (Lite) that helps your PoC facility become better and more efficient.
Concluding remarks
I know that most PoC 3D printing happens with the best intentions in offering optimized, personalised care. The main purpose of this article is to increase the awareness of the risks involved. A QMS, if anything, forces you to think about risks and how to mitigate them. It will support all your business processes and should make your facility better and your resulting devices of consistent, high quality.
And do please perform the following Litmus test: “Do I feel 100% comfortable of using my PoC 3D printed device when my wife / child is the patient?”
So I strongly urge HealthCARE Facilities to adopt 3D printing with CARE. Only print models or guides under a Quality Management System. Leave printing of implants to the medical devices industry.
References
Clinical Professor (Specialised in Robotic Joint Replacement)
8 个月Erik, Thank you for an eye-opening article. Well written with lots of valid thoughts on 3D printing.
Research Engineer p? Region Sk?ne
1 年It's a great article, and will help lots of hospitals to know what they are in for when they decide to start a 3D print department at the hospital. I just don't agree on the last sentence of the article, implants can definitely be manufactured by hospitals.
Additive Manufacturing Medical Device Engineer.
1 年Indeed, hard to validate! I would say careful selection of CTQ attributes is key here, as well the decision to verify versus validate. At small volumes verification is still somewhat feasible.
Biotech Disruptor | GTM | Leading with Purpose
1 年Really well written; thanks for sharing Erik Boelen!
Medical Device Additive Manufacturing | Regulatory and Product Development Expertise
1 年Nice overview. Curious, have you tracked the first uses of the phrase “point of care 3D printing” to some of the early TGA documents? Clinicians tend to bristle at this term since, as you mentioned, it’s not bedside and I usually point them to TGA’s early work as the origin of the term rather than any company marketing materials. Perhaps you’ve found something different?