Master’s Thesis 30 credits June 2020
Manufacturing in Cardiology
A qualitative study of barriers and facilitators from a managemental point of view
Implementing Additive Manufacturing in Cardiology
Annette Wolff and Simon Sandgren
Additive Manufacturing (AM) is a fast-developing technology, that despite its potential is yet to be applied by the mainstream healthcare market. Compared to other clinical areas where AM is applied, cardiology has a negligible market share, why this thesis aimed at identifying barriers and facilitators for AM implementation, as well as presenting a framework with factors to consider when attempting to implement AM.
A literature review outlined aspects currently considered in literature, in relation to the barriers and facilitators of implementing AM in cardiology. To identify barriers and facilitators to AM implementation in cardiology, and to complement the literature review, two exploratory case studies were carried out, which had a comparative design. The case studies took place in Sweden, whereof one has already implemented AM in cardiology, and the other one has not. Purposive sampling was applied to choose the two involved hospitals, while convenience sampling and snowball sampling were used for selecting interview participants. The findings were analyzed using a thematic analysis.
Results show that barriers and facilitators can act on an individual, organizational, and industrial level. Barriers and facilitators were divided into the themes Management, Technology, Network, Behavior, and Market. Aspects falling under the theme Management were mentioned most frequently among the respondents, suggesting that such barriers and facilitators play a significant role in implementing AM, while findings placed in the Network and Behavior theme respectively were not previously addressed in literature. Barriers include, among others, low involvement of leaders, little cross disciplinary collaboration, and lack of innovation resources in health care. Facilitators include, among others, having an innovation culture that supports initiation of projects, providing an easy and intuitive system for ordering a 3D model, and promoting the technology among potential users.
Concluded is that AM implementation faces numerous barriers and facilitators which should be considered before an implementation endeavor. Addressing these on an individual, organizational, and industrial level most likely facilitates the process of AM implementation, leading to a successful and sustainable change in the organization.
Keywords: additive manufacturing, 3D printing, barriers, facilitators, framework, implementation, health care, cardiology
Supervisor: Marcus Lindahl Subject reader: Anders Brantnell Examiner: David Sköld
Faculty of Science and Technology
Ångströmlaboratoriet Lägerhyddsvägen 1 House 4, Level 0
Box 536 751 21 Uppsala
+46 (0)18 – 471 30 03
+46 (0)18 – 471 30 00 Web page:
Popular Science Summary
Additive Manufacturing (AM), commonly 3D printing, is a technology that creates 3D objects through adding layers to one another. It is a relatively young technology gaining more attention in health care since the turn of the millennium. Specifically, concerning heart diseases, the technology’s potential is not fully exploited yet. Compared to other health care areas, the usage of 3D printing is far less, although the technology shows promising results, and heart diseases are among common causes of death. Therefore, this thesis aims to investigate what is hindering (barriers), and what is supporting (facilitators) the usage of AM regarding heart disease treatments.
First, current literature was reviewed to detect known issues connected to the process of integrating (implementing) AM in health care. Then, interviews at two university hospitals in Sweden, one using AM for heart diseases, and one, not using AM for heart diseases, were conducted to complement and update the information found in the literature. Through a thematic analysis of the interview data, five main themes influencing the implementation were identified. In literature, social aspects like personal attitude, relations, or leadership methods gain low to no attention while considering AM implementation barriers and facilitators.
Nevertheless, throughout the interviews, these aspects were emphasized. Specifically, the location and administration of a 3D printing facility played a significant role since this implicates its accessibility for medical doctors, who are the technology’s final users. Further, the interviews showed that AM’s acceptance was higher if leaders were supportive and users (medical doctors) educated about the technology.
Based on the findings, a framework was developed which proposes possible factors to consider when managing the introduction of 3D printing. The factors concern individuals, the organization, and the industry, and can either be barriers or facilitators to the implementation, depending on the framework user’s setting. Since existing, generalist frameworks in literature are complex, the developed framework only focuses on AM implementation, and presents factors in an intuitive way.
This thesis work is done as a part of the Master Programme in Industrial Management and Innovation at Uppsala University and was conducted by two Master students. The workload was distributed evenly; parts of the literature review and theory were written individually and iteratively rewritten and corrected by the other one. The interviews, analysis, and discussion were conducted together, and thus, the results show a common work effort.
The thesis is done in collaboration with a project called AddLife – a competence center for Additive Manufacturing within Life Science – of which Uppsala University is part of.
Firstly, we would like to thank our subject reader Anders Brantnell at Uppsala University for the opportunity to write this thesis as well as for his valuable feedback and thoughts throughout the work. We would also like to sincerely thank the personnel at two case hospitals in Sweden, that despite a busy schedule (and even more so during the Corona crisis), found time to participate in our interviews. Your experiences and insights were the very essences of our work, and we are grateful that you wanted to share them with us.
Lastly, we would like to thank all other people that in one way or another, have helped to guide this work and provided thoughts and feedback.
Uppsala, June 2020
Annette Wolff & Simon Sandgren
Table of Contents
List of tables ... vi
List of figures ... vii
List of appendices ... viii
List of abbreviations ... ix
1 Introduction ... 1
1.1 Background ... 1
1.2 Problem formulation ... 2
1.3 Objective ... 4
1.4 Research question formulation ... 4
1.5 Delimitations ... 5
1.6 Thesis structure ... 5
2 Literature review ... 6
2.1 Implementation of Additive Manufacturing in health care ... 6
2.1.1 Economic considerations ... 6
2.1.2 Standardization and legal considerations ... 9
2.1.3 Application cases and their applicability... 11
2.1.4 Technological considerations ... 14
2.1.5 Cardiology ... 16
2.2 Gaps in research ... 18
3 Theory ... 21
3.1 Implementation Science ... 21
3.1.1 Rationale for selecting determinant frameworks... 23
3.1.2 Conceptual Model ... 25
3.1.3 Consolidated Framework for Implementation Research ... 28
3.1.4 Active Implementation Frameworks ... 30
3.1.5 Integrated Technology Implementation Model ... 33
3.2 Comparison of frameworks ... 36
4 Methods ... 39
4.3.1 Sampling method ... 40
4.3.2 Data collection method - Interview guide ... 41
4.3.3 Interview Participants ... 42
4.4 Data analysis – Thematic analysis ... 44
4.5 Ethical considerations ... 45
4.6 Trustworthiness ... 45
4.7 Limitations ... 46
5 Results ... 48
5.1 Comparison and analysis of both cases ... 48
5.2 Main themes derived from case studies ... 50
5.3 Barriers and Facilitators of Implementing Additive Manufacturing in cardiology .. 57
5.4 Framework ... 60
6 Discussion ... 65
6.1 Barriers and Facilitators of Implementing Additive Manufacturing ... 65
6.2 Factors applicable on the individual, organizational, and industrial level ... 71
7 Conclusions ... 75
7.1 Limitations ... 77
7.2 Implications ... 78
References ... 79
Appendices ... 85
List of tables
Table 1. Economic considerations ... 9
Table 2. Standardization and legal consideration. ... 11
Table 3. Application cases ... 13
Table 4. Technological considerations ... 16
Table 5. Cardiac considerations ... 18
Table 6. Details of the different levels ... 23
Table 7. Summary of the reviewed determinant frameworks ... 38
Table 8. Interview participants ... 43
Table 9. Summary of commonalities and differences between Hospital A and Hospital B ... 50
Table 10. Barriers and Facilitators of AM implementation ... 58
Table 11. Details of each level in the developed framework ... 60
Table 12. Case studies in relation to implementation science literature ... 68
List of figures
Figure 1. Technology adoption life cycle ... 3
Figure 2. Three objectives of the use of theoretical approaches in implementation science and the five categories of theories, models, and frameworks ... 22
Figure 3. Visualization of the Conceptual Model ... 25
Figure 4. Consolidated Framework for Implementation Framework ... 29
Figure 5. Overview of the five frameworks of the Active Implementation Frameworks ... 31
Figure 6. Integrated Technology Implementation Model ... 34
Figure 7. Themes identified from interviews ... 51
Figure 8. Summary of the Additive Manufacturing implementation framework ... 61
List of appendices
Appendix A. Implementation determinants and outcomes in eight determinant frameworks 85
Appendix B. Interview guide ... 87
Appendix C. Initial themes identified at Hospital A ... 88
Appendix D. Initial themes identified at Hospital B ... 89
Appendix E. Codes and themes of each interview ... 90
Appendix F. Management main theme including codes ... 109
Appendix G. Technology main theme including codes ... 110
Appendix H. Network main theme including codes ... 111
Appendix I. Behavior main theme including codes ... 112
Appendix J. Market main theme including codes ... 113
Appendix K. Summary of barriers and facilitators from literature and case studies ... 114
List of abbreviations
2D Two Dimensional
3D Three Dimensional 4D Four Dimensional
AIF Active Implementation Framework AM Additive Manufacturing
CAD Computer-aided design
CFIR Consolidated Framework for Implementation Research
CT Computed Tomography
EBP Evidence-based practice FDA Food and Drug Administration
ITIM Integrated Technology Implementation Model MDR Medical Device Regulation
MRI Magnetic Resonance Imaging PDSA Plan-Do-Study-Act
RP Rapid Prototyping
SEF Socio-ecological framework TA Thematic Analysis
UK United Kingdom
US United States
Medical Device Regulation Theme
An Additive Manufacturing process which combines biomaterials with cells to create structures that imitate natural tissues.
Fabrication of objects through the deposition of material using a print head, nozzle, UV-light, or other printing technology. Until present, the term 3D printing has, in particular, been associated with machines that are low-end in price and/or overall capability.
In non-technical contexts, this term is often used interchangeably with Additive Manufacturing.
A manufacturing process that makes parts from 3D model data and builds it up in a layer-by-layer fashion. In non-technical contexts, this term is often used interchangeably with 3D printing.
A material designed to interact with the body (i.e. a biological system). It can be anything from hydrogels or other polymers to metals and composites.
A branch of internal medicine concerned with disorders of the heart, such as congenital heart disease, coronary artery disease, and valvular heart disease. Cardiologists deals with both diagnosis and treatment.
Key individuals that support the innovation and facilitate the adoption of it.
A factor is used as a more general term for something influencing AM implementation and can either be interpreted as a barrier or a facilitator.
In a descriptive manner, a framework presents an overview of factors believed or found to influence implementation outcomes.
Regulation in Europe to ensure the quality and safety of medical devices being produced in or supplied into Europe.
A theme is a pattern of meaning and can be divided into sub-
The following chapter gives a brief introduction to Additive Manufacturing in health care, use cases, and implementation aspects, followed by the objective and research questions guiding this work.
Additive Manufacturing (AM), or 3D printing, creates objects from 3D model data, by joining materials layer upon layer, as opposed to subtractive manufacturing technologies (ISO/ASTM 52900, 2018). In other words, instead of removing matter (e.g. scraping, dissolving or turning) from raw material to get the desired product, an AM process will add material (e.g. plastics and metals) one layer at a time to create the desired product (Kietzmann et al., 2015). The first layer-by-layer printing technology, stereolithography, was developed in 1986, followed by selective laser sintering and fused deposition modeling in the years thereupon (Ashish et al., 2019; Jamróz et al., 2018; Shafiee and Atala, 2016). Several processes today derive from these technologies, which, since 2015, maintain the ISO/ASTM 52900 standard (ASTM International, 2020; Chua et al., 2020). It covers seven categories of AM processes and their associated technologies. The choice of process depends on the material to be printed, the tolerances, the object’s size, and the required surface finish, among others.
Initially, 3D printing was invented for quickly creating prototypes or basic models, also referred to as Rapid Prototyping (RP). The primary goal was and is to accelerate the speed of product development, specifically concerning increasingly complex products (Chua et al., 2020). Over time, RP was not only used for producing prototypes but also used for manufacturing purposes, which is why the term Additive Manufacturing is used more commonly today. (Chua et al., 2020; Gibson et al., 2010)
AM’s usage in medicine started in the late 1990s, with important developments being a synthetic 3D urinary bladder scaffold modified with the patient’s cells and a functioning miniature kidney printed with cell-seeded bioink, i.e., 3D bioprinting (Ashish et al., 2019;
Jamróz et al., 2018; Hong et al., 2018). The development is ongoing, and recent developments are a 3D printed bionic ear and an artificial heart valve in 2013, as well as 3D printed drugs in 2015 (Ashish et al., 2019). Several clinical areas such as dentistry, orthopedics (Barrow, 2018), and cardiology (Haleem et al., 2018) use AM nowadays. In dentistry, there are relatively many studies exploring AM and diverse application possibilities, shown in Bhargav et al. (2018). In cardiology, most papers (17 papers) are published in 2017 according to a review by Haleem et al. from 2018, which shows that AM in cardiology is a younger and less explored research field.
According to Haleem et al. (2018), the first paper was published in 2000 with a rising interest until 2017 (the last year before the paper’s publication). Ramola et al.'s (2019) systematic review on various AM applications in health care also shows a rising interest in cardiology, with an increase of papers since 2011.
The potential of printing anatomic models in cardiology was already seen in the 1990s (Nguyen, 2017). Nguyen (2017) shows in his paper different cases in which 3D printed heart models helped as a preoperative roadmap. The base for 3D prints is a digital 3D model, which commonly is created by technologies like 2D echocardiography, cardiac magnetic resonance imaging (MRI), and computed tomography (CT). In cardiology, 3D printed models are helpful for complex and rare cases, as well as for the management of cardiac tumors. Luo et al. (2017) mention preoperative, interoperative, and postoperative evaluations, as well as device development and medical education as current applications of AM in cardiology.
With regards to medical education, currently, the Canadian Hospital for Sick Children offers the most advanced surgical training courses during which participants use 3D printed heart models of congenital cardiac defects (Nguyen, 2017). Since 2015, they have conducted eight courses, and the interest is growing (Proto3000, 2018). In Sweden, Skåne University Hospital in Lund came furthest owning five 3D printers and having an order system that allows doctors to order 3D (heart) models the same way as they order radiographs. Recently, Skåne University Hospital attracted attention related to AM in cardiology; Swedish news showed a case of a child having a complicated heart defect, which was solved with the support of a 3D printed heart model (Heiberg, 2020a; TV4 Nyheterna, 2020).
Apart from presurgical training, tissue engineering also grows with AM technology, including research on blood vessels, complex vascular networks, and the tissue’s vascularization (Shafiee and Atala, 2016). However, using 3D bioprinting in clinical applications is still in its infancy.
Initial results are promising, but it will take more time to fabricate 3D bioprinted vascular conduits or valves, which can grow into the patient’s body. (Nguyen, 2017) In recent years, 4D printing (i.e. 3D printed objects with the ability to change its properties over time) has received more interest, with pioneer studies conducted by Spiegel et al. (2019) and Chua et al. (2020).
This development also requires improvements regarding bioprinting materials and 3D bioprinting technologies. According to Chua et al. (2020), 4D printing still lacks consensus regarding terminology, and probably will in the foreseeable future. Nevertheless, AM shows potential for further research and for more advanced applications in the future of cardiology.
1.2 Problem formulation
Relating AM’s adoption in cardiology to the technology adoption life cycle, which became famous through Rogers’ book “Diffusion of Innovations” in 1962, the chasm from early adopters to the early majority, shown in figure 1 on next page, is yet to overcome (Rogers, 2003).
Though AM has the potential to serve as a useful tool for clinical practice and medical education, some challenges need to be addressed. Luo et al. (2017) list several challenges of implementing AM in cardiovascular medicine like a lack of referencing guidelines for
documentation or medical device), as well as the cost and time efficiency of using 3D printing (Luo et al., 2017).
Figure 1. Technology adoption life cycle. (Mohammed, 2017)
Apart from cardiology, Ahsan and Rahman (2019) investigated the challenges of implementing 3D printing in medical device supply chains and found that technological, strategic, organizational, and operational factors affect the implementation process. Thus, there is a variety of factors that interplay, and which influence the success of AM adaption in health care.
Grol and Grimshaw (2003) point to the many available strategies to change health care practice, saying that ”the difficulty of introducing innovation in patients’ care makes a critical platform on which to identify the most effective and efficient approaches to achieve change in practice.”
Moreover, there are different levels on which these factors can influence implementation.
Scholl et al. (2018) provide a scoping review about organizational- and system-level characteristics believed to influence shared decision-making implementation. They propose organizational-level characteristics like leadership, culture, resources, priorities, as well as teams and workflows and system-level characteristics like policies, clinical guidelines, incentives, culture, education, and licensing. Concluded from Scholl et al.'s (2018) scoping review, organizational- and system-level characteristics should be carefully considered, and implementation barriers and facilitators should be addressed and could be guided by implementation theory. Although their review was about the implementation of shared decision making in health care, one can assume that similar issues like the health care system’s
complexity or changing individuals’ behavior, apply when attempting to implement AM in health care.
As of 2018, the global 3D printing health care market was valued at $973 million and is projected to be valued at $3,692 million by 2026, thus having a compound annual growth rate of 18,2% from 2019 to 2026 (Kunsel and Sumant, 2019). Dental implants have the highest market share (34%) by revenues on the health care 3D printing market, followed by general implants (19%), prosthetics (13%), hearing aids (9%), surgical guides (8%), medical components (6%) and some other areas with less than 5% (Barrow, 2018). Considering cardiovascular medicine, AM is mainly used for “device innovation, teaching tools, procedural planning, and functional flow models” (Haleem et al., 2018, p.434). These applications are barely included in the market overview listed above, although millions of people require help regarding cardiovascular medicine each year (Haleem et al., 2018).
The field of cardiology has a negligible market share in 3D printing and has just recently got some attention from researchers. However, Sweden is far behind the US, UK, and Singapore in terms of published papers (Ramola et al., 2019). Although AM offers many opportunities and benefits, it is yet to be applied by the mainstream health care market, especially in cardiology.
The projected growth in the coming years indicates the potential for the technology, although it is still facing challenges. Such challenges regard both the technology itself, including its cost efficiency and usability, but more importantly, the adoption process, including barriers and facilitators of implementing the technology. All in all, 3D printing implementation in cardiology is both an interesting and relevant field to investigate further.
AM technology is progressing quickly and is doing so in a fast-changing environment.
Therefore, the thesis aims to create an overview of the latest 3D printing challenges in health care, more specifically, in the field of cardiology in Sweden. By gathering the latest information through interviews, the authors strive for detecting influencing factors that will build a base for further research on strategies for implementing AM in health care. Furthermore, this study will connect innovations in health care with implementation science literature to create a determinant framework supporting managemental decisions of AM implementations in prospective projects.
1.4 Research question formulation
Given the problem formulation and the objectives of this work, the following research questions will be addressed:
1) What are the barriers and facilitators of implementing Additive Manufacturing in the
This study will be delimited to investigate barriers and facilitators of AM implementation within the health care sector in Sweden, more specifically within the clinical field of cardiology.
Given the scope of this thesis and the time constraints, two case studies will be done, one at a hospital yet to implement AM in their cardiology division, and one at a hospital that has already implemented AM. Moreover, identified barriers and facilitators will not be valued in relation to one another nor will an explanation of any causal mechanisms between them be provided.
Chapter 4.7 gives a more detailed overview on the thesis’ limitations, specifically in relation to the chosen methodology.
1.6 Thesis structure
The thesis consists of seven chapters that are briefly described below:
Chapter 1 – Introduction: This chapter briefly describes AM in health care, use cases, and implementation aspects. Moreover, this chapter accounts for the objectives and research questions guiding this work.
Chapter 2 – Literature review: This chapter gives an overview of aspects currently considered in literature in relation to barriers and facilitators of implementing AM in health care and cardiology in specific, as well as it describes gaps in current research.
Chapter 3 – Theory: This chapter describes the research field of implementation science and presents the existing frameworks for review and the rationale behind choosing them. The chapter ends with a comparison of the selected frameworks that accounts for similarities, and discrepancies between them, advantages, and disadvantages as well as potential limitations of the reviewed frameworks.
Chapter 4 – Methods: This chapter describes the methodological approach taken for this study.
It explains why the respective design of study, data collection method, and data analysis method were chosen to conduct qualitative, exploratory, and comparative case studies which investigate barriers and facilitators in empiricism. Further, ethical considerations, trustworthiness, and the study’s limitations are accounted for.
Chapter 5 – Results: This chapter accounts for the findings from the interviews conducted within the two case studies, as well as the result and analysis of the findings.
Chapter 6 – Discussion: This chapter contextualizes the research questions to the respective literature background and reasons about the results’ meaning, importance, and relevance as well as its contribution to existing literature.
Chapter 7 – Conclusions: This chapter includes the conclusions of this work, which are based on the literature study, the results from the case studies, as well as the previous reasoning in the analysis and discussion sections. Moreover, the study’s limitations and implications are presented.
2 Literature review
First, this chapter gives an overview of aspects currently considered in literature in relation to barriers and facilitators of implementing AM in health care and cardiology in specific. Second, gaps in current research are shown, arguing for the relevance of this thesis work.
2.1 Implementation of Additive Manufacturing in health care
The literature review is separated into five parts. The first four sections: Economic considerations; Standardization and legal considerations; Application cases; and Technological considerations describe barriers and facilitators concerning the implementation of AM in general, followed by the section cardiology, which reviews these aspects in cardiac applications in specific. A clear cut between the five sections cannot be done as some of these issues are overlapping.
2.1.1 Economic considerations
Economic considerations comprise aspects of AM implementation that are connected to the AM market itself, financial and strategic aspects for single players (e.g. a hospital) on the AM market, as well as long-term aspects (e.g. market positioning) resulting of AM implementation.
In health care, there is a demand for customized products (Weller et al., 2015), providing even better medical care. However, conservative manufacturing methods are often energy-intensive and cause material waste, as well as processes, are complex and, therefore, expensive (Kalaskar, 2017). Thus, customized medical products are rarely manufactured nowadays, but AM might offer the opportunity to produce complex parts cost-effectively at a low volume (Rath and Sankar, 2017).
Weller et al. (2015) discuss in their paper general economic implications of 3D printing, independent of any specific industry, at a firm as well as industry level. They expected the AM market to reach a mainstream level between 2016 and 2020, including more experience (e.g.
skilled labor) and support in the market environment (Weller et al., 2015). A later review by Sireesha et al. (2018) shows that the early majority’s adaption is expected to start between the earliest 2022 and latest 2030 depending on different adaption rate scenarios. These examples show the difficulties in predicting the market development and adaptation of AM. Especially, the health care market is difficult to assess due to stronger regulations and quality standards, which are shown in more detail in section 2.1.2. Ramola et al. (2019) investigated about 70 papers (published between 2007 and mid-2018) in their literature review about AM in health care. They found that AM use cases increased, albeit it has yet to explore its full potential. The increasing interest in research and usage in empiricism emphasizes the trend of adopting AM
Regarding the implementation’s cost aspects, different opinions are existing in current research.
Weller et al. (2015) assess an investment decision as highly strategic due to high initial costs.
Other authors agree that investment in AM is connected to high expenses, and there might be a lack of reimbursement (Bryant, 2018; Ramola et al., 2019). Depending on how a hospital is funded and hence, its financial possibilities, the decision process of implementing AM might look different. Vervueren (2017) writes about the Materialise World Summit; a conference organized every second year to exchange information. In 2017, different stakeholders of the AM industry took part in the panel discussion, also underlining the issues of regulation and cost of 3D printing applications. Dr. Morris – a physician from the Mayo Clinic, USA – expresses that costs are “an issue, especially when people can’t see the big picture” (Vervueren, 2017).
He adds that there is evidence that AM reduces costs and operating time, but more studies are needed to convince the majority and to overcome outdated perceptions (Vervueren, 2017).
Further, clinical evidence is missing, which shows how AM affects patient outcomes (Bryant, 2018; Vervueren, 2017). However, Huang et al. (2013) who wrote a literature review on the societal impact of AM from a technical perspective, emphasize in their paper that there is evidence that customized health care products “improve population health and quality of life”
(Huang et al., 2013, p.1191).
Moreover, Ramola et al. (2019) write in their paper that AM is still costly compared to traditional approaches. Though, they also mention that only a few researchers have considered cost-related issues showing whether 3D printing is cost-effective or not, compared to traditional approaches (Ramola et al., 2019). At least, in comparison to traditional methods, there is no need for tools or molds, for instance, since 3D models can be printed directly from digital data (Tofail et al., 2018; Weller et al., 2015). Another issue related to investment costs is skilled labor and training, which are required to apply AM successfully (Weller et al., 2015). The importance of interdisciplinary knowledge rises since practitioners need to assess the technology’s capabilities, and technicians need to understand the field of application like health care issues while applying AM. Thus, the question of whether AM is costlier than other technologies cannot be answered. Scientific evidence and cost-related considerations concerning AM’s implementation and usage seem to be missing in a broader range in current research.
Huang et al. (2013) argue that AM will shorten the design cycle process and the delivery lead time. In other words, product design iterations can be done without a high-cost penalty leading to faster prototyping in general (Bryant, 2018; Huang et al., 2013; Shafiee and Atala, 2016;
Tofail et al., 2018; Weller et al., 2015). Thus, the marginal costs of creating innovations and high product variety are comparatively low. Tofail et al. (2018) complement saying AM provides the opportunity for “excellent scalability.” Though, AM has high marginal costs considering production expenses like the use of raw material and energy consumption (Weller et al., 2015). Further, Rath and Sankar (2017) argue that producing and assembling 3D models is still a time-consuming process.
Even though AM might be a time-consuming process, the delivery time is lower compared to traditional methods since AM enables local production (Weller et al., 2015). Thus, several authors conclude that complex designs and anatomical models are realizable in a shorter time
(Bryant, 2018; Huang et al., 2013; Bagaria and Chaudhary, 2017; Weller et al., 2015). Supply chains are less complex, and hence, they might also be less costly. Huang et al. (2013) also find evidence for increased efficiency through simplified supply chains in their paper. However, Özceylan et al. (2018) mention that there is still a gap in analyzing the effects of using AM in supply chains and a general lack of case studies. Therefore, the authors investigate two supply chain networks, one traditional and one, including 3D printing in the health care industry. Their results also show that 3D printing supply chain networks outperform traditional methods considering lead time, cost reduction, and the number of customers served successfully.
(Özceylan et al., 2018) Though their research also has limitations and in general, there is still a lack of case studies considering AM supply chains in health care.
Since economic, environmental, and social sustainability are interconnected, some related aspects are mentioned in existing research. Attaran (2017), for instance, lists in his paper, among others, AM implementation challenges, AM’s impacts on supply chains, and discusses advantages and disadvantages compared to traditional manufacturing. Mentioned sustainability-related aspects are reduced transportation costs and emissions, reduced waste, and reduced inventory since AM is mainly used as on-demand manufacturing (Attaran, 2017).
Other authors also underline the reduction of scrap (Tofail et al., 2018; Weller et al., 2015) and the reduction of using raw material and decreased energy consumption, as well as AM, pollute less “terrestrial, aquatic and atmospheric systems” (Huang et al. 2013, pp.1200-1201).
However, issues about the disposal of used models as well as safety aspects while using AM may arise, thus hampering the implementation process (Bryant, 2018). These sustainability aspects as well as how a hospital positions itself at the market contribute to a hospital’s image.
Being among the first movers of adapting AM and acting sustainably creates another image than being part of the laggards (see figure 1) and not considering sustainability aspects at all.
To summarize: three main factors were identified in economic considerations. There are issues considering the market characteristics like how far the AM market has progressed and which opportunities are existing, issues considering the implementation and running costs for single adopters and issues considering an adopter’s image and, thus, positioning on the market. Table 1 on the following page lists the above-mentioned main identified aspects which influence AM implementation (positively as well as negatively), currently considered in AM literature.
Table 1. Economic considerations
Factor Sub-factor Aspect
Technology adoption Estimation issues, different adoption rate scenarios
Number of manufacturers Existing AM network
Demand Customizable products
General interest in technology
Strategic decision Reimbursement
Training of labor, interdisciplinary knowledge Clinical evidence, cost-related evidence in research Existing alternatives
Resource consumption (e.g. material, energy, time) Simplification of innovations
Supply chain costs
Place of production
Complexity of supply chain steps Efficiency, existing case studies
Usage of resources Protection of users Waste considerations
Competitors Customers Network
2.1.2 Standardization and legal considerations
Standardization and legal considerations comprise what standards and regulations are existing for AM, how easily applicable they are, and in which way they influence the implementation process positively or negatively.
Considering AM standards, the most popular one is ISO:ASTM 52900:2015, which sets standardized terminology and definitions, updated with a second edition in 2018 (ISO/ASTM 52900 (en), 2018). The two standard development organizations ISO and ASTM International also developed the Additive Manufacturing Standards Structure, mainly covering the fields of materials, processes, and equipment and the treatment of finished parts (AMFG, 2018). These internationally set standards seek to reach unified communication about AM and a shared understanding of the technology heading for a simplified adoption among users. Though, it is
still criticized that AM adoption “is hampered by challenges concerning part quality, consistency, and certification” (AMFG, 2018, para 5).
Weller et al. (2015) also claim that quality standards and guidelines are lacking, and intellectual property rights and warranty concerning AM’s usage are limited. Furthermore, a report published by Deloitte still shows a lack of IT standards concerning AM in 2019 (Proff and Staffen, 2019). Ramola et al. (2019) also mention in their literature review different applications for which standards are lacking. First, they detect missing regulations concerning the handling of printers and printed medicine. Second, standards for pre-surgical training models are lacking as different printing methods generate slightly different models. Thirdly, they mention variations in 3D printed devices requiring a more standardized approach. (Ramola et al., 2019) The second and third issues show the problem of reproducibility. Nevertheless, Bryant (2018) emphasizes that standardization efforts are made in North America, for instance, where a 3D printing group is working on a standardized application of AM through analyzing different use cases.
Another aspect concerning the implementation of AM is legal and regulatory issues. In North America, for instance, there is the Food and Drug Administration (FDA), bringing up regulations. 3D printed medical devices are classified into one of three classes, which require respectively different tests and trials. Moreover, there are different laws in different American regions, which makes the approval path quite complicated. (Ashish et al., 2019)
At the Materialise World Summit, the discussion covered regulations of AM in health care as well. Some participants agreed on the application case (e.g. implants versus medical devices) of AM technology to be a driving factor for regulations. Further, a physician from the Mayo Clinic USA emphasized the importance of not letting regulations deciding on how AM in health care evolves. He said standards set by the AM community should guide regulatory bodies to adapt them since people writing the regulations are mainly non-professionals in the field.
In Sweden, the Medical Products Agency is mainly responsible for controlling and monitoring drugs and other products connected to drugs, medical engineering products, cosmetic products, tattoo colors, and other goods with the intended use for cosmetic and hygienic purposes. In the annual report of 2018, three new fields have been identified for which new regulations are necessary, one of them is AM. (Läkemedelsverket, 2019) In the annual report of 2019, AM is not mentioned anymore, and there is no information about AM regulations on the Agency’s website. Nonetheless, though Sweden does not have country-specific regulations in Sweden and the entire European Union, the European Medical Device Regulation (MDR) applied as the main regulation to be considered in connection to AM implementation.
The MDR categorizes medical devices into four different classes (I, IIa, IIb, and III) depending on the purpose of usage, the time of application, and the risks included. Annex VIII of the MDR comprises 22 rules in total, defining the classification system. Most non-invasive devices, for
be implemented in May 2020 but was postponed by one year because of the ongoing coronavirus crisis (European Commission, 2020).
To summarize: the ISO:ASTM standard might help to overcome issues related to unified communication and definitions, but clarifications are still needed concerning quality and IT standards used for AM. Further, regulations can be complex, and the MDR, which is applied in Sweden as part of the European Union, is a new regulation; hence, it might be difficult to identify related issues. Table 2 lists the identified standardization and legal aspects of AM implementation in more detail.
Table 2. Standardization and legal consideration.
Factor Sub-Factor Aspect
ISO /ASTM standards
International and unified understanding
Processes and equipment Treatment of finished parts
Intellectual property rights and warranty
IT standards Design tools
Legal and regulatory aspects
Approval path Regulation creation Classification of products
Purpose of usage Time of usage Invasiveness MDR regulation
2.1.3 Application cases and their applicability
Application cases and their applicability comprise current possibilities as well as perspectives of AM applications. Some applications are not realizable in a consumable and widespread way yet. In some clinical areas, application cases can be limited unless an organization applying AM also pursue research or educational purposes. Depending on the technology user’s goals, the variety of application cases can be supporting or limiting the implementation decision.
As already mentioned in the introduction, several applications are facilitating the way of how physicians treat their patients. Shafiee and Atala (2016) discuss in their review current and
future applications of 3D bioprinting, as well as some initial applications of 3D printing. Within the field of bioprinting, they identify applications like drug delivery, personalized medicine, and living organs. Further, they mention maxillofacial applications where 3D printing can be used for ears, eyes, or jaws with methods of tissue engineering, prostheses, surgical planning models, and implants like reconstructed mandibular structures. Additionally, they list applications for internal organs like livers, hearts, or kidneys. Considering kidneys, cells show viability after 3D printing steps leading to further printing opportunities. (Shafiee and Atala, 2016) Although there are the perspective and first promising results of producing living organs, 3D printing is mainly used for non-living 3D models in this field. 3D models support pre- surgical planning, consultation, simulation, and educational purposes. Further, the authors mention cancer applications, since printing with hydrogels, for instance, helps to model cancer cells resulting in advanced cancer research. Moreover, hardware and instrument production is mentioned in relation to cancer applications. (Shafiee and Atala, 2016)
Bagaria and Chaudhary (2017) investigate in their study the experiences of five surgeons during their first 50 orthopedic surgeries with the support of 3D printed biomodels in surgical planning.
The surgeons reported that they got additional valuable information compared to conventional methods like an enhanced understanding of complex pathoanatomies (Bagaria and Chaudhary, 2017). To quote:
“It [3D models] was useful in preoperative planning, rehearsing the operation, surgical simulation, intraoperative referencing, surgical navigation, preoperative implant selection, and inventory management.” (Bagaria and Chaudhary, 2017, p. 2501)
That shows how AM models might contribute to better educated and more confident doctors who even can develop a slight routine of rarer cases. Thus, the quality of medical treatments might increase. Nevertheless, the doctors participating in the study also claim that using 3D models did not change their surgical plan significantly (Bagaria and Chaudhary, 2017). Thus, depending on an individual’s experience, it might be redundant to produce 3D models.
Tofail et al. (2018) mention in their report, key applications of AM in different industries. In health care, they conclude pre-surgical planning, customizable orthopedic implants, and prosthetics, printed body parts for medical education and training as well as biodegradable living tissues for medicinal product testing as main applications (Tofail et al., 2018). Jamróz et al. (2018) also identify models for surgical planning and education as main application cases since models provide information on individual organ features resulting in less patent loss.
From a manufacturer’s point of view, AM can be applied for rapid prototyping of devices, tools as well as for anatomy-based models for testing and training according to an engineering director from Medtronic’s cardiac-related business (Bryant, 2018).
Even though there are lots of cases that are realizable with the help of AM, there are cases that are not applicable today. Hong et al. (2018), for instance, mention the issues of vascularization, which is not resolved yet. However, no other technology resolves this issue, either. Ramola et
A more recent application scenario of AM in health care is during periods of shortages. In 2020, the COVID-19 pandemic results in a worldwide shortage of different medical products. AM supports the production of face shields, nasopharyngeal swabs, and door opening hooks, for instance (Heiberg, 2020b). Furthermore, the company Copper3D developed a mask and shared the printing file for free. Another notable example of AM during the crisis regards a 3D printed valve for respiratory intensive care in hospitals in Italy (Carlota, 2020). These applications are specifically useful for COVID-19 issues. Though, some parts like the face shield or the door opening hook might also be useful for reducing other diseases that might come up in the future.
These examples show the efficiency of AM technology in emergencies since people – independent of the location – just need a digital file and a 3D printer to react quickly to shortages (Carlota, 2020).
Summarized, the possibilities of AM applications compared to conventional methods probably outweigh, and therefore, most applications work as facilitators to AM implementation.
Concerning organs, the application cases, which are currently practicable, are limited, and depending on the user, it might not be sufficient for the technology’s implementation.
Depending on the users and their level of experience, education, and research focus, the necessity of implementing AM might also be questionable. In table 3, the current main applications of AM in health care are summarized. The summary is based on applications mainly mentioned in current literature and is not intended to be exhaustive.
Table 3. Application cases
Factor Sub-Factor Aspect
Drug screening and delivery Personalized medicine Living organs
Cell engineering/ patterning
Maxillofacial Orthopedic Implants
Cancer Product testing
Surgical training and medical education Inventory planning
Medical devices and instruments
Rapid Prototyping Tool testing Adaption process
2.1.4 Technological considerations
Technological considerations comprise technical (hard- and software) and material issues connected to the main process steps. AM technology consists of three main steps: pre-printing, printing, and post-printing (Shafiee and Atala, 2016). The pre-printing step mainly includes imaging, digital editing, and design. Most commonly computed tomography (CT) and magnetic resonance imaging (MRI) are employed (Hong et al., 2018; Shafiee and Atala, 2016), as well as sometimes, 3D scanners or ultrasound imaging are used to generate digital anatomic 3D models (Hong et al., 2018). Then, these data are converted into standard AM files readable by 3D printers, such as STL files (Hong et al., 2018; Huang et al., 2013; Shafiee and Atala, 2016).
In the case of designing medicinal tools, for instance, traditional CAD software can be applied to create 3D models. During the printing phase, material selection is important, dependent on the printer type and the intended use (Shafiee and Atala, 2016). During the post-printing process, the maturation time for biological functionality with regards to bioprinting might be considerable (Shafiee and Atala, 2016).
Regarding conventional 3D printing for medicinal tools or instruments, for instance, post- processing issues like polishing of the printed 3D object might be important. Some aspects connected to AM’s technology were already mentioned in the economic considerations and application cases. Therefore, some aspects are mentioned rather superficially and can be found in more detail in one of the two previously mentioned sections.
In the pre-printing phase, existing technology (e.g. CT and MRI) can be employed, resulting in a low effort to create digital data for 3D printing. Tofail et al. (2018, p.23) list the “direct translation of design to component” as a benefit of AM. Therefore, as mentioned in the economic aspects, adaptions and improvements to existing files can be made without much effort (Huang, 2017; Shafiee and Atala, 2016; Weller et al., 2015). Though, Chua et al. (2020) call for new AM systems, specifically with regards to bioprinters. According to them, there is a need for combining AM with other technologies like electrospinning to create better results in the field of bioprinting (Chua et al., 2020). Weller et al. (2015) also identify a lack of proper design tools and guidelines, as do Jamróz et al. (2018). The latter claim a need for better visualization and improved rendering of digital images in order to exploit AM technology’s entire potential. Ashish et al. (2019, p.2) also mention drawbacks of the STL file format, such as the non-inclusion of “type and properties of the material, color, surface texture, units, or any other feature details,” for what reason other file formats (AMD or 3MF) were developed.
Hence, the technology’s user has to figure out which format is the best to use for his or her specific application, which might result in confusion and increased effort.
Regarding the actual printing process, most materials applied today are limited to the manufacturing of static objects (Spiegel et al., 2019). Concerning 3D bioprinting, there are material issues like limited availability of bioink, which is ink consisting of living cells (Chua et al., 2020; Zadpoor and Malda, 2017). Bioink has numerous requirements, and it is hard to maintain the ink’s condition during AM processes, which include, amongst others, changes in
(Hong et al., 2018; Spiegel et al., 2019). Not just biomaterials are criticized in literature but also the diversity and flexibility of 3D printing materials in general (Kalaskar, 2017).
Kalaskar (2017) thinks the resolution still is lacking in many cases. Chua et al. (2020) also assess the resolution of 3D printed objects as a challenge, especially if it comes to nanoscale applications. Further, material consistency and mechanical characteristics – like for hydrogels used in bioprinting – are issues to investigate further (Chua et al., 2020) as are the application of multiple materials within one print (Ashish et al., 2019; Chua et al., 2020). Zadpoor and Malda (2017) agree that there are lots of requirements for bioink which have to be considered.
Even though AM offers the opportunity to design almost any internal shape or microarchitecture of a 3D object, it is still unclear which microstructure leads to the best performance of a certain biomaterial (Zadpoor and Malda, 2017). Though, other authors agree on AM’s benefit of manufacturing complex internal structures (Bryant, 2018; Rath and Sankar, 2017) and thus, the production of lightweight and flexible models (Tofail et al., 2018). Several authors estimate the overall product development, including prototyping and production time to be reduced through AM (Bryant, 2018; Tofail et al., 2018). However, other authors mention the printing process in specific to be too time-consuming and not optimized yet (Kalaskar, 2017;
Ramola et al., 2019).
With regards to the post-printing phase, Tofail et al. (2018) see the potential of AM to be a zero-waste production method. Through its direct manufacturing approach, additional post- processing can be minimized. Further, AM provides good scalability in terms of changing the product sizes or customization degree without significantly increasing cost (Tofail et al., 2018).
However, there should be safety considerations whether materials in powder forms can be harmful to people if inhaled in post-printing processes or during the printing process itself (Bryant, 2018).
To summarize: there are issues worth to be considered during each step of AM manufacturing.
Concerning an implementation decision, it is necessary to know which printer, material, and software are needed in specific to buy the most suitable equipment. Further, existing infrastructure might facilitate the implementation and only needs to be complemented. Since its application in health care is relatively young, there are still time-consuming issues within the processing steps for which (initially) more resources are needed. Table 4 on the following page shows the aspects within the three phases of manufacturing a 3D print, which might hinder or support the successful use of AM in health care.
Table 4. Technological considerations
Factor Sub-Factor Aspect
Conversion (file format)
Design Tools and their features
Behavior during printing process Research
Resolution Consistency Internal structures Time
Though most of the above-mentioned barriers and facilitators also apply in the clinical area of cardiology, this section gives an overview of some specialties regarding cardiac applications.
One of the most important papers summarizing AM in cardiology is probably the review by Haleem et al. (2018), which describes current applications, including the technology’s limitations and the technology’s development in the field of cardiology. Amongst others, they investigated the development of published articles concerning AM in cardiology, showing an increasing interest in the field from 2014 (Haleem et al., 2018).
In respect to economic considerations, the authors mention additional time and cost efforts compared to general patient treatments. However, this additional effort may lead to reduced operating times for complex surgeries. Further, the development time of a 3D printed heart is perceived as rather short. However, the authors also mention the need for interdisciplinary knowledge and that AM will enhance collaboration between different clinical areas like
“cardiologists, radiologists, cardiac catheterization specialists, and surgeons.” (Haleem et al., 2018, p.439)
On the contrary, another review by Valverde (2017) assesses the preparation time of 3D models as critical. According to him, only counting the reduction of surgery time against time invested in 3D printed models is difficult since other benefits like patient safety are hard to assess economically (Valverde, 2017). Valverde (2017) agrees that multidisciplinary teams are necessary for a successful application of AM, especially incorporating engineering skills.
with 3D expertise” (Anwar et al., 2018, p.309) as the primary cost aspects to be considered.
Martelli et al. (2016), who wrote a systematic review on articles concerning 3D printing in (cardiac) surgeries, concluded as well that additional costs and the time needed to prepare 3D prints are the most limiting factors for routine use.
Regarding legal issues, the negative consequences of modeling 3D prints vary depending on the use of the respective model. Employing a flawed model used in teaching compared to one used for detailed surgical planning has different implications (Vukicevic et al., 2017).
Therefore, Rath and Sankar (2017) mention that it is still in debate who is finally responsible for errors in a customized model. Since the technology’s different manufacturing steps require multidisciplinary teams, there are several people involved who might be responsible. Further, outsourcing processes might be complicated due to local and international data protection regulations (Rath and Sankar, 2017).
As mentioned in the thesis’ introduction, there are diverse application opportunities for AM in cardiology. 3D cardiac applications are “proliferating” and exposed to intensive research, according to Haleem et al. (2018). Mainly, AM in cardiology is supporting surgical planning or preoperative evaluation (Anwar et al., 2018; Haleem et al., 2018; Proto3000, 2018; Valverde, 2017; Vukicevic et al., 2017), training and teaching (Haleem et al., 2018; Jamróz et al., 2018;
Valverde, 2017; Vukicevic et al., 2017) and medical consultations like explaining diseases to patients or colleagues involved in surgeries (Anwar et al., 2018; Haleem et al., 2018; Valverde, 2017). Further, but less popular fields in which AM is involved, are device innovations or flow models (Haleem et al., 2018; Vukicevic et al., 2017). Different cases are benefiting from AM like congenital heart diseases used in pediatric cardiology, complex anatomical concepts or phenomena like double-outlet right ventricle, aortic disease treatments, or cardiac tumors (Jamróz et al., 2018; Valverde, 2017). Another case is hemodynamic considerations describing the blood flow influenced by factors like vessel size, for instance (Haleem et al., 2018).
Although the previously mentioned authors mention the benefits of applying AM in cardiology in education, amongst others, a paper by Huang (2017) contests the educational benefits of 3D printed cardiac models compared to traditional ones. Showing results of recent empirical studies on students’ outcomes of being taught by 3D models, he concludes that “3D printing technology did not show superiority in knowledge acquisition when compared to other groups” (Huang, 2017, pp.670-671). Therefore, there is still debate on whether AM provides a benefit in teaching methods (Huang, 2017). Though, Valverde (2017) shows that classes using 3D models for teaching were positively perceived by students leading to higher satisfaction and retention rates.
The technology’s process steps are the same as described previously. After the medical image acquisition, the segmentation process is more important. Segmentation means the delineation of cardiovascular structures, thus separating them from irrelevant noncardiac parts like bones or lungs, for instance (Valverde, 2017). This process may take time, up to 12 hours, depending on the complexity of the model and software employed (Vukicevic et al., 2017). Thus, there is still potential for improving this step and especially, the software. In cardiology, often, multi- colored models are important to separate different parts of interest for showing cardiac tumors, for instance (Valverde, 2017). Thus, there might be higher requirements for printers and materials compared to other clinical fields.
Summarized, the issues occurring in cardiology overlap mostly with ones occurring in health care in general. Considering the application cases, cardiology is limited to simulation models for consultation, education, and planning in practice currently, but there are promising research fields as well. Since the heart is a sensible part of the human body, legal issues about responsibility in cases of failure are more emphasized. Table 5 summarizes the main aspects which specifically apply in the field of cardiology. Though, most aspects of tables 1-4 also apply in the field of cardiology.
Table 5. Cardiac considerations
Factor Sub-Factor Aspect
Preparation time Surgical time Printer Software Personnel Material
Data protection Guideline
Pre-surgical planning Training and education Medical consultation
Congenital heart disease
Complex anatomical phenomena Aortic diseases
Technology Software Segmentation
Material Multi-color print
2.2 Gaps in research
Haleem et al.'s (2018) review showed a rising interest in AM in the field of cardiology from 2016 until the mid of 2018 when their paper was published. In this paper, the researchers also show area wise research efforts on AM in cardiology. The research field Medicine contributes with 47% of the published articles the largest amount. Other fields mentioned are Engineering (11%); Chemical Engineering (8%); Materials Science (8%); Biochemistry, Genetics and
sciences, including a managemental point of view on AM in cardiology. Although the articles reviewed above comprise economic, legal, or technological considerations, the articles’ main interest remains from another point of view, like a specific technological issue to be resolved or a medical problem to be simplified. Thus, the literature on implementation factors or strategies from a managemental point of view is limited. Further, incorporating aspects of implementation science, which will be described in more detail in the next chapter, is not considered yet in AM implementation.
With a lack of social sciences, there is also a lack of market considerations. Current literature barely looks at how the market around AM innovations and applications in cardiology is built.
It is not clear which relation hospitals applying AM have to their suppliers. Are these one-time buyer-supplier relationships, or do they learn and develop commonly in the long run? Further, literature does not describe how people using AM in cardiology educate themselves. Are they part of existing educational networks, are there local opportunities or external opportunities offered by suppliers, for instance?
Moreover, are national or international societies, like the “Svenska Kardiologföreningen” (The Swedish society of cardiology), involved in knowledge exchange, or are they generally supporting AM implementation in cardiology? These questions are examples of open questions not addressed yet in literature. A good network and easy access to education, for instance, could work as a facilitator to AM implementation. Through investigating barriers and facilitators of implementing AM in cardiology, some of these raised questions might be answered.
Further, organizational issues, including the organization’s structure, relations between leaders and clinicians, or the localization of 3D printing facilities, are not considered yet. Moreover, individual characteristics of leaders, clinicians, researchers, or external people might be important concerning AM implementation.
Comparing table 3 (AM application cases) with table 5 (cardiac considerations) shows that cardiology mainly deals with applications not implantable in the human body but with heart models used for planning, consultation, or education. Since AM is developing fast, it is interesting to find out whether other applications are planned or already applied in cardiology.
Nevertheless, this thesis work cannot answer all of the above-risen questions. It instead aims to create a broad overview of aspects that either hinder or support the implementation of AM in health care, specifically in cardiology. Currently, there is a qualitative study published that assesses the clinical value of 3D printed heart models for pre-surgical planning, education, and communication through interviews with different health professionals (Lau et al., 2018).
However, as AM in cardiology has received relatively little interest in literature (especially in Sweden), which is based on lacking publications, it will be interesting to explore AM in cardiology in Sweden.
In short, current gaps in research are summarized as follows:
• Low involvement of social sciences incl. management sciences as well as implementation science concerning AM in cardiology