Consequences of the adoption of
blockchain technology for
intermediaries in the Dutch
Healthcare industry
Master thesis presented by
Marta Vargas Sánchez Student ID: 11376732
Supervisor: Dhr. Prof. Dr. Peter van Baalen 23 - June - 2017
Statement of Originality
This document is written by Student Marta Vargas Sanchez who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
© Copyright by Marta Vargas Sánchez 2017 All Rights Reserved
ACKNOWLEDGMENTS
I want to acknowledge and thank the people who trusted and supported me in making this thesis. Thanks Madrid for the support, Berlin for the encouragement, E.2.22 for all the resolved doubts, and home for all the important things. You have been the pillar where I could rest and take a breath.
Possunt quia posse videntur.
P·VERGILIVS·MARO
ABSTRACT
— The aim of this paper is to analyze the impact that the adoption of blockchain technologies will have on intermediaries in the Dutch healthcare sector. The focus of the paper is to study if blockchain technologies can disrupt the Dutch healthcare industry by playing a disintermediating role on health insurance companies and therefore, accelerate new business models as collaborations to increase the trust between institutions and patients in a traditional inwards industry. These three propositions are based on the underlying theoretical framework of the theories of disintermediation and disruptive innovations. After a research strategy in the form of a case study and an iterative coding process, the results of these conceptualized propositions suggest that the use of blockchain technologies will not replace current health insurance companies – as intermediaries, but will support them by accelerating the creation of new forms of collaborations, leading to an increase in transparency and trust. This paper contributes to the literature by applying the concepts of blockchain, disruptive innovations and disintermediation to the Dutch healthcare sector with a focus on health insurance companies.
Keywords— Intermediaries, Healthcare, Disruption, Blockchain, Trust, Transparency,
TABLE OF CONTENTS ABSTRACT ... 4 1. INTRODUCTION ... 6 2.BACKGROUND ... 8 2.1 Blockchain as a disruptive technology ... 8 2.2 Blockchain in the insurance industry ... 10 2.3 Healthcare system in The Netherlands ... 12 2.4 Blockchain in Healthcare ... 15 3. THEORETICAL FRAMEWORK ... 19 3.1 Theory of disruptive innovations ... 19 3.1.1 Disruptive innovations applied to health care ... 21 3.2 Theory of disintermediation ... 23 4. METHODOLOGY ... 25 4.1 Philosophical assumption ... 25 4.2 Research strategy: Case study ... 25 4.2.1 Research framework ... 26 4.2.2 Access to the field ... 28 4.2.3 Participant selection procedure ... 29 4.3 Method ... 33 4.3.1 Semi-structured interviews ... 33 4.3.2 (Annual) reports, emails and brochures ... 34 4.4 Ethical considerations ... 35 4.5 Coding process ... 36 5. ANALYSIS ... 38 5.1 From coding to coherent narrative ... 38 5.2 Findings ... 39 5.2.1 Uses cases ... 39 5.2.2 Propositions ... 40 6. DISCUSSION ... 52 7. CONCLUSION ... 56 7.1 Theoretical contributions ... 56 7.2 Limitations ... 57 7.3 Practical implications ... 58 7.4 Recommendations for further research ... 59 REFERENCES ... 60 APPENDICES ... 64 Appendix A: Information about the participants ... 64 Appendix B: General protocol for interviews ... 67 Appendix C: Interview Transcripts ... 68 Appendix D: Open Coding of Interviews & Reports ... 69 Appendix E: Axial coding – Interviews & Reports ... 71 Appendix F: Definitions of code categories ... 72 Appendix G: Narrative Quotes ... 73
1. INTRODUCTION
It was 2008 when Estonia brought to life the Electronic Health Registry, a whole new e-service that added value to its people by combining healthcare data from different sources, providing quick data access for both doctors or patients and rapidly obtaining examination results, while maintaining privacy throughout the entire process. This e-solution is based on a disruptive technology – Blockchain- that came to life already 9 years ago, to primarily support a new cryptocurrency, the Bitcoin, developed by an anonymous programmer or group of programmers, known under the pseudonym Satoshi Nakamoto. However, it is the technology behind it what is worth to study. Blockchain based itself in a publicly distributed system for recording and storing transaction records, without central authority to verify the process: transactions are open to public and privacy is secure by established peer-to-peer cryptographic techniques (Swan, 2015).
In the healthcare sector, the benefits behind the adoption of blockchain technologies have been attracting more and more attention from private and public institutions. In the current system, there is a disconnect between one healthcare provider and another, so whenever a patient wants to switch his healthcare provider or need to provide its own data for medical reasons, the patient’s health data must either be transferred manually or the patient will be forced to repeat analysis or even to undergo slow and excessively bureaucratic processes (Linn & Koo, 2014). There is an array of incompatibility back-end systems and disintegrated data trails limit a patients’ capacity to access their medical history (Clayton M Christensen, Bohmer, & Kenagy, 2000). With the adoption of blockchain technology, the whole process would not only gain in effectiveness but also transparency. The patient would seamlessly share permissions to access important information attached to its identity, and make advance promises of payment in return for new treatments (Hwang & Christensen, 2008).
Because of the scarcity of scientific discussion about the uses of blockchain technology in the Dutch healthcare industry, this dissertation aims to study – from
technologies will have in the Dutch healthcare industry. In pursuance of extending the existing understanding of the blockchain potential and its ramifications that affect the healthcare industry, the following research question is posed:
What are the consequences of the adoption of blockchain technology for intermediaries in the Dutch Healthcare
industry?
In addressing this question, a qualitative case study has been performed, in which 12 institutions were carefully chosen for participation based on their direct involvement in the Dutch healthcare industry. In-depth interviews were conducted with each participant, with the goal of comprehend their experiences – if any- with blockchain technology. The thesis will follow the following structure: in Section 2, the current state of blockchain will be outlined, to familiarize the reader with the technology and its uses in the insurance industry. The Dutch healthcare market will be explained, followed by the potential uses of blockchain in the healthcare sector. In Section 3, the theoretical foundation will be presented, where attention is given to the theory of disintermediation and the theory of disruptive innovations. In Section 4, the research methodology is explained – describing underlying philosophical assumptions, research strategy, methods, as well as access to the field and ethical considerations. The consequent analysis in Section 5, starts by describing the coding process carried, and then presents the results of the investigation in the form of a case study narrative. These findings are then discussed in Section 6, focusing particularly on confronting the findings with the theories exposed. The study is concluded in Section 7, with a discussion of the theoretical contributions, practical implications, limitations and suggestions for future research.
2.BACKGROUND
2.1 Blockchain as a disruptive technology
If you wanted to learn about ancient history in 1994 you would need to go and consult an Encyclopaedia, either at home or in the central library. This would have cost you some time and money either to pay the library fee or the book itself. Only ten years later the information flow changed dramatically. In 2004, you would only need a web-enabled device that would connect you to Wikipedia and give you instant access to the information, no matter where you were and for free. There was an evolution in the transfer of information, that switched from centralized content to an open world of creation and distribution. Decentralized content reduced the cost of content generators because there are millions of people collaborating in the creation information sharing. Access is no longer limited because of physical barriers. Thus, the generation of content is faster and available worldwide. However, right now, we are even reaching the next level of information transfer: Blockchain.
Blockchain has been defined in many ways, always related with Bitcoin, as its creation was primarily to support the existence of this cryptocurrency. It is the public ledger of all Bitcoin transactions that have ever been executed (Swan, 2015), and it can only be updated by consensus of a majority of the participants in the system. Once entered into a blockchain, information can never be modified or erased (Schatsky & Muraskin, 2015). First, a transaction involves two contracting parties exchanging digital assets, contracts or money. When it is accepted, it is added to a protected block with other transactions that have happened in the last 10 minutes and sent out to the entire network. Miners - members in the network with high computing power - compete to validate the transactions by making changes to one variable until the network accepts the solution. This is called “proof of work” and the first one to solve it receives a reward. The validated block of transactions is added to the chain in a chronological order, giving everyone the chance to prove who owns what at any time, while preventing the alteration of blocks that have already been validated, as they would not match anymore with their predecessor in
Figure 1. Blockchain: How it works Source: Deloitte University Press
Blockchain allows, and encourages, disintermediation and decentralization (Swan, 2015), the disappearance of intermediaries; the ones who made large investments on systems of records and the processes to support them. The procedure would look vastly different when the trader could directly contact the seller while the blockchain would take care of the transfer of the asset and the
registration of the new ownership. Hence, the technology that underpins bitcoin, has emerged as an object of intense interest in the financial services industry and beyond (Schatsky & Muraskin, 2015). Several driving key principles increase its potential: it’s reliable and available; transparent; unchallengeable; irrevocable and digital (Schatsky & Muraskin, 2015). The financial services industry aside, other industries are only just starting out on blockchain development. This paper wants to analyze the potential development and adoption of blockchain in the Dutch healthcare industry and its potential consequences for its market intermediaries. 2.2 Blockchain in the insurance industry
“When I saw what the fundamental principles of the blockchain provided, it was just patently obvious to me that it would make sense around reducing fraud
related instances of valuables”.
(Leanne Kemp, Chief Executive Officer, Everledger)
Over the past year insurance and technology have been intensively interrelated due to blockchain. It has become very usual to associate hashtags like #fintech and #insurtech in social media to #blockchain (Ramada-Sarasola, 2016). Despite the threat of disintermediation, blockchain may offer great potential for the insurance industry, because much of this transformation relies on data (Shelkovnikov, 2016). Certainly, it can help to overcome the competitive challenges many incumbents face, from improving customer engagement to strengthen companies in digital mature markets. Customer relationship can improve with blockchain, as it secures patient data and enhances transparency and tariffs/claims fairness.
Insurance industry is facing ever-tighter regulation due to the increasingly threat of fraud by individuals or organised crime (Shelkovnikov, 2016). Blockchain could also resolve this, through smart contracts.
Smart contracts are transactions occurring in the blockchain that go beyond simple currency transactions. Additionally, these contracts have more extensive
a better understanding, smart contracts are characterized by three attributes, autonomy and self-sufficient -- as once initiated, contracts run by themselves with no human intervention -- and decentralized nature that removes the need for any type of trust between parties; once the agreements of the contracts are settled, actions will be automatically triggered as they occur (Swan, 2015). By using smart contracts, clients and insurance companies could go through claims in a more effective, decentralized and safe manner. However, there are many considerations raised by smart contracts regarding the regulators’ position to distinguishes between these new code contracts and the more flexible human ones.
One of the characteristics of blockchain is its network verification and historical record, therefore, the once complex and drawn-out claims process could now be recorded on the blockchain once authenticity and uniqueness have been verified, avoiding claims duplication or fraud. Each trust agent in the network verifies the transactions using cryptographic algorithms and if there is a corrupt attempt, the agents will not arrive to a consensus, preventing the transaction of being incorporated in the blockchain (Technology, 2016). Thus, smart contracts could reduce the immense amount of daily bureaucracy the insurance industry holds by self-executions only responding to certain pre-established triggers. Back-end activities could be automated without the need for a central database model, increasing efficiency by ceasing data duplication and decreasing claim processing costs (Ramada-Sarasola, 2016).
Blockchain could also disrupt the insurance industry in some other areas. Due to the disintermediation that blockchain offers, some small intermediaries would become obsolete. Small companies in charge of administrative or billing processes would see how this technology eliminates the need of face-to-face activities such as identity validation or executive transactions made by notaries, lawyers or insurance brokers. (Ramada-Sarasola, 2016). Furthermore, as in any industry, the insurance industry seeks for an increase of effeciency in their processes. The combination of “Internet of things” and blockchain could offer an immense new world of available data interconnected through all the entities in the industry (Shelkovnikov, 2016).
These last two points will be discussed more in depth in section 1.4 Blockchain in Healthcare.
Insurance products could also be disrupted as blockchain generates a better understanding of pricing and customer engagement. Data would be accessible to all due to the decentralized public ledger, thus, companies could use analytics to study pricing and offer new types of insurance more customer tailored (Ramada-Sarasola, 2016). Customers would perceive a greater degree of transparency and fairness of tariffs which would lead to an increase of trust in the insurance companies (McKinsey&Company, 2016). Reaching new markets, previously underserved is also another one of the possibilities that the use of blockchain in insurance could provide. A higher number of people in developing countries own a mobile device (Kahn, Yang, & Kahn, 2010) and there exists a high demand of products and services where micro insurance could benefit from. However, while administration costs and policy cancellations are very high, loyalty is weak, rising the reluctance of insurance company to operate in those markets. Most of these problems could be avoided using smart contracts and mobile platforms (Ramada-Sarasola, 2016).
2.3 Healthcare system in The Netherlands
Until 2006, the Dutch health system was a mixed scheme based on public insurance, together with a range of private insurance companies covering additional treatment. After the reforms, the focus shifted to the demand side, with the creation of three markets, patients, healthcare providers and insurance companies. The intricate regulations in healthcare in the Netherlands can be divided into regulations regarding public health; quality of healthcare services; and self-regulation of health insurers and healthcare providers.
The government has taken a step back from a controlling to a supervisory role. The Ministry of Health, Welfare and Sport still has an important role in health policy development and implementation. For example, the government has a key influence on cost development in the sector by determining the content of the
certain tasks have been taken over by new relatively independent structures or agencies such as insurance companies, which oversee collecting basic and premium packages from patients and buying services from health care providers for their clients. The system was modified to increase its efficiency by focusing on three main goals: quality of care (effective, safe and patient-centered), accessibility to care (reasonable costs for individuals, travel distance and waiting times) and affordability of care (cost control of the system) (M. Van Den Berg, Groenewegen, & Groenewegen, 2016). Nonetheless, there are already some weak points that need to be addressed. Quality of care is not yet a leading principle in the negotiation process between insurance companies and health providers, the focus of which is mostly on price and volume.
Figure 2. Actors and markets in the Dutch healthcare system since 2006 Source: HIT Netherlands
Patients do not have so much influence either, the information on quality of care, freedom of choice of healthcare provider and the option to switch insurer because of quality issues is still hardly used. On the other hand, health insurers have gained in negotiating power by clustering in four main groups (representing about 80% of the insured patients). Moreover, whereas hospitals are merging to increase
Health systems in transition The Netherlands
24
insured people and healthcare providers – operate in three markets: (1) for health insurance, (2) for health services provision and (3) for healthcare purchasing (see Fig. 2.2). In the health insurance market, health insurers offer the basic insurance package, which is obligatory for all citizens. The healthcare purchasing market is where health insurers can negotiate with providers on price, volume and quality of care. In the health services provision market, providers offer care that patients can choose to use. In their policies, health insurers may impose restrictions on the patients’ free choice of provider (usually in return for a lower premium).
Fig. 2.2
Actors and markets in the Dutch healthcare system since 2006
Source: Authors’ compilation.
Freedom of choice is essential in this system. To freely choose their health insurer and providers, patients need to be reliably informed about insurers and providers. Therefore, the government has put increased effort into making information available on waiting lists, quality and prices of care through the Internet (see also Section 2.9). The content of the basic health insurance package is fixed, but insurers can compete on the price of policies and the quality of care offered, as long as they observe both the obligation to accept applicants and the ban on premium differentiation. They have freedom in the content of
Healthcare provision market Health insurance market Healthcare purchasing market Dutch Health Care Authority
Government (regulation and supervision)
Insured/
patients Providers
their negotiating power, many small providers, such as general practitioners (GP) and physical therapy practices, are not allowed to combine their efforts and have their associations negotiate for them (M. Van Den Berg et al., 2016). The introduction of managed competition has not yet balanced bargaining power amongst parties.
Regarding information systems in healthcare in the Netherlands, the demand has increased over the last years (M. Van Den Berg et al., 2016). Different health providers are investing in new systems of information exchange, both within their organization and with partners in the supply chain. Therefore, there could be a potential for blockchain adoption between these players. There is already an information system that allows general practitioners to share information with pharmacist to generate data for disease prevention and research, close to the blockchain promise of a shared decentralized database for medical researches. For example, the NIVEL Primary Care Database (NIVEL Zorgregistratie, NZR) gathers data from a large sample of GP practices to oversee GP care in the country for health policy purposes. However, the industry still faces resistance for data sharing; some years ago, a national roll-out of the Electronic Patient Record (Electronisch Patiënten Dossier, EPD) was blocked in the Senate after vigorous debate and opposition. The reason was that the privacy of patients was insufficiently guaranteed. The cryptography of blockchain could improve this situation and encourage the use of Electronic Patient Record. Nonetheless, this will not happen until blokchain proves itself to be an enough trustworthy technology to manage such delicate information. One positive point is that blokchain allow patients to access their medical records, whereas the existing system called Care Infrastructure (Zorginfrastructuur), only allows GPs, pharmacists and specialists to exchange and see information from patients who have explicitly given their consent, but not the patients themselves (M. Van Den Berg et al., 2016).
Lastly, an extra motivation to study the potential adoption of blockchain in the Dutch healthcare industry was the government health budget priorities for last 2016. The focus was on empowering citizens for a better health choice; cost driven
policies; less bureaucracy and therefore, smaller transaction costs, the stimulation of research in fields as mental illnesses and energizing e-Health applications – all points that could be beneficed of the adoption of blockchain.
2.4 Blockchain in Healthcare
It is a very interesting time for health care and information technology (IT). There is a hype around blockchain and its potential uses in the health care industry. This paper wants to study these potential uses and impacts on the Dutch health care market. Blockchain has been defined as the technology capable of addressing the interoperability challenges in health IT systems as well as to enable the different players in the industry to securely share electronic health data (Linn, 2014).
The promises and pitfalls of this technology are interrelated. Where advocates of blockchain would claim how medicine could become much personalized, detractors stare at its immature infrastructures or the not safe enough patient controlled data (Tierion, 2016).
Disruptive innovations in healthcare never came easy. The reason is not only the reluctance that patients may have when things are related to their personal medical data, but also from powerful institutions such as insurance companies, health providers or doctor specialists who fight against simpler alternatives to expensive procedures just because those options threaten their occupations (Clayton M Christensen, Bohmer, & Kenagy, 2000)
.
Blockchain in healthcare is still in a very early stage, requiring education around use-cases that can benefit from it, unlike in finance where it is a worldwide phenomenon. For instance, the use of blockchain in healthcare could gather data from the population, in which individuals are already collecting it by themselves with the use of wearables sensors or applications, and be securely stored in a healthbank. People are tracking and obtaining more information about their health than before (i.e. daily blood pressures or blood sugar levels), and this daily personalized data could engage the patient more in his own health care whereas allowing physicians to improve individualized care and treatment plans (Linn,
2014). Blockchain can enable for the first time the long sought longitudinal health record that contains every episode of care from childhood to old age in every location healthcare was delivered. This capability would radically reduce medical errors and improve care quality as well as empower individuals to have full control over their own health records. Furthermore, Dutch government policy has been encouraging individuals, such as the elderly and those suffering from Alzheimer’s disease, to stay at home for as long as possible, supported by home care (M. Van Den Berg et al., 2016). This home nursing could be extremely promoted with wearables and e-Health applications’ progress.
Mobile applications and wearable sensors can facilitate a 24 hours-a-day monitoring of high risk patients, giving care givers the possibility, through “smart” applications, of reaching out to the patient and better coordinate their treatment options. Even nowadays, healthbank users are not only able to save their data on the platform, but can also make it accessible for medical research (Mettler, 2016). Clinical trials would become more transparent and accessible and health researchers could also develop studies from other researchers who previously started and did not finish due to lack of time or money. Real-time shared data would deliver broad diverse data from different geographic populations while allowing researchers and public health resources to quickly detect and isolate emergency situations for public health.
Healthbank thereby becomes a unique data-trading platform, with new opportunities in oriented strategies (Mettler, 2016). This patient-empowerment also reveals itself as an important tool for improving customer engagement. Patients’ fears about losing control of personal data as soon as it is handed over to any health provider or insurance company and their frustration with the need to repeat data entry processes can be addressed by blockchain (McKinsey&Company, 2016) – the distributed framework for patient digital identities, which uses private and public identifiers secured through cryptography, creates a singular, more secure method of protecting patient identity.
Marta Vargas Sánchez Master Thesis, Spring 2017 Student ID: 11376732 Blockchain technology in the Dutch healthcare industry
Figure 3. Potential roadmap for adding medical data to the blockchain
Source: Blockchain: A Comprehensive Look at Blockchain Applied to Healthcare, Article Two—
Information Exchange and the Need for Trustless Collaboration, Deloitte.
Moreover, according to the Dutch National Security Centre, cybercriminals pose the biggest threat to the healthcare sector (Cyber Security Assessment Netherlands csan 2016, 2016). There has been as well an increase in private fraudsters, for instance in the insurance field (see Figure 4, below). Thus, the immutable nature of blockchain could play an important role in the battle against cyberattacks.
Nevertheless, there is a significant averseness from the population to disclose their health data in the blockchain. Due to its young nature, immature infrastructures and scarcity of blockchain expertise, patients will not be likely to hand over their data so freely because of the probable consequences of failure. For example, the disclosure of patient health data to the wrong individual, or the loss of data in transfer processes.
Managing access to healthcare data is more complicated than managing payments and the backdoor of care givers accessing a patient health records in case of emergency represents a security risk (Tierion, 2016).
7 trust between entities are two reasons why it can
help solve the interoperability problem better than today’s existing technologies. As shown in Figure 1, an interoperable and comprehensive health record on the blockchain would most likely be pulled directly from existing EHRs in hospitals and physician offices. Today’s health records are typically stored within a single provider system. With blockchain, providers could either select which information to upload to a shared blockchain when a patient event occurs or continuously upload to the blockchain.
Patient-generated data could also be added to the blockchain. Previously agreed-to data standards would be applied to data entering the blockchain through smart contracts (decentralized applications that automatically execute actions based on blockchain activity), resulting in readable and consistent data from all sources. Trust issues
blockchain’s automated data verification. No intermediary is needed and blockchain users do not have to communicate. Participants would have control over who accesses the data, which would be
tamper-resistant once inside the blockchain. Concerns around privacy and security are major obstacles to sharing data. Health care organizations worry about the Health Insurance Portability and Accountability Act (HIPAA) and related regulations. While breaches are a valid concern, data sharing (when done with the proper precautions and consents) is not a violation of HIPAA. Blockchain can provide a more secure environment to store and access data. For example, to ensure information on the blockchain is shared only with authorized users, patients can grant access to their information to physicians, insurers, and others by providing private keys to unlock the data to these select entities.
Source: Blockchain: A Comprehensive Look at Blockchain Applied to Healthcare, Article Two—Information Exchange and the Need for Trustless Collaboration, Deloitte.
› Providers perform medical services for a patient
› Clinical data from a patient interaction recorded in the existing EHR system
› A standardized set of data with the patient’s corresponding public ID is directed to a blockchain API
› A smart contract then processes the incoming transaction for blockchain recording
› Network permissions validate that the submitter has access to the blockchain
› The data block is formed and all copies of the blockchain are updated
Figure 4. Number of incident investigations and instances of fraud in the Dutch insurance market Source: Factsheet fraudecijfers 2015, Anti Insurance Crime Bureau
This paper will study the situation in the Dutch healthcare industry for the adoption of blockchain and the possible reactions of the market participants and technology leaders, for the operational transformation.
3. THEORETICAL FRAMEWORK
In this section, the theoretical framework for the deductive analysis performed in the thesis will be elaborated upon. First the theory of disruptive innovations will be discussed, as well as the disruptive innovations applied to healthcare. This will be followed by the theory of disintermediation. Both theories are interconnected as the last one is considered itself a disruption.
3.1 Theory of disruptive innovations
An innovation will get traction only if it helps people get something that they’re already doing in their lives done better.
Clayton M. Christensen
Applied to technology, there exists a significant distinction between sustaining technologies and those called disruptive (C M Christensen, 1997); sustaining innovations are the result of dominant players in the market. These companies focus on improving their products or services so they can be sold for higher profits to the best customers. However, the speed of sustaining innovation almost always exceeds the capacity of customers to absorb it (Clayton M Christensen et al., 2000). This generates an opportunity for new entrants to introduce disruptive innovations. In contrast with sustaining innovations, the disruptive product is just simpler, more convenient and cheaper, fulfilling the needs of less-demanding customers previously ignored by the market.
The term ‘disruptive innovation’ should not be tangled with ‘radical innovation’. The three main characteristics that Christensen explains on disruptive innovations starts with the first period after implementation: a disruptive innovation initially provides inferior performance compared to existing products available in that industry. It is therefore often not seen as a threat or potential replacement by existing market actors and users. The second attribute of these types of innovations is that they serve markets that did not exist before. Therefore, existing producers are often not interested in disruptive innovations, as they do not
serve their existing customers. Thirdly, once the disruption product is established, it will tend to have a very steep improvement trajectory over time and customers from sustaining companies will start using them because the products meet their needs. (C M Christensen, 1997)
Figure 5. The theory of Disruptive Innovation
Source: C.M. Christensen, The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail
(Boston: Harvard Business School Press, 1997).
The main difference between a sustaining and a disruptive company is the value proposition of their business models. It is one of the four components of a business model (Hwang & Christensen, 2008); over time an established company creates a business model around its first value proposition, determining the resources, processes and profitability the company will operate with in the market. Hence, the business model will define if the company can effectively operate against new incumbents or not (Hwang & Christensen, 2008). When startup companies introduce their disruptive innovations in the market, not every original market leader is able to successfully adapt its business model fast enough to become a leader in the new disruptive situation.
3.1.1 Disruptive innovations applied to health care
In the healthcare industry technology is always applied in a sustaining style (Hwang & Christensen, 2008). Many dominant companies are overshooting the needs of average customers with innovation on sustaining products. Major health providers have developed remarkable technological advances to improve the needs of a small amount of very sick people, whereas much less investment has been done in the research of providing health care to a larger population with less serious diseases.
In other industries disrupted innovations enabled less-skilled people to do tasks historically only performed by experts on the field (Clayton M Christensen et al., 2000). An example is the disruption of computing hardware when minicomputers were replaced by more affordable and simpler personal computer (PC). In the 1960s, people needed to bring their information in form of cards to university or corporate centers where skilled computer scientist or technicians could help them to run the information. After the introduction of personal computer, although less powerful, people could perform their jobs without external help, at their own offices or homes. As PCs improved, becoming more powerful, less and less people used the expensive mainframe.
The health care industry could benefit from disruption innovation, by allowing less-skilled personnel as nurse practitioners to treat simple diseases usually treated by physicians, or in addition, allowing primary care physicians to treat simpler illnesses that used to involve specialists. (Clayton M Christensen et al., 2000)
Applied to Figure 5, there is a bottom less-demanding tier for patients with simple infectious diseases who could be treated by less-skilled professionals whereas the upper more-demanding tier containing more complex diseases treated by specialist and experts on the field (Figure 6).
The practice of user networks in healthcare institutions or major health providers would facilitate the exchange of information and care advices between professionals from different expertise levels. There is a vast amount of patient and insurance data that could be used to build an information network to learn, find
partners or share data. (Hwang & Christensen, 2008). One of the promises of blockchain technologies in health care is Data Integrity and Security. The technology has demonstrated in the financial field that reliable, accountable computing is a reality due to decentralized network of peers in a public ledger (Yue, Wang, Jin, Li, & Jiang, 2016).
Figure 6. Disruptive Innovation of Health Care Professions
Source: C.M. Christensen, R. Bohmer & J. Kenagy, Will Disruptive Innovation Cure Health Care?
(Harvard Business Review, 2000).
The volume of patient data increases per year (Tierion, 2016), and sharing health and patient data between institutions is defiant. However, blockchain systems have strong data integrity, meaning that once the information has been recorded in the ledger it will remain there forever. Likewise, operating in a public blockchain network, a health care one in this situation, any health care professional or scientist verified with the integrity and timestamp of any data, file or health knowledge in this case, can use the open source without relying on a trusted third-party. The purpose of this paper is to analyze if blockchain technology, as a disruptive innovation, could be a reality inside the healthcare industry and how the intermediaries in the industry, thus, health insurance companies would react to it.
3.2 Theory of disintermediation
The term of disintermediation was coined in the banking industry in late 1960s to describe the phenomenon of dividing savings from bank accounts that generated low interest rates and invest them in other financial institutions with larger profits. New entrants created companies focused on fulfilling the investment needs of bank customers, hence the population stopped considering bank institutions the only intermediary part between small depositors and the financial market. The financial market was disrupted and due to the loss of some of their formerly captive market, banks adapted themselves by paying higher interest rates to their customers. The banking industry has experienced major changes because its interactive communications (Gellman, n.d.), thus the healthcare industry may as well experience it, as every player in the market is continuously interacting between each other. The possibility of disintermediation is growing deeper in healthcare supply chains. Disintermediation is in the core of blockchain technology (Seppälä, 2016) and it offers important economic benefits, however, it also requires that manufacturers and providers develop new expertise. The industry changed in 2006, therefore, there is no longer a traditional task distribution in the supply chain but a supply chain driven by the consolidation of hospitals and cost driven strategies. Increasingly, hospital groups and manufacturers are seeking to do business directly with each other to avoid distributor markups (Practice, 2004) supporting the idea of disintermediation models being a solution for the more and more unpredictable world, where agility is growing to be more important than stability (Seppälä, 2016). The concept of disintermediation has some irrefutable effects. First, both value and power could be redistributed in the network as a result. Furthermore, with the introduction of Internet of Things and wearables, the healthcare industry may become closer to a more distributed model, where information comes from many different parts and services are offered closer to the consumer, and the roles of producers and consumers could be more concurrent rather than mutually exclusive. This would lead to demand and supply meeting more efficiently, and to giving
patients more power to choose what kind of services they want to consume (Seppälä, 2016).
Nevertheless, when disintermediation involves IT, the current intermediaries do not go away, because they are in a good position to benefit of the new intermediation, this is called re-intermediation (Avital, King, Beck, Rossi, & Teigland, 2016).
This could also be possible with blockchain in healthcare. The paper of Avital et al: Jumping on the Blockchain Bandwagon: Lessons of the Past and Outlook to the Future Panel (2016) states that there are established companies that are most likely to appropriate and apply blockchain to their own interests, while restricting solutions that may replace them. In other words, they are more likely to retain power by adopting the solutions that blockchain offers while obstructing those innovations to replace them in the market (Avital et al., 2016). This paper will analyze three propositions conjectured from these two theories. The research strategy and framework will be discussed in the next section.
4. METHODOLOGY
In the following section, the research design for studying the impact of blockchain adoption in the Dutch healthcare industry will be elaborated upon. Therefore, the researcher will discuss the use of case studies as the research method for this thesis, along with the participant selection procedure. This will be followed by a section in which the researcher outlines and evaluates the methods for data collection (interviews and content analysis), as well as the strategy for data analysis. To conclude, the followed coding process will be described.
4.1 Philosophical assumption
There are different philosophical assumptions and worldviews regarding truth, reality, knowledge, values or writing styles. Researchers should explore these to determine which they believe and how it could impact their research methods, because aware or not, beliefs and assumptions are always involved while doing research (Pratt, 2009). Those assumptions will often come to serve as a foundation of the research by influencing the theories chosen, the way the information is pursued and the analysis of the entire research process (Creswell, 2013).
To comprehend the approach to research, it is important to acquaint that, instead of adhering to notions of an “objective reality”, the researcher embraces the pragmatism and social constructivist idea of different subjective realities (Creswell, 2013). Due to the young nature of the research topic, the inquirer understood that executing qualitative research was the right choice as it would allow herself to focus on the several meanings that individuals raise and assign to social issues (Pratt, 2009). The choice is especially appropriate for issues lacking of previous information from the literature, or issues that are not easily measured by quantitative methods (Creswell, 2013).
4.2 Research strategy: Case study
In this section, the researcher will detail why case study was chosen as the most appropriate research strategy for this thesis, as every qualitative researcher needs a strategy to enhance the rigor and sophistication of the research design (Creswell,
2013). The explorative nature of case studies makes them particularly suitable for emerging topic areas where slight previous empirical work is accessible (Eisenhardt, 1989; Gibbert & Ruigrok, 2010; Yin, 2009a). Given the state of blockchain technology research in healthcare to date, case studies have been identified as the most appropriate research strategy to follow. Case studies focus on a concrete entity – an individual or small group – nevertheless, it is still possible to center attention on more abstract phenomenon such as relationships, projects or processes, that enables comparison of different sources and perspectives describing the same case issue at hand (Creswell, 2013). The focus of the case itself appeals upon multiple sites, situations and sources of information to detect and shape rich case descriptions and themes (Creswell, 2013). It is indeed the deeper connection with reality that allows the research to arise concepts or propositions with a novel, relevant and empirically validity (Eisenhardt, 1989); some of those groundbreaking insights revolutionizing the field of management (Gibbert & Ruigrok, 2010). Nonetheless, it is key in case studies to ensure credibility in its operational procedures (Gibbert & Ruigrok, 2010). Therefore, to determine the quality of this case study it is the duty of the researcher to exhaustively describe each footstep taken from here.
4.2.1 Research framework
As mentioned in the theoretical framework section, the researcher focuses her study on the theories of disintermediation and disruptive innovations, to test how the adoption of blockchain technologies in the Dutch healthcare industry could disrupt the current intermediaries’ scenario or not. Disintermediation is in the core of blockchain technology (Seppälä, 2016), as it allows the members of the system to exchange information or transactions in a safe manner, thanks to its cryptographically secured network. This enhances transparency and trust. Thus, the researcher asks herself what will happen with the entities that until now, have overseen that trust with society’s approval. In the Dutch healthcare industry, these bodies are the health insurance companies and the health providers, who connect
these intermediaries? Is it going to be disruptive enough to modify the current business models? What is the added value of implementing a still unknown technology in one of the most regulated1 industries in the world? All these
questions asked by the researcher find a common ground in the theoretical approach (see figure 7 below)
Figure 7. Research framework - Propositions
Proposition #1:
As disintermediation is known as cutting intermediaries from a supply chain (Avital et al., 2016), the first proposition explores if blockchain technology will be a threat to the current intermediaries in the industry, focusing on health insurance companies.
“Disintermediation is a direct consequence of the adoption of blockchain technology for health insurance companies in the Dutch healthcare industry.”
Proposition #2:
However, disintermediation is considered a disruption as well, and the researcher was keen on knowing if the adoption of blockchain technologies will develop new business models in the Dutch healthcare industry.
1
Blockchain in the Dutch
healthcare industry disintermediationTheory of
Theory of disruptive innovations Proposition 1
Proposition 2
“Blockchain technologies will accelerate collaborations between players in the Dutch healthcare industry.”
Proposition #3:
Finally, blockchain technology is a new phenomenon with little academic literature about it. One of the reasons for disruptive innovations to be adopted is because they are simpler, quicker and cheaper, when fulfilling the demands of the less powerful entities. There is a powerful benefit in the use of the innovation against the current product or service. The healthcare industry is considered an opaque sector where the information does not flow naturally between players and where patients do not control their health data. An increase in transparency of the use of the data, who owns, sees and modifies it, may impact the market and shorten the distance between entities and patients.
“Health insurance companies and health providers will establish trust between institutions and patients using blockchain technologies.”
4.2.2 Access to the field
When the researcher moved to the Netherlands before starting her studies, the differences in healthcare between her home country –Spain- and the Netherlands strongly hit her. In Spain, the industry still works as it did in the Netherlands until 2006, and the market consists of not so many players. Therefore, when the researcher studied blockchain, she could not help but to wonder if the care industry could potentially adopt the technology and if so, the ramifications of such implementation. Significant to mention as well that the Netherlands ranks 4th in the Digital Economy and Society Index - DESI 2017 (Economy, 2017). The Netherlands ranks highest in connectivity; Dutch citizens are very active users of the Internet and have the right skills to do so. Also, the digitization of Public
the researcher’s curiosity to study the effects of blockchain as a new technology in healthcare in the Netherlands.
Having purposefully delimited the geographic area of research, the researcher was able to identify and approach different players in the industry. First, she started with the health insurance market, initially believing that it would be a difficult market to reach without any referral. After learning through Zorgverzekeraars Nederland -the umbrella organization of nine health insurers in the country- about the different companies, their sizes, partnerships and projects, the researcher began to call and send emails to each one of the nine insurers. With the help of a brief and graphic attachment of the thesis topic, in some occasions the researcher managed to narrow the search and reach someone in the IT department of the specific company, who enthusiastically agreed for an interview. In other occasions the e-mail was replied with an apology of not being able to spare any time for the research. Nonetheless, the researcher also received some referrals from guest speakers in the University of Amsterdam that the researcher approached and discussed the topic with. This method was of great benefit to reach big IT companies, otherwise very difficult to be reached through unsolicited ways.
4.2.3 Participant selection procedure
In contrast to quantitative, hypothesis-testing research relying on statistical (random) sampling, qualitative researchers typically engage in purposive sampling (Eisenhardt, 1989). This means that potential interviewees are subject to a participant selection procedure that is progressive and supported on the theory implemented, where the researcher must make a wide range of sampling decisions to ensure participant suitability (Gibbert & Ruigrok, 2010). As the reasoning behind these careful sampling decisions is considered invaluable for determining the overall quality of the research (Gibbert & Ruigrok, 2010), the researcher wishes to make these selections transparent to the reader. As previously discussed in the background section, the whole Dutch healthcare industry was profoundly transformed in 2006 with the creation of three markets: patients, healthcare
providers and insurance companies. As patients would belong to a different research method than case studies, for example, surveys or experiments, the researcher concluded to contact employees from health insurance companies and care providers.
For 11 years now, the market is dominated by insurers operating on a non-profit basis. The umbrella organization of (currently nine) health insurers is Health Insurers Netherlands (Zorgverzekeraars Nederland, ZN2). The three largest
insurance groups have a market share between 11% and 30% (Verbond van Verzekeraars, 2015); hence the importance to obtain data from insurance companies inside that cluster. However, as Porter (2008) stated competition among existing firms in an industry is significant as well when analyzing the whole sector. Therefore, the researcher decided to also contact medium and small size insurance companies to get their judgement on blockchain in the healthcare industry. The market share distribution of the health insurance umbrella organization is as follows:
Figure 8. Market share of Dutch health insurance market Source: (NZa, 2016)
Figure 9. Overview of the 25 health insurance companies in the Netherlands Source: NZa, 2016
The researcher contacted nine health insurance companies, and got interviews with five, two of which are among the three biggest insurance companies based on market share, and three smaller ones with the aim of comparing if company size makes a difference for the adoption of blockchain. This sample covers a representative part of the health insurance industry market because it is more than half the number of companies inside the umbrella organization. For the health providers, the researcher wanted to interview big health providers such as hospitals. Nevertheless, from the broad national sample and because of the early nature of blockchain technology, the researcher decided to limit her interviews to academic hospitals, as they are more prone to innovate and experiment. As a student of the
MARKTSCAN ZORGVERZEKERINGSMARKT 2016 | 5
Aantal zorgverzekeraars
In 2016 zijn er in totaal 25 zorgverzekeraars. De meeste daarvan vallen onder een overkoepelend concern. Op dit moment zijn er negen concerns. In 2016 zijn er (tot het moment van dit schrijven) geen wijzigingen ten opzichte van 2015 in de aanwezige concerns of zorgverzekeraars. In de loop van 2015 zijn alleen de juridische namen van OZF en Zilveren Kruis gewijzigd, waarbij de toevoeging ‘Achmea’ is weggelaten.
Aanbod concerns en zorgverzekeraars
Concern Zorgverzekeraar
Achmea Avero Achmea Zorgverzekeringen N.V. De Friesland Zorgverzekeraar N.V. FBTO Zorgverzekeringen N.V. Interpolis Zorgverzekeringen N.V. OZF Zorgverzekeringen N.V.
Zilveren Kruis Zorgverzekeringen N.V.
ASR ASR Basis Ziektekostenverzekeringen N.V.
CZ Delta Lloyd Zorgverzekering N.V.
OHRA Ziektekostenverzekeringen N.V. OHRA Zorgverzekeringen N.V.
OWM CZ Groep Zorgverzekeringen U.A.
DSW-SH OWM DSW Zorgverzekeraar U.A.
Stad Holland Zorgverzekeraar O.W.M. U.A.
Eno Eno Zorgverzekeraar N.V.
Menzis Anderzorg N.V.
Azivo Zorgverzekeraar N.V. Menzis Zorgverzekeraar N.V.
ONVZ ONVZ Ziektekostenverzekeraar N.V.
VGZ IZA Zorgverzekeraar N.V. IZZ Zorgverzekeraar N.V. N.V. Univé Zorg N.V. Zorgverzekeraar UMC VGZ Zorgverzekeraar N.V. N.V. VGZ Cares
Zorg en Zekerheid OWM Zorgverzekeraar Zorg en Zekerheid UA
zorgverzekeraars
9
concerns25
2
4
1
3
5
2
University of Amsterdam the researcher had the chance to approach AMC (Academic Medical Center), a university hospital affiliated with her university. After a general exploration about the current situation of blockchain in the country the researcher learned that the area of Nijmegen stands out regarding blockchain experiments, with a high participation in hackathons3.
Therefore, the researcher approached REshape Center, an innovation department under the responsibility of the Radboudumc University Nijmegen Medical Center. The difference in number of participants between health providers and health insurers is explained by the researcher’s consideration of health insurance companies as a purer intermediary than care providers. Although both are connectors, the second one provides the care itself, therefore its disruption seems less plausible. However, as blockchain is a relatively new phenomenon, the researcher decided to also contact players in profounder contact with the technology and its uses and values, actors who could be considered comparable to “blockchain experts” –while acknowledging that, because of its young nature, expertise comes in another level than in different sectors. After exploring the existence of blockchain innovations into their strategies, Philips Electronics Nederland B.V and IBM were interviewed.
In terms of selecting an appropriate number of participants, research scholars stress that there is no “magic number” in qualitative research (Eisenhardt, 1989; Gibbert & Ruigrok, 2010). Instead, the most important issue to consider is whether the data will provide and substantiate meaningful and significant claims (Tracy, 2010). In providing some guidance, however, Eisenhardt (1989) suggests that a number between 4 and 10 usually works well, noting that with more than 10, it quickly becomes difficult to cope with the complexity and volume of the data. Although, at the beginning the researcher was not sure about how many companies would agree to be interviewed about this new and –in many cases- unknown technology, the participation reached 11 interviewees and one last individual who
offered to answer some questions through e-mail. Although acknowledging that the amount exceeds the advised 10 participants, the researcher made the decision of accepting the additional work load to obtain a fuller insight from a new and scarcely studied topic as blockchain technology in healthcare. Table 1 (Appendix A) provides an overview of all the 12 participants, containing self-reported information on the interviewees.
4.3 Method
One, if not the biggest, difference between quantitative and qualitative analysis is that the last one does not rely on a single data source but in multiple forms of data (Eisenhardt, 1989). To focus on the research question, the researcher must discipline while navigating through a wide array of data collecting options. Eventually, the aim is to triangulate or to adopt multiple data sources, methods, and theoretical lenses to look at the same phenomenon (Gibbert & Ruigrok, 2010; Tracy, 2010). It is a method conceived as valuable as it increases scope, deepens understanding, and allows different sides of an issue to be analyzed (Tracy, 2010). Tracy (2010) stated that triangulation serves to capture different data and reveal different insights to provide greater nuances that open a more complex understanding of the issue. On the other hand, triangulation may also serve to converge data in a way that may demonstrates validity and strengthens substantiation for the emerging propositions construction (Gibbert & Ruigrok, 2010). This thesis groups semi-structured interviews alongside content analysis of each company’s official website and conferences or discussions as methods for data collection.
4.3.1 Semi-structured interviews
One of the methods of the data collection is in-depth interviews, conducted to understand if and how participants experience the blockchain phenomenon on their job positions and companies. Here, the aim was to directly talk to individuals from the IT department or at least within the closest contact to informatics or technology (see overview in Table 2 – Appendix A).
The interviews were recorded (pre-authorization) and semi-structured in nature, closer to an informed dialogue rather than “a survey administered out loud”(Leech, 2002a). This does not mean that questions were not carefully devised. In fact a general protocol was written prior to the interview sessions (see Appendix B) , but the researcher deliberately aimed to keep the questions open, broad and general so that the interviewees themselves could build the meaning of a situation (Tracy, 2010). The flexibility of the interviews allowed the exploration of topics as they emerged in the conversation, so the researcher chose to not control the discussion too much in order to not miss unexpected points (Leech, 2002b).
All the interviews were done during the period of end of April to June 2017, usually at the office location of the interviewed company. The length of the interviews was not planned but on average the interviews lasted between 30 and 45 minutes. In total, 6 hours and 30 minutes of interviews were audio-recorded for the research. As suggested by Leech (2002), the researcher also spent some more minutes before and after the interview talking with the participants, with no other goal but to establish some affinity and trust between the interviewee and the researcher. Nevertheless, these small and easy talks also helped the researcher to understand the issues surrounding the studied industry, and to obtain a bigger picture of the phenomenon. Lastly, once the interviews were conducted and transcribed, they were sent back to each participant for validation and clarification with an anonymity-clause option. By doing so, the researcher could ensure that participants recognized the transcripts as reliable statements of the interview done (Bryman, 2004; Gibbert & Ruigrok, 2010).
4.3.2 (Annual) reports, emails and brochures
As previously mentioned, the researcher obtained the data for her research from different sources, semi-structured interviews are only one of them. During her research, the researcher had access to annual reports, informative emails and brochures which served as additional input for the analysis. Content analysis is a
written communication (B. Berg, 2009; Stemler & Colors, 2001). The content analysis of the mentioned data was undertaken to grasp how companies, institutions or individuals try to publicly communicate their responses to disruptive technologies through different routes as official websites or reports. In conducting the content analysis, each text was analyzed with the goal of bringing to surface different meanings, themes, or discursive dominations (Creswell, 2013). However, it is important to mention that the analysis and meanings are a result of an overall interpretation made by the researcher, and should not be deemed absolute or conclusive, but a tool to corroborate the findings from the interviews. To provide an overview of the documents obtained, a table (Table 3) has been created below:
Company name Type of document Name of the file
Participant 2X Annual report 2017 Policy terms and conditions 2017
Participant 4Y Company report AMC - Academic medical center
Participant 5X Company brochure Not defined
Participant 6Z Annual report 2016 Philips annual report 2016
Participant 8Z Annual report 2016 2016 IBM Annual Report
Internal report Healthcare rallies for blockchain
Participant 9Y Annual report 2016 Annual report 2016
Participant 11B Internal report Report on eHealth objectives for 2016
Participant 12B Email Participant 12B
Table 3. Overview documents from participants
4.4 Ethical considerations
To achieve ethically sound research, anonymity has been offered and granted to participants who asked for it, to protect them from information disclosure that could adversely affect them (Gibbert & Ruigrok, 2010; Tracy, 2010). Safeguarding participants from unnecessary exposure and ensuring their reliability has been prioritized with the goal of allowing the participants to feel more comfortable speaking openly about their experiences once they were granted anonymity, thus increasing the quality of each interview. To this end, information that could be related with the participants has been changed and only the appointed examiners of
the thesis will know about these details. For the same reason, the public version of this paper will not contain the full interview transcripts nor the coding schemes. 4.5 Coding process
Through the coding process the obtained data can be improved, categorized and organized allowing its own analysis (Miles & Huberman, 1984). It is a way of index in the data. Here, the researcher aims to gradually focus the data without leaving it out of meaning or context in which it occurred (Eisenhardt, 1989). Coding involves the breaking down of all the obtained data into units which are then grouped according to their characteristics. To achieve this, the researcher followed a process of open, axial and selective coding where the data was organized inductively into core units of information while linking those variables with categories and themes from the bottom up. A computerized coding tool (NVivo) has been used to perform the coding process. Although coding schemes were guided by the research question and the theoretical approaches of “theory of disintermediation” and “theory of disruptive innovations”, the codes were not completely predetermined prior to data analysis. Only some of them emerged through a deductive but iterative process where the researcher worked back and forth between the theories and the raw data, finding key terms – codes- until a set of themes and categories were established (Creswell, 2013).
The first stage started with open coding of the data, which involved a search through the data to see what patterns are emerging. Together with the reading and re-reading of the raw data, thoughts and ideas were written down developing an initial coding scheme. This process was carried out on each unique source of data, therefore, the researcher got to know each text in a familiar way (Eisenhardt, 1989; Yin, 2009b). After the first contact, the researcher inquired into the theoretical approach for concepts that could support the results obtained from the raw data concepts (Appendix D). Afterwards, the process moved to axial coding, where found common statements were used to form provisional themes and categories. The
identify interrelationships with the two theories from the theoretical framework (See Appendix E and Appendix F). To deductively summarize the analysis and compile the relationships and patterns that emerged through open and axial coding, a selective coding was used (Creswell, 2013; Miles & Huberman, 1984) (See Appendix G).
Raw data (text) Source Appendix
General protocol for interviews Researcher B
Coding (tables/schemes) Source Appendix
Interview transcripts Audio recording C
Open coding - Interviews & reports Interview transcripts D Axial coding - Interviews & reports Open coding scheme E Definitions of code categories Axial coding - Interviews F
Narrative (quotes) Source Appendix
Quotes used to inform narrative Both selective coding schemes G
5. ANALYSIS
While quantitative research is mostly concerned with the testing of hypotheses and statistical generalizations (Jackson, 2009), qualitative research focuses instead on understanding the nature of the research question rather than on the quantity of observed attributes (Strauss & Corbin, 1994). Therefore, analyzing case study data is one of the least guided and most difficult aspects of undertaking case studies (Yin, 2009b). The abstract concepts should link to the theoretical foundations; therefore, the researcher frequently revisited the theoretical approach that support the research. As suggested by Miles & Huberman 1984 and Yin 2009, the researcher presents her findings following a case study narrative, to avoid that the big volume of the results becomes overly complex. In the next two sections the used code process and the analysis strategy will be explained.
5.1 From coding to coherent narrative
Several scholars suggest using narratives to present case study findings, as every case study has a story to tell (Yin, 2009). The goal is to build a logical chain of “evidence” by organizing information in a cognitively appealing order (Creswell, 2013; Miles & Huberman, 1984; Tracy, 2010; Yin, 2009b). Designing the structure and order of that narrative, and determining which data to incorporate and display, is an analytical activity that allows the researcher to tell their story (from the data) about how it all fits together (Miles & Huberman, 1984; Pratt, 2009). Here, it is worth noting that choices for inclusion or exclusion of data and the order of its presentation are never value free. To reinforce the analysis, the researcher is well advised to incorporate large amounts of data into the research story itself (Pratt, 2009; Tracy, 2010). This is rhetorically challenging, however, and there is a risk that the overwhelming volumes of inputs become overly complex (Tracy, 2010). To avoid this, this thesis contains a summary with every quote drawn upon in the analysis (Appendix F); making explicit reference to each throughout the text. Presenting it in this way does not make the reading tedious while explaining complex data. At some
