Digital Health Ecosystems for Sensor Technology Integration - A Qualitative Study on the Paradox of Data Openness

Abstract

Mobile health startups develop innovative, sensor-based solutions that continuously collect health data. To generate added value from these large amounts of data, an integration of the solutions into the healthcare system is essential. In this context, the collaboration between interdependent healthcare stakeholders is required which can be enabled by structures considered as digital ecosystems. To understand the conditions for ecosystem participation, more specifically the incentives and disincentives for data openness, we conducted 30 interviews with four healthcare stakeholder groups in Germany and analyzed the data using a Grounded Theory approach. As a result, we developed a conceptual model that describes the integration of mobile sensor-based health solutions into digital health ecosystems. Thereby, we improve the understanding of incentives and disincentives for data openness on the collective ecosystem level, the ecosystem-stakeholder-group level, and the individual user level. Practically, we contribute by outlining important market entry barriers for mobile health startups.

Full text

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
1
Digital Health Ecosystems for Sensor
Technology Integration - A Qualitative Study
on the Paradox of Data Openness
Completed Research Paper
Anne-Katrin Witte
Technische Universität Berlin
Straße des 17. Juni 135,
10623 Berlin, Germany
a.witte@tu-berlin.de
Daniel Fürstenau1,2,3
1Copenhagen Business School
Department of Digitalization
Howitzvej 60, 2000 Frederiksberg
dfu.digi@cbs.dk
2Freie Universität Berlin
School of Business & Economics
Garystr. 21, 14195 Berlin
3Einstein Center Digital Future
Wilhelmstr. 67, 10117 Berlin
Rüdiger Zarnekow
Technische Universität Berlin
Straße des 17. Juni 135,
10623 Berlin, Germany
ruediger.zarnekow@tu-berlin.de
Abstract
Mobile health startups develop innovative, sensor-based solutions that continuously
collect health data. To generate added value from these large amounts of data, an
integration of the solutions into the healthcare system is essential. In this context, the
collaboration between interdependent healthcare stakeholders is required which can be
enabled by structures considered as digital ecosystems. To understand the conditions for
ecosystem participation, more specifically the incentives and disincentives for data
openness, we conducted 30 interviews with four healthcare stakeholder groups in
Germany and analyzed the data using a Grounded Theory approach. As a result, we
developed a conceptual model that describes the integration of mobile sensor-based
health solutions into digital health ecosystems. Thereby, we improve the understanding
of incentives and disincentives for data openness on the collective ecosystem level, the
ecosystem-stakeholder-group level, and the individual user level. Practically, we
contribute by outlining important market entry barriers for mobile health startups.
Keywords: Digital health ecosystem, data openness, mobile sensor-based health technology
Introduction
The past years have seen a rise in innovative technologies, primarily driven by mobile health startups
introducing monitoring or diagnostic applications enabled by smart algorithms and mobile sensors. Sensor-
based solutions allow to continuously acquire “big data” (Raghupathi and Raghupathi 2014), help create
high-definition risk profiles (Torkamani et al. 2017), support patient engagement (Chiauzzi et al. 2015),
allow more accurate interpretation of disease symptoms (Raij et al. 2011), and facilitate self-tracking
(Gimpel et al. 2013). In this way, care services become more preventive, cost effective, and precise.
However, “big data” created by sensor-based tracking and tracing is also met with resistance (e.g. in the
case of the Covid-19 tracking application) for reasons often related to unwanted surveillance of one’s private
behaviors (Raij et al. 2011) as well as unilateral value claims by private companies (Zuboff 2019). Therefore,

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
2
the incentives to collect and share health data openly are pronounced differently and depend on the
stakeholders’ perspective. Policies in favor of a user’s privacy protection, allowing for a choice of with whom
their health data is shared, run counter to the companies’ preference towards (more exclusive) access to
“big” health data and to developing more innovative offerings (Nambisan et al. 2019). This “paradox of
openness” (Arora et al. 2016) describes the tension between the creation and appropriation of value and
illustrates that focusing on the openness of health data on one level, for instance on the individual user
level, ecosystem-stakeholder-group level or collective ecosystem level exclusively, is not feasible. This calls
for a more nuanced, stakeholder-differentiated view on sensor-based health technologies.
The purpose of this paper is to conceptualize incentives and disincentives, as well as barriers for data
openness and sharing, in the context of integrating sensor-based health technologies into digital health
ecosystems. Sensor-based solutions allow for a stronger interconnection of different actors and enable, but
also require structures that are commonly discussed under the term “ecosystem.” An ecosystem can be
defined as a “set of actors with varying degrees of multi-lateral, non-generic complementarities that are not
fully hierarchically controlled” (Jacobides et al. 2018). While some researchers have highlighted that these
ecosystems often emerge in the context of digital platforms through which the ecosystem participants
become connected (Parker et al. 2017; Song et al. 2018) others have described more dispersed forms of
coordination in innovation ecosystems (Adner 2017; Giudici et al. 2018; Jacobides et al. 2018). In the
healthcare field, ecosystem-based coordination via digital means is only in a nascent stage in many
European countries, as illustrated by the Digital Health Index published by the Bertelsmann Stiftung
(2018). In considering digital health readiness, actual data use, and policy activity, the taillights of the
countries studied are formed by Switzerland, France, Germany and Poland which are ranked lowest (in
descending order). To enable service provision for patient value, defined as the betterment of the patient’s
condition (Rantala and Karjaluoto 2016), the interdependencies of healthcare stakeholders need to be
considered. There is a (non-digital) ecosystem that in many countries such as Germany (our country of
focus), centers around three stakeholder groups: patients, medical service providers, and health insurance
companies. The patients pay a monthly fee according to the insurance contract with their health insurance
and are treated by medical service providers in a case that they may fall sick. Medical service providers are
reimbursed by the health insurance according to their contracts and depending on the provided treatment
(Busse et al. 2006, p. 2). All interactions between these stakeholders are strictly regulated and subsidized,
as is the use of medical devices (Busse et al. 2006, p. 18).
Within Europe, and particularly in Germany, the development of provider-specific solutions and the lack
of a national digital health infrastructure creates a heterogeneous technological landscape. This causes
conflicts when the cooperation between different solutions or the consumption of services by other
providers is needed (Benedict and Schlieter 2015). It further aggravates the fragmentation of the market by
creating proprietary data formats and silo solutions. From a theoretical perspective, this absence of
prospering digital ecosystems calls for research into the rules and roles, and monetization, as well as how
actors are connected, which have been identified as important requirements for ecosystem formation
(Jacobides et al. 2018). In particular, we posit that technology-oriented streams of research on health
ecosystems (Benedict and Schlieter 2015; Vesselkov et al. 2019) should be extended by considering the
ecosystem concept also in a socio-economic and strategic light, as well as considering how sensor-based
hardware (Olla and Shimskey 2015), in contrast to software-based solutions, is integrated into digital
ecosystems and which specific challenges arise. In parallel, country-specific characteristics and their
associated legal and regulatory requirements play an important role, e.g. particularly in Iceland (Islind et
al. 2019) or Finland (Vesselkov et al. 2019) where similar studies with a slightly different focus have been
conducted. Furthermore, new measures at the national and European levels are currently being announced
over short intervals, such as with the European health cloud Gaia-X (The Economist 2020). This, in
contrast, creates high degrees of uncertainty for startups who want to enter the market. Since business
models cannot be pre-planned safely, incentives exist to collect “big data”, despite consequences on the
individual level, as this could maximize the chances of a company’s survival. This may alleviate the tension
between those positive effects enabled by sharing sensor-generated health data openly and the concerns by
different stakeholders on different levels to do so.
To address these issues as well as their implications, we conducted 30 interviews with stakeholders in the
emerging mobile sensor-based health technology (MSHT) ecosystem in Germany. This ecosystem includes
the actors, activities, and those relationships involved in providing sensor-based solutions for use with
patients in order to create added value for them. Interviews were coded using the Grounded Theory

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
3
approach, and within that process we viewed the collected material through three purposeful lenses: (1)
incentives and disincentives of ecosystem participation from a multi-stakeholder perspective (Jacobides
et al. 2018), (2) ecosystem participants’ stances toward openness and control (Nambisan et al. 2019), and
(3) the principles of the data economy (Zuboff 2019). From the systematic coding of the interviews and our
subsequent theory building effort, we have derived a conceptual model on the integration of sensor-based
health solutions in a digital health ecosystem. Furthermore, we identified that design and governance
strategy, in light of the paradox of openness, consist of two main phases within which different forms of
restrictions and controls apply. In this regard, we have identified different incentives and disincentives for
data openness and data sharing which are dependent on the stakeholder group and/or ecosystem level.
We contribute a phase- and stakeholder-differentiated consideration of the incentive differences in the
strategic design of healthcare ecosystems, as requested by Nambisan et al. (2019). Mobile health startups,
health insurances, and medical service providers have very different incentives for data openness and data
sharing which play out on different levels; namely, the collective ecosystem level, the ecosystem-
stakeholder-group level, and the individual user level, all of which we will argue should be considered in
future research. From a practical viewpoint, we inform MSHT integration in the context of digital health
ecosystems by identifying stakeholder-specific barriers to market entry, which can be addressed through
various framing, nudging, and regulatory strategies.
Theoretical Foundation
The concepts of mobile sensor-based health technology (MSHT) and digital ecosystems are described in
the following to define the scope of our study and to show their relation to the existing literature. We then
go on to develop our own perspective, synthesizing important insights from digital ecosystems, data
openness and the role of data in the context of health service improvements and business model
development.
Mobile Sensor-based Health Technology
On one side, medical and research grade sensor devices generally promise a high accuracy and are often
targeted at unhealthy or elderly patients and are designed for the management of a certain disease e.g.
diabetes (Gao et al. 2015). But these devices can be expensive, their outer appearance can be bulky and it is
hard for the user to set up the device independently. In stark contrast, the term fitness tracker is constantly
evolving and is generally defined as wearable technology that is worn on the wrist (Swan 2012), which is
more accessible to the average consumer and often mentioned in contexts in the fitness sector to help
healthy, young users track their daily lifestyle data (Gao et al. 2015). The convergence of the usability of
fitness trackers and the accuracy of medical sensor devices is demonstrated by various attemps of mobile
health startups to enter the healthcare market with medically certified end user products and services. In
this context we conceptualize the term mobile sensor-based health technology (MSHT). The word
“mobile” implies that the sensor is flexible and wearable so that it can be worn on the user’s body
continuously. The device can be used independently without the support of medical professionals and is
targeted at the end-consumer (e.g. patient). The word “sensor” represents the integration of any type of
sensor technology with the goal of capturing vital parameters of its user e.g. inertial measurement units
(linear and angular motion) or galvanic skin response sensors (skin conductivity). The exact sensor position
on the user’s body varies, as does its shape e.g. as wristbands or headbands. The concept of “health
technology” defines that sensor technology is applied in a medical and health related context.
Digital Ecosystem
In the context of digital ecosystems there are different research paradigms. Deriving from the platform
evolution framework of Tiwana et al. (2010), Schreieck et al. (2016) went on to frame key concepts and
issues for future research in connection with the design and governance of platform ecosystems. In contrast
to Tiwana (2010), they broadened the rather technical definition of platform “architecture” and morphed it
into “design”, which includes a conceptual blueprint of the ecosystem as a whole (Schreieck et al. 2016).
Within their findings, they reveal the issues with an “individual level of analysis to consider characteristics
of actors” as well as “the role of data as boundary resource” (Schreieck et al. 2016). Hein et al. (2019)
describe different foci on digital platform ecosystems within the existing literature e.g. technical, social,

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
4
economics or business paradigms only. In contrast to this single-paradigm research they introduce a
nuanced approach that integrates “the intra-organizational technical perspective on digital platforms and
the inter-organizational economic, business and social perspectives on ecosystems” (Hein et al. 2019). Such
an approach is close to our own understanding of ecosystems. In the context of our research question (see
Introduction), we consider technical aspects of MSHT, social aspects of stakeholder characteristics and
collaboration incentives, as well as business aspects of economic efficiency and subsidization of treatment
costs within the healthcare system to be of great importance. For this reason, we differ from existing
contributions that focus on single paradigms and instead follow this new approach to digital platform
ecosystems.
The conceptual work of Adner (2017), Kapoor (2018) and Jacobides (2018) extends platform-centric views
by broader and more conceptual definitions of the ecosystem construct. Adner (2017) views ecosystems as
a combination of “ecosystems as a structure” and “ecosystem as an affiliation” approach. So a digital health
ecosystem encompasses several healthcare actors that perform interdependent activities. According to
Kapoor (2018) “an ecosystem encompasses a set of actors that contribute to the focal offer’s user value
proposition”, which refers to the collaboration of several healthcare stakeholders to create patient value.
Like in biological ecosystems (with the term ecosystem originating from (Moore 1993)) there is an evolution
over time that influences the ecosystem’s members in regards to their collaboration, innovation and
competition. This somewhat mirrors the historical development of a healthcare system which then still
heavily influences the way it is working today. The product or service offer can be designed with (or even
without) a technological architecture that is based on a platform (Kapoor 2018); in this context, the digital
health ecosystem can also be designed in a decentralized way. Jacobides (2018) focuses on the types of
complementarities and resulting mechanisms and defines ecosystems as “a set of actors with varying
degrees of multi-lateral, non-generic complementarities that are not fully hierarchically controlled”
(Jacobides et al. 2018).
While these previous contributions have highlighted important aspects of ecosystems from a platform-
centric and multi-lateral coordination point of view, there has been limited attention paid to the context in
which these ecosystems do and do not emerge. One important context condition for the integration of
MSHT is the high regulation on the healthcare market that is, for instance, caused by medical device
certification guidelines and data security standards. Secondly, there are social policy objectives (e.g.
high quality of care) that meet economic policy objectives (e.g. promotion of entrepreneurial activity)
that are often contradictory (Saltman et al. 2002). These contradictions and context conditions create an
interesting opportunity to study the (non-)formation of ecosystems in situ and to understand the factors
that hinder or promote the creation (and simultaneous non-creation) of ties between ecosystem actors.
For developing our own perspective on ecosystem design and governance strategy, Jacobides et al. (2018)
provide a detailed description of different governance and regulation mechanisms. Digital ecosystem
success and the behavior in it are influenced by the rules of engagement, as well as the nature of
interfaces and standards which include open-versus-closed and imposed-versus-emergent ecosystems.
Standards within an ecosystem can either be proprietary or sector-wide and are defined either (a) de facto,
especially if they are not based on technology or (b) de jure, especially if there are many ecosystem
members. For each, there is a certain degree of ecosystem membership control e.g. by a central hub. The
rules for membership within the ecosystem may vary over time which also relates to the modularity and
nature of complementarities within an ecosystem.
Governance is closely related to decision-right allocations (who is responsible for ‘what’) as well as
the enforcement of desirable behavior, which in this context can be called control. This is complicated by
the absence of the authority structure of a central actor (Jacobides et al. 2018). Extending upon the notion
of Jacobides et al. (2018) and borrowing from Tiwana et al. (2010), we refer to control in an ecosystem
context as the formal and informal mechanisms implemented by an ecosystem firm to enco urage
desirable behaviors among other ecosystem participants. Formal mechanisms can relate to output or
process control, controlling what other firms produce (or use) and which processes they must follow
(Tiwana et al. 2010). Informal control refers to social pressure, as well as the development of shared norms
and values which are imposed on or emerge from ecosystem members. One of the most important
governance mechanisms are boundary resources (Schreieck et al. 2016) which consider technological as
well as social aspects of platform ecosystems (Eaton et al. 2015). Data that is provided by platform users
and can be accessed by complementors (Gawer 2014) is a boundary resource that is gaining importance in

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
5
practice. This is especially relevant in the MSHT context where huge amounts of health data are collected
that need to be made accessible to complementors and interoperable within all ecosystem members to
facilitate (co-)creation of value.
Closely related to the governance of an ecosystem is its openness. While openness has been discussed in
different domains regarding different objects (Arora et al. 2016; Baldwin and Hippel 2011; Lessig 2001), it
has also been picked up by Jacobides et al. (2018) in their agenda for future research on ecosystems. We
focus on data openness as one important concept that is subject to different views and opinions. In line with
Lessig (2001), we define data as open if its use is “free,” meaning that one can use it without permission
from anyone else or if the permission one needs is granted neutrally. This does not mean that openness
implies sharing without costs, but it implies non-discriminatory access to data, such as when it is governed
by pre-defined neutral licenses. Managing such data sharing requires making design and governance
decisions that maintain the tradeoff between promoting generativity to facilitate complementors’
contributions and retaining control to prevent undesirable platform use (Vesselkov et al. 2019).
Generativity can be defined in this context as the ability to spark unbounded growth, facilitated by large,
uncoordinated audiences (Zittrain 2008, p. 70). The seeming tension that arises between benefits from
open data sharing on the one hand, and conflicting interests to do so on the other hand, has been called the
“paradox of openness” (Arora et al. 2016), and recent research has called for new perspectives on this
paradox in the healthcare context, which our contribution sets out to do using the context of MSHT and its
integration into emerging digital health ecosystems.
Any description of these governance mechanisms would be incomplete without mentioning the
technological preconditions enabling the emergence of ecosystems as well as their evolution in the first
place. Jacobides et al. (2018) note that one important prerequisite is modularity, the decomposition of a
system into smaller components that one can “mix-and-match” relatively easily (Schilling 2000). As Zuboff
(2019) insightfully notes, modularity is a circle of behavioral data created from “users,” analytics, and
service improvements, enabling the creation of surplus from rendered behavior. This, in turn, can be
used to create new data-driven business models balancing the tension of openness and control, thus
designing specific distributions of value claims, which are more “unilateral” (favoring the one-sided
monetization of prediction-based insights by private firms) or more “bi- or multi-lateral” (favoring the
distribution according to universally accepted and agreed societal standards). What is interesting in the
MSHT ecosystem is that stakeholders can take the role of data producer and of consumer, so that they
become data prosumers (Vesselkov et al. 2019). Similarly, the monetization of data and developing
prediction-based business models is clearly limited, making this an interesting study context. The main
tenants of our perspective form a stakeholder-differentiated view on incentives and disincentives for data
production, sharing, and usage, as well as a simultaneously social, technical and economic view on
ecosystem emergence and evolution. Highlighting this, figure 1 synthesizes the different views into our own
perspective on MSHT integration into the healthcare ecosystem, which the remainder of this paper sets out
to explore and deepen.
Methodology
In the following we explain the methodological approach of our qualitative interview study. The motivation
for this paper is to investigate the phenomenon of mobile sensor-based health technology integration from
four different stakeholder perspectives within the context of digital health ecosystems (Myers and Avison
2002). The benefit of a qualitative research approach is that the cultural and social context in which
decisions take place can be apprehended well (Benbasat et al. 1987). In the “natural context” of the
Integration of
technology innovation
Mobile sensor-based
health technology
User as “data producer”
and “data consumer”
Data analytics
Improvement of services
Healthcare system
context
High regulation
Social policy objectives
Economic policy objectives
Ecosystem design &
governance
Strategic direction
Rules of engagement
Decision-rights & control
(Data) openness
Figure 1. Core Constructs to Be Explored and Deepened in Research Study

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
6
healthcare system this is particularly relevant because personal experiences and legal requirements often
highly influence stakeholders’ decisions (Myers 2019). The best way to understand stakeholders’ actions
and motivations as well as the context in which they take place is by talking with people (Myers 2019, p. 5).
For this reason, we chose expert interviews that are semi-structured and guideline based to enable an in
depth-review. The phenomenon under study is relatively new, such that across (and even within)
stakeholder groups it is likely that different terminologies and phrases are emerging. For this reason, the
underlying philosophical assumption of our research is interpretive (Klein and Myers 1999) since we need
to interpret these meanings to be able to grasp the respective phenomenon.
While designing the semi-structured interview guideline for data collection, three aspects were given
special attention (Baur and Blasius 2014, p. 567): to avoid sudden changes in topic in order to establish a
narrative flow, to give the interviewee enough time to speak by providing a clear structure and a limited
number of questions and to encourage the interviewee to narrate freely. During the preparation of the
interview guideline we followed the four steps that were introduced by Cornelia Helfferich (2011) which
resulted in 15 questions. These are grouped under six
headings (Introduction, MSHT characteristics, Health data
integration, Technology integration, Process integration
and Future development) to enhance the structure of the
interview. There was a pre-test of the interview guideline to
assure a clear wording and an overall common thread. For
every stakeholder group, the detailed interview guideline
was adapted slightly to match the related role or field (e.g.
company/hospital). At the beginning of the interview there
was a brief introduction of the researcher as well as the
research project to clearly state the purpose of the interview
(Myers 2019, p. 133). The concept of “mirroring” was
applied during the conversation (Myers and Newman
2007). This means that phrases and words employed by the
interviewee are subsequently used by the interviewer to
phrase their following questions or comments. At the end,
there was the opportunity to ask further questions and then
the interviewee would be thanked for their provided
insights (Myers 2019, p. 133). All interviews were conducted
in the German language by one researcher.
The initial case selection is inspired by the European Connected Health Alliance Ecosystem (European
Connected Health Alliance 2019) that identifies several stakeholder groups within its ecosystem (see figure
2). In total four stakeholder groups have been included within our study because they represent the
phenomenon of MSHT integration (Corbin and Strauss 1990) and have a high accessibility. The first (1)
stakeholder group included are companies which are represented by mobile health startups (MHS).
Startups are conceptualized as “young, growth-oriented firms that engage in innovative behavior” whose
growth rate can be higher than that of mature companies (DeSantola and Gulati 2017). Mobile sensor-based
health technologies are innovative products/services and there are no prominent large corporations, but
rather small businesses operating in the healthcare sector targeting the end user. These businesses try to
enter the market with a new digital technology they try to integrate into the existing healthcare system. The
second (2) stakeholder group is represented by statutory health insurances (SHI) because they pay the
service providers for the treatment of patients while respecting the legal guidelines for reimbursement. The
medical service providers (MSP) are the third (3) stakeholder group that is included in our review. This
comprises general physicians, medical care centers and hospitals which have to choose the information
technology that is included in medical service provision (and at the same time act as a “business” which
compares costs and revenues). As a fourth (4) stakeholder group we also decided to include institutes &
incubators (I&I). They act as supporting stakeholders within the healthcare system and advise, consult and
conduct research within healthcare. Therefore, they provide a holistic view of the overall healthcare system.
To enable a triangulation of subjects (Rubin and Rubin 2005, p. 67) and counteract elite bias (Miles and
Huberman 1994; Myers 2019) we aim to include a variety of experts on different hierarchical levels within
each stakeholder group that represent a variety of perspectives. In general, only stakeholders that are active
in the German healthcare market are considered to ensure the comparability of the legal framework.
(3) Health &
social care
(Providers,
professionals)
(1) Companies
(Startups, large
corporates)
Policy makers
(2) Payers
(Statutory &
private health
insurances)
Third sector
Patients
&
Citizens
(4) Advisors
(Institutes,
incubators,
research)
Figure 2. Stakeholder Map

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
7
Overall, experts were defined as persons that are either German or English speaking employees, have at
least one year of working experience (or in the case of very young companies, that have been working there
from the beginning) and either have a technical or a medical background. Potential interviewees were
contacted directly via mail or LinkedIn (# interview requests = 89). A brief description of the research
project and its overall objectives was attached in the initial message to make sure the interviewee felt
confident enough to answer the interview questions. The interviews (n=30) took place between October
2019 and April 2020 and were either conducted in person (1/30), via video conference software (2/30) or
via telephone (27/30). An overview of the interviewed stakeholders is depicted in figure 3. The interviewees
have an average experience of 7,2 years within their field. This comprises the shortest experience of 9
months and the longest experience of 40 years within the medical domain. After the interviewees gave their
informed consent (Payne and Payne 2004, p. 68), all interviews were recorded using the Philips DVT2710
dictating machine. The average interview duration was 46 minutes, which includes the shortest interview
of 32 minutes and the longest interview with a duration of 72 minutes. To enable the qualitative analysis of
data each audio file was transcribed.
The data analysis of interview transcriptions started parallel with the collection of further interview data.
We paid special attention to the representativeness of concepts in the choice of interview partners (Corbin
and Strauss 1990). To allow for an inductive development of theory that is based on empirical data, we
chose the Grounded Theory approach (Martin and Turner 1986). Grounded Theory is frequently used in
information systems research “to study technological change and social technical behavior in emerging
research domains” (Wiesche et al. 2017). This resonates with our objective to study the integration of
constantly evolving innovation (MSHT) into a digital healthcare ecosystem that depends on the
collaboration and acceptance of various stakeholders. Additionally, Grounded Theory offers a high degree
of flexibility (Birks et al. 2013) and there are various methodological approaches within the existing
literature. We chose the Straussian approach because it provides a frame for students that prefer to work
with a preset structure (Mey and Mruck 2010) as well as an “unambiguous process guidance” (Wiesche et
al. 2017, p. 689). The paradigm for coding should be consistent with the research question (Mey and Mruck
2010) as well as with the researchers position (Birks et al. 2013). Initially, we consider the a priori chosen
coding paradigm of the Straussian approach (Corbin and Strauss 1990), which are conditions, context,
strategies and consequences to fit our data well (Birks et al. 2013). The coding of data is performed
according to the three steps proposed by Strauss & Corbin (1990). To support the process of qualitative data
analysis Atlas.ti (v.8 for Windows) software was used as well as memos to keep track of our hypotheses and
questions throughout the coding process. In the first step of open coding, a sentence-by-sentence analysis
was performed by assigning concepts to the text fragments of the interview transcripts and continuously
comparing them with one another (Corbin and Strauss 1990). In order to illustrate the process of open
coding, the text “[…] but the entire mobile data systems with the platform behind them and medical
applications, they must be reimbursable for companies like us, otherwise we will be dependent on investors
for a long time to come. Germany in this area is financed by investors and German investors are not so
willing to take risks anyway.” was coded as follows: reimbursement conditions, access to financing, risk
tolerance. After open coding (# total codes = 239), the codes were checked for duplicates and (especially
(1) Mobile health
startups
n = 10
(2) Health insurances
n= 7
(3) Medical service
providers
n = 6
(4) Institutes &
incubators
n = 7
Statutory health
insurances (7) which
comprise five different
public corporations, in
two cases there were
independent interviews
(n = 2) with employees of
the same health
insurance that work in
different departments
Independent institutes
that conduct research and
support actors within the
healthcare system (5) and
incubators that assist and
advice mobile health
startups (2)
Providers of medical
services with different
company characteristics:
physicians with own
practice (3), medical care
center (1), privately-held
hospital (1), publicly-held
hospital (1)
Application areas:
stress management (1),
sleep monitoring (1),
health risk analytics/
insights (3), patient
monitoring (2), mental
health (1), rehabilitation
monitoring (1), diabetes
management (1)
Sensor types:
smartphone sensor (2),
individual sensor (7),
commercial sensor (1)
Figure 3. Interviewee Numbers and Affiliations

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
8
those with rare occurrence) reviewed (# total codes = 204). If possible, this also includes merging
semantically similar codes e.g. hacking and information security (# total codes = 181). Next, to form
categories and sub-categories, conceptually similar codes were grouped (# total concepts = 107) e.g. human
intervention and replacement of humans both describe the concept of the level of human involvement. In
the second step of axial coding, relationships that are grounded in the data are assigned between categories
and their subcategories. This is done by following the coding paradigm of conditions, context, strategies
(action/interaction) and consequences (Corbin and Strauss 1990). We deviated slightly from the original
paradigms within the coding process because we felt that the theory development would be enhanced by
renaming some of the categories (Mey and Mruck 2010). We therefore merged context and conditions to
context conditions (because within the phenomenon under study the conditions arise from the context) and
customized consequences to outcomes. In the final step of selective coding, all identified categories are
unified around a core category (integration of mobile sensor-based health technology), which has the most
relationships to the remaining categories and represents the central phenomenon of the study (Corbin and
Strauss 1990).
Integration of Mobile Sensor-based Health Technology into a Digital
Health Ecosystem
We turn to the results of our analysis regarding the integration of mobile sensor-based health technology
into digital health ecosystems. Table 1 displays 107 concepts and their number of appearances in brackets
(# appearances) as a result of all three coding steps. Additionally, their assignment to the 18 sub-categories
(right table column in italics) and the four main categories (bold table captions) that are included in the
final conceptual model are displayed. Direct quotes from the interviews are presented in quotation marks
and the stakeholder group of the interviewee is indicated in brackets. The over-arching categories that
emerged were context conditions, the integration of mobile sensor-based health technology, digital health
ecosystem design and governance strategy, and outcomes. The following sections detail these findings and
point to apparent tensions and contradictions.
We define the first category context conditions as the overall factors and requirements for the integration
of MSHT into the healthcare system. Society’s mindset describes the overall attitude of the public for
instance if they are willing to track personal health data. The healthcare system illustrates special
characteristics within the (German) healthcare system e.g. the status quo of the system “[…] more in the
sense of a preventive system and not in the sense of a repair system. And that we get away from this shallow
medicine that we are making now” (MSP), if there are specific reimbursement conditions and system
evolution over time, for instance “Then I always try to explain to them that these [traditional measurement
devices] were not created because one was clever at the time, but because it developed historically.” (MSP).
System agility describes the speed and flexibility of actions, e.g. “and [in other countries] it makes me feel
like things are getting back to the hospital bed faster, into use. In Germany there is still a lot of bureaucracy
and forms.” (MSP). Data regulation includes data protection of the user against privacy impairments
through e.g. unauthorized data access and information security which refers to the characteristics of
technical and non-technical systems that store and process information. In this context, the right of the
user to data deletion might collide with medical data archiving obligations. Quality requirements
represent the need for medical certification if sensor devices are used for a medical purpose. Depending on
the certification procedure and certification classes, the effort can be very high regarding costs and time.
Next to this, the benefit of the application needs to be proven: “so you really have to have determined the
benefit of this application in care […].” (I&I).
Integration of mobile sensor-based health technology illustrates the core phenomenon of the
conceptual model. The user of the technology includes everybody that is affected by its integration e.g.
patients, physicians, caregivers, etc. Every user has a self-image and some intrinsic motivation to use the
technology which can also depend on technology affinity. The application of the technology is location
independent and “the usability must of course be designed for this use case, e.g. the app and also the devices
themselves must be built so that they work as intuitively as possible.” (MHS). The sensor device is low -
threshold and can be easily integrated in the user’s everyday life e.g. “But how can I make it so easy for the
user that s/he has no additional effort at all and therefore just uses it?” (I&I). The technology that is used
includes a mobile sensor and is therefore able to monitor user data continuously, often in combination with
an app or smartphone. Depending on the maturity of the sensor technology, the devices have a certain

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
9
degree of error-proneness. Each sensor device provides some kind of data acquisition, for instance “we
record movement data 24 hours a day” (MHS) with a specified data accuracy. Data diversity describes the
collection of many different types of data e.g. “when certain patterns come together in patients, which can
also be sensor data, but also laboratory data, diagnosis constellations, etc., then it becomes interesting.”
(MSP). After health data is captured it needs to “flow” between different stakeholder groups to enable their
collaboration. This data flow includes the import of data into a stakeholder’s systems or devices, for
instance “The physician in turn, when s/he receives the measurement results, can import them into his/her
practice EDP system […].” (MHS). The imported data must be stored so e.g. “We store 99% of the data on
the server, which is then retrieved by the app on a regular basis, but a few things are stored locally, of course,
to make it fast.” (MHS) is a possible solution for data retention. If data from different sources is imported,
the consolidation of this data is necessary, for instance “[…] if I can really import, merge, perhaps with other
data from the medical field, then I can of course do much, much more exciting analysis.” (SHI). There is
also the possibility for data export that is often provided by generating pdf- or csv-files. For import as well
as export of data, data transmission is essential as well e.g. “so the sensor on the arm joint is connected to
the smartphone app via Bluetooth. And the app is then connected to our servers via mobile phone network.”
(MHS). In contrast to that data sharing is independent of the technological basis but rather focuses on the
stakeholders that are involved, for instance “[…] transferred to a server to ensure access for doctors,
relatives, patients etc.” (MHS). Within data processing, there is medically relevant and valid health data
that facilitates data analysis e.g. “and there [on our servers] the calculation runs with our models, which
we are currently training, also with Machine Learning.” (MHS). Depending on the amount of data there
are different means of data analysis e.g. “So I'm a friend of big data pools and I'm also a friend of big data
analytics and deep learning systems.” (MHS). After data processing, which describes the types and means
of analysis, data interpretation (which can also be incorrect) describes the derivation of instructions for
actions, user feedback, etc. Large data amounts also enable predictions within the healthcare system, for
instance “And especially in the area of predictions, this is very, very meaningful data […].” (SHI). Another
important feature is data visualization e.g. “The added value is created by making the activity profile visible,
through the companion app […].” (MHS) and the way it is implemented: “A graph that goes from very fit
very slowly to increasingly sick, that's what makes a difference.” (MHS).
Digital health ecosystem design and governance strategy describes (inter)actions of the
stakeholder groups that influence each other as well as the entire ecosystem. Policy control describes legal
guidelines within the ecosystem. To define guidelines, the necessary knowledge and experience within the
respective area are essential (which may still need to be acquired). Someone needs to be responsible for
compliance control of these guidelines, but also to define them in the first place, e.g. “until the legal
framework for [...] is established and how concretely the whole thing is defined. It is difficult to combine all
sorts of things without a clear definition.” (I&I). Every stakeholder can interpret the guidelines in a different
way, with one interviewee saying, “Now comes the joke: but the interface is interpreted a little differently
by each company, so it's actually not a standard. You always have to tinker with it a bit.” (MHS). If there is
a system transition within healthcare from an analog to a digital ecosystem, then this process needs to be
managed carefully. Data openness describes the interoperability of data within the system which requires
semantic, as well as syntactic standards, that enable the “flow” of the data between stakeholder groups. If a
stakeholder possesses the data sovereignty, s/he has the right to decide who can e.g. access it. Also,
application programming interfaces (APIs) can enable the interoperability of systems even if there are
different data formats. If the data format of a sensor is not compatible with other applications, there can be
a lock-in. All data that is captured needs to be collected at a certain point which is recognized by every
stakeholder (e.g. electronic medical records). We learned from one interviewee that “[…] what possibilities
arise from these patient files to really be able to integrate such data and make it available to users is
enormous.” (SHI). Community engagement describes the commitment of the stakeholders within the
ecosystem and what incentivizes them to participate. For the stakeholders to collaborate there needs to be
trust amongst each other e.g. “We cannot afford to be dependent on interfaces that some manufacturer
supplies and possibly delivers uncertainly […].” (MHS) or trust in the technology, for instance “We trust
that this data is accurate enough.” (MHS). To accept new technologies e.g. sensor devices, stakeholders
must be open to change, “[…] you have to open yourself again internally and say ok, do you really still need
it?” (MSP). Every stakeholder has a defined role within the system that can change over time. Additionally,
an introduction and explication of innovation is necessary e.g. “[…] and forget to bring in the person who
will somehow prescribe, use or explain it to the users every day.” (SHI). Within the system, stakeholders
can assume responsibility e.g. for a certain role. “But if s/he makes it available to the health insurance and

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
10
receives a reimbursement, e.g. s/he gets money from the community or something is paid, then I think I
can give data back to the community.” (I&I) describes the (necessary) community spirit or the feeling of
belonging within the ecosystem. To be able to succeed within an ecosystem and process transactions,
stakeholders need to be able to orientate themselves and navigate through the ecosystem. The product-
market strategy describes the product/service design of stakeholders within the ecosystem. Monetization
illustrates if and how stakeholders receive a certain amount of money, as explained by one interviewee “[…]
it must also be economically viable, it must be accountable, the cash flows must be mapped.” (MHS). The
access to customers describes how product/service offers reach potential customers, which also depends
on the target group. The market access represents how stakeholders enter the healthcare market. In this
context access to financing is also particularly important for mobile health startups, “because you are
normally always externally funded and have to invest some development time to bring the product to
market.” (MHS). To cooperate with other stakeholders, the alignment of their strategies is of importance so
“that we have entered into cooperation with startups and that we have said that what they have, the app or
whatever, fits in with our focus […].” (SHI). The business models within the fitness and lifestyle market
differ from those in a medical context.
Context conditions
Public perception (26), Self-tracking need (3), Willingness to optimize (2)
Society’s mindset
Reimbursement conditions (63), System agility (36), Status quo (23), Country-specific
differences (17), Public law (15), System evolution (12), Self-organization (5), Solidarity
principle (5)
Healthcare system
Data protection (60), Information security (41), Data deletion (28), Data archiving (12)
Data regulation
Proof of benefit (34), Certification classes (22), Certification procedure (21), Certification effort
(20), Quality label (7)
Quality
certification
Integration of mobile sensor-based health technology
Self-image (35), Usage motivation (31), Health & data literacy (18), Technology affinity (11)
User
Simplicity in use (43), Use case specific (39), Location independence (31), Low-threshold
application (18)
Application
usability
Mobile sensor (100), Error-proneness (20), Technological maturity (3)
Technology
Type of data collection (110), Data accuracy (60), Continuous monitoring (29), Data diversity
(17)
Data acquisition
Data retention (57), Data sharing (38), Data transmission (37), Data export (20), Data
consolidation (13), Data import (10)
Data flow
Data analysis (34), Data validity (20), Amount of data (16)
Data processing
Data visualization (44), Data interpretation (43), Prediction (22), Data based decision (16)
Data
interpretation
Digital health ecosystem design and governance strategy
Legal guidelines (63), Compliance control (26), Definition of requirements (24), Assignment of
responsibility (15), Necessary knowledge/experience (14), Design sovereignty (7),
Interpretation of requirements (5), System transition (4)
Policy control
Data sovereignty (87), Semantic & syntactic interoperability standards (80), Data collection
point (51), (Open source) APIs (45), Lock-in (9), Data harmonization (3)
Data openness
Stakeholder (participation) incentives (43), Trust (41), Stakeholder collaboration (28),
Stakeholder acceptance (26), Openness to change (21), Role definition (19), Assumption of
responsibility (17), New actors (17), Introduction & explication of innovation (13), Community
spirit (6), Orientation within the system (6)
Community
engagement
Product/service design (46), Access to customers (42), Monetization (41), Access to financing
(25), Market power (19), Boundary between fitness and medical devices (18), Target group (16),
Market access (16), Strategic alignment (15)
Product-market
strategy

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
11
Outcomes represent possible consequences that the (non-)implementation of the strategies can cause at
different levels of the (digital) health ecosystem. Shared system values describe a common understanding
of values within the entire ecosystem, which includes a certain level of agreement from (all) stakeholders
in regards to specific topics. It also describes how risks and uncertainties are handled within the ecosystem.
There needs to be a protection against misuse, to which all stakeholders adhere to and prevent that “[…]
the data is simply used in a discriminatory manner against the user.” (SHI). Additionally, decision rules
that are clear to everybody have to be implemented, especially if the opportunities outweigh the risks of a
new technology. Transparency within the ecosystem can facilitate a paradigm shift that fundamentally
changes the status quo. Big data algorithms and pattern recognition can be used to conduct research for
the common good, for instance “You gain new knowledge about clinical pictures and the course of the
disease. By comparing with other, perhaps anonymized, data.” (SHI). The nature of stakeholder
participation in the ecosystem can either be on a voluntary or obligatory basis. System efficiency illustrates
the modification of the existing processes within healthcare which can cause changes in time and changes
in cost expenditure. Also the level of human involvement can be modified within medical service provision
e.g. “And with us in the program, it's not a physician right now, it's our virtual coach […].” (MHS). The
stakeholder burden is influenced and either results in relief or overstrain, which can also be connected to
high complexity, for instance “So I see this lack of clarity and this flooding and overburdening of health
applications […].” (MHS). Further, the healthcare quality (of service provision) for the patient can change,
which would then include the definition of “quality” or “value” within value based care. Large amounts of
continuously collected health data enable personalized medicine and might lead to a behavior change of
the user, for instance “I measure it, I make myself aware of it and that's why I change my lifestyle.” (SHI).
User empowerment enables a focus on prevention – “We are very much in the prevention area and it is
about enabling patients or insured persons to better manage themselves from the outset.” (SHI), which
finally impacts the quality of life.
Discussion
We turn to discussing how we can conceptualize incentives and disincentives as well as barriers to data
openness and sharing. This becomes relevant when integrating MSHT into digital health ecosystems. Figure
4 shows our conceptual model and illustrates the relations between the four main categories and 18 sub-
categories. Relations between main categories are depicted by bold grey arrows and relations between sub-
categories are represented by small black arrows. The four main categories are context conditions, the core
phenomenon in form of a health service improvement cycle, digital health ecosystem design and
governance strategy and outcomes.
Barriers to Entry
Before a mobile health startup enters the ecosystem, we identified that it faces obliging to strict guidelines
(formal controls) and other contextual prerequisites. The perception of society is influenced by data
regulation especially connected to the healthcare system, for instance, that data leaks have a deterrent
effect on the overall trust in MSHT. Moreover, quality certification which describes medical device
certification guidelines, is influenced by data regulation (because respective requirements are included in
the guidelines but are also applied in other contexts) as well as by the context of the healthcare system. For
instance, in the fitness and lifestyle sector it is not mandatory. Medical device regulation is necessary if the
device/software is used for a medical purpose: “The approval as a medical device is the prerequisite for me
Outcomes
Transparency (33), Ethics (24), Protection against misuse (21), Modernization/digitization level
of the system (19), Research for common good (17), Decision rules (13), Nature of participation
(10), Risk tolerance (10), Paradigm shift (8), Sustainability (4), Level of agreement (3)
Shared
value system
Process modification (67), Level of human involvement (52), Changes in time expenditure (48),
Changes in cost expenditure (37), Stakeholder burden (25), Complexity (15)
System efficiency
Quality change in care (70), Prevention (42), User empowerment (33), Behaviour change (24),
Quality of life (17), Personalized medicine (10), Value-based care (10)
Healthcare quality
Table 1. Results of Grounded Theory Coding Process

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
12
to be allowed to market it at all.” (MSH). If they are certified it is still complicated to enter the healthcare
ecosystem “which has been a very, really closed shop so far, and is very narrowly limited.” (SHI). Up until
now the conditions for reimbursement are defined in selective contracts with SHI according to the
insurances’ requirements. In the future, the reimbursement of MHS products/services in Germany will be
controlled by a recently introduced law (Digitale Versorgung Gesetz) which also includes strict guidelines,
as well as cost and duration efforts. We learned from one interviewee that “[…] these certification processes
take so much time that even good solutions are bled dry financially before they hit the streets.” (I&I). These
are the legal guidelines that were only recently passed, but the MHS business models highly depend on
these changes. For this reason, they have to be flexible and agile to adapt to changing conditions in the
future. Within the quality certification, the MHS disclose their algorithms, data accuracy and service design
so the “proof of benefit” can be evaluated by medical device certification authorities. For example, “[…]
software products for digital work, server portals, mobile apps etc., algorithms must be certified as medical
devices […] as a manufacturer you have to guarantee that digital data – which can have a therapeutic effect
– must be a medical device […].” (MHS). This elaborate process creates high entry barriers for the
ecosystem.
Health Service Improvement Cycle and the Challenge of Data Openness
Once a mobile health startup enters the market, we find the core phenomenon to be depicted by the data
cycle that is enabled by MSHT. The user (data producer) applies sensor technology that captures diverse
health data in a continuous way, such as one’s heart rate. The easier the technology is to use, the more often
the user will employ it and a higher wear time results in the collection of a larger amount of (continuous)
health data. For advanced calculations, in most cases the data is transmitted to a (central) storage and
analysis location. Within this data collection point, there is a consolidation with other health data. To offer
added value for the user, an interpretation of their health data is necessary (through the data consumer, for
instance MHS) which is simultaneously seen as improving the service offer e.g. “your ECG data indicate
atrial fibrillation”. To communicate the results, a clear and graphical presentation of (aggregated) health
data is essential, which is often depicted on an (external and larger) screen. With the amount of accessible
health data, the possibilities for algorithmic and interpretative improvements increase, while the likelihood
for error is reduced. This allows for an additional product/service improvement for the user which will
result in a higher adoption rate and in turn to more availability of users’ health data. Also, MSP can be data
producers, for instance, by adding captured patient health data to the central data collection point. MSP
can also be data consumers, for instance, by looking at independently collected patient data for treatment
improvement. Additionally, SHI can take the role of a data consumer to improve their service for insured
persons e.g. offering individual recommendations. While the technology supported acquisition of data, as
well as its processing and interpretation, are carried out by one stakeholder group, the flow of the data and
its consolidation is happening on a collective ecosystem level. The more data producers participate, the
larger the “data treasure” becomes that can be used to facilitate service improvements. There are still strict
Core phenomenon: Integration of mobile
sensor-based health technology
User
Policy
control
Digital health ecosystem design
and governance strategy
Community
engagement
Product-
market
strategy
Shared
value system
Outcomes
Healthcare
quality
System
efficiency
Quality
certification
Context conditions
Data
regulation
Society’s
mindset
Healthcare
system
Collective
ecosystem
level
Ecosystem-stake-
holder-group level
Individual
user level
Data flow
Data
interpretation
Technology
Data
acquisition
Application
usability
Data
processing
Data
openness
Figure 4. Conceptual Model of MSHT Integration

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
13
policies regarding data protection, but in the context of health data standards, whose interoperability is
essential to generate digital health innovation, there is no clear definition of a uniform standard (formal
control), nor of an informal control through data standard agreements between (all) ecosystem members.
For this reason, low data standard (formal and informal) control causes low data openness. In this stage,
MHS are active on the healthcare market and (partly) reimbursed by healthcare insurances. There is, for
one, “the issue of interoperability […]. Standards are a very big topic, which plays a role here.” (SHI). At the
moment there is no uniform data standard to adhere to. It is opined that, “If only you could agree on such
a standard and everyone would adhere to it really seriously. In this respect, we always create something
new for everyone.” (MHS). Even if two stakeholders use the same standard or interface, there are often
differences between them due to the interpretation of guidelines and the consequent creation of individual
solutions e.g. “Theoretically, yes, but practically, there is nothing. Everyone has an individual solution and
everyone has the problem […] and then everybody builds an own standard on the standard.” (MHS). As a
consequence, we find that there is low data openness within the healthcare ecosystem. The missing
definition of a data standard and compliance control causes interoperability issues, in that “[…] if I have
any standard, then I can work with it, then I can transfer it to other standards, so that is much more
important than anything else. It must be standardized. It must be a format that works across everything.”
(I&I).
In this context, “the business model is important to us, that the digital health service is not designed to
exploit or sell the data.” (SHI). The predominant type of monetization for MHS is to be reimbursed by
(statutory) health insurances, for instance “B2C, which is difficult in Germany because we have a low self-
payer willingness, and B2B2C, e.g. about the refund […].” (I&I). In related fields where companies collect
large amounts of data through products/services for which there is no direct willingness to pay (like MHS
do), there are often business models enabled by a surplus of behavioral data (Zuboff 2019). Within
healthcare it is strictly forbidden to use data for other purposes than those they were originally collected
for. For this reason, MHS cannot receive compensation from third parties for their selling of data. If they
would do so, this would likely be accompanied by a loss in trust, e.g. “to make people feel, yes it is a solidarity
system, that is why you get the therapy, but that is why you are not part of any bigger business models that
you do not know anything about or where you cannot defend yourself against.” (I&I). This illustrates the
difficult process for MHS to develop sustainable business models within the German healthcare system.
This is also a challenge in terms of service improvement through big data analysis. While the demand for
high control of ecosystem access is reasonable (e.g. for MHS) in the sensitive context of the healthcare
system, the low data openness within this ecosystem tends to be caused by an absence of control, even
though there are also calls for action. This is highlighted in rationals such as, “[…] ok all other countries do
it, we just have to agree on it and participate. It is not as if it were somehow impossible to define it and to
make progress on it.” (I&I). While one challenge in the context of a uniform data standard within healthcare
is its definition, there are also different needs regarding data openness within the system or the “paradox
of openness”. It illustrates that incentives for openness of data differ depending on stakeholder and
ecosystem level (Nambisan et al. 2019). To reveal potential conflicts across levels, we chose to explain the
incentives for high and low openness on three different levels. The collective ecosystem level describes the
overall environment of the digital health ecosystem encompassing all members. Within the ecosystem-
stakeholder-group level, for instance, activities that are performed by one stakeholder group, such as data
interpretation by MSP, are described. The individual user level includes categories that change depending
on individual users who can be members of the same stakeholder group e.g. different patients prefer
different MSHT usability characteristics.
Collective Ecosystem Level. With the motivation to conduct research for the common good e.g. to
enhance the understanding of disease symptoms, there need to be large amounts of health data, so “one
acquires new knowledge about disease patterns or progression. By comparing with other, perhaps
anonymous data, for example, how are others dealing with these diseases […].” (SHI). This enables an
earlier intervention for other patients, so “that, if you go back to the beginning of the disease, this data can
help others who may be showing symptoms.” (I&I). In order to compare and analyze such data, we must
use the same data standard. As one interviewee aptly stated, “Of course, if I want to compare data now, I
need to have them in the same format in the best case, so that I can run an AI over them […].” (SHI). If
there is more data, then there is also the possibility for different compensation models, such that, “If you
look at compensation, for example, we always had the discussion, wouldn't it need quality-based
compensation? Where you said that is totally difficult, because you would need good data for quality -

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
14
oriented compensation. You haven't had that before and you may have it with the digital health applications
in the future because you have completely different data availability, for example.” (SHI). In this context
there is the concept of value-based care where existing incentive systems can be modified so that the patient
benefits from high quality care. Still, a standard for “value” within care needs to be defined, which may be
even more difficult than the definition of a data standard. Thus, overall, there are mainly benefits in the
context of high data openness on the collective ecosystem level.
Ecosystem-Stakeholder-Group Level. Within the three stakeholder groups that are actively involved
in data openness (I&I are excluded), the incentives differ. Mobile health startups’ product/service offerings
have to be applied by medical professionals in a simple and intuitive way, otherwise there is no wide
adoption. As one interviewee suggested “[…] we can connect with other applications, let’s say create a
certain standard of interfaces […] to adapt these processes in care in such a way that it does not mean more
work for the carers, but actually makes things easier.” (MHS). Overall, if all data were to be compatible, the
sensor devices might become a generic complement (Jacobides et al. 2018) within the ecosystem and the
added value would be created by an app (for example) that analyzes the data which can be imported from
any source. In this case, there is more competition between MHS that offer services for similar disease
patterns e.g. diabetes management. If every application can import the data, a company has to differentiate
itself from other companies. Currently, SHI have the power to decide which product/service is reimbursed,
which is one of the only business models for MHS within healthcare at the moment. For MHS, data
openness creates another approach to monetization and sustainability, as well as value-based business
models, “so there would just have to be a business model where you can earn money, for example, by
organizing care better, more efficiently, etc., by providing better individual, data-supported care than with
collective care.” (SHI). It would appear that MHS are also interested in data openness, as long as their data
is not used by competitors.
At the moment, statutory health insurances are not allowed to access or own any health data of the user
e.g. collected by MHS. Even though public opinion varies and users are afraid that SHI possess health data
via one’s usage of an app that is reimbursed. We were told that “most [patients] think we already have data
[…]. That is actually the perception of most of the insured with whom we speak […]. However, we have
nothing.” (SHI). If SHI possess continuous sensor data from a user, s/he might be worried that insurance
fees are adapted according to the “healthiness” of behavior. At the moment, SHI have a relatively passive
role and act as payer within the system. In the future however, they wish to change their role to an (active)
healthcare partner or (neutral) healthcare navigator transitioning “the topic of health insurance away from
the pure cost carrier and we are only the payer in the system, but we want to become a player, so become a
health service provider, a navigator.” (SHI). This would also include an improvement of the communication
with insured persons and improve the service experience (cf. MHS) e.g. “how to approach them with such
a digital service and how to pitch it to them.” (SHI). At the moment, SHI are the player, where the data of
all different MSP comes together. Still, service improvements are not possible with the data that is provided
today nor with the overall speed of the system. Instead, “today we have accounting data, routine data. That's
nice that we have them, but they are very limited, both in quality and in terms of having them on time. [...]
In other words, if we think about active supply management now, they will only help us to a very, very
limited extent.” (SHI). It would seem that for SHI to be interested in data openness, they must be able to
offer predictions e.g. regarding health risks for insured persons (if they wish to receive any).
Within the group of medical service providers, the interest in openness varies widely depending on single
stakeholders “because there are many practices that are, as I said, a bit older or have existed for a longer
time and are perhaps not as open to new things. Because it also leads to deterrence in terms of cost.” (MSP).
In addition, there can be overstrain through huge data amounts and stakeholders, such as physicians, may
not have the time to analyze them. Relevance and quality of the data have to be assured, as well as there
being intelligent algorithms which can aggregate the findings, otherwise there might be, according to one
interviewee, “Paralyzing by analyzing” (MSP). In the context of treatment, almost all interviewees agreed
that MSP always make the final decision and are responsible, so digital health innovation only supports but
never replaces human intervention. It may be that “that's why we can't go there and only do medicine on
the computer, because then that intuition will be lost […]. We just have to empower, empowerment through
intelligent data is also the future.” (MSP). But MSHT can replace some of the existing services that are
provided by MSP, e.g. surgery aftercare at the patient’s home using sensor-based devices. In this context
new business models need to be developed which can also be seen as an opportunity. For example, “[…] to
bring me a market advantage in certain patient segments.” (MSP). If there is high data transparency, there

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
15
are also new ways of control, for instance “I believe that the fact that the data will also become much more
transparent will in turn have an effect on the medical service provision of the traditional players, because I
may suddenly have to justify myself as an individual physician for something I have or have not done.”
(SHI). Further, with respect to different interests regarding the degree of transparency, one interviewee
believes that “[…] many actors are not interested in transparency, you have to say that. In the end,
standardization leads to enormous transparency.” (I&I). The take away being that data openness is heavily
tied to cost and time expenditure, as well as to an overstrain through large data amounts for MSP. As a
result, adjustments of business models are necessary. To facilitate these changes, the right incentives in the
context of data openness and transparency must be created.
Individual User Level. The user possesses health data sovereignty and controls data privacy. As we were
told by one interviewee, “I [patient] have the data sovereignty. And also the possibility to provide my
information to other players, because I decide whether I want to or not.” (SHI). There is often an emphasis
on the voluntary participation of the user in donating data within the system “and then, we all have to get
used to the fact that the patient decides to whom s/he gives the data.” (MSP). For this reason, only the
collective action of many users donating their data can realize data openness on a collective ecosystem level
(Constantinides and Barrett 2015) in order to improve disease prediction. Next to this, there is tension
between value creation and value appropriation (Nambisan et al. 2019): If a user donates (sensitive) health
data, the value is created via big data analyses and training of algorithms and the collective ecosystem
appropriates value by improving treatment. From a societal perspective, donating data is a good thing, but
there is no “direct and quantifiable” value appropriation for the user. Looking at the overall incentives, there
must be added value for the user in data donation which exceeds privacy concerns. One interviewee
explained that, “I am not a friend of the fact that the data is donated voluntarily […]. And I would like a
compensation or a kind of reimbursement model for the data producers.” (I&I). In summation, apart from
an intrinsic motivation, a user’s incentive to donate data and enable data openness is rather low within the
current ecosystem design.
Conclusion and Future Research
In summary, our qualitative study analyzed data from interviews with 30 healthcare stakeholders using a
Grounded Theory approach. As part of a conceptual model we developed, we pinpointed a health service
improvement cycle and how it works in the context of integrating mobile sensor-based health technologies
into digital health ecosystems. This consists of data acquisition, flow, processing and interpretation for
continuous service improvements and demonstrates the difficult process of developing sustainable business
models within the German healthcare system. In this context, we also identified the importance of
ecosystem design and governance strategy in the context of the “paradox of openness”, which includes
stakeholder group-specific incentives and disincentives for data openness and data sharing on three levels.
These are the individual user, the ecosystem-stakeholder-group, and the collective ecosystem level.
Practically, our findings can inform the derivation of stakeholder-specific tactics and implementation
strategies to overcome barriers for embedding MSHT into the healthcare system. For instance, they may
help to better address the needs of medical service providers, health insurances, and patients via
consideration of their specific requirements. This can also guide business model development for mobile
health startups, as well as foster alignment of stakeholders and cross-group collaboration.
While the study is limited by the number and selection of stakeholders and stakeholder groups, this is
generally in line with existing studies (Mantzana et al. 2007). We would also like to acknowledge that as a
limitation, most interviews were conducted via telephone, posing a possible barrier to catching facial
expressions and non-verbal cues. There might be potential bias caused by personal opinions and
perceptions of interviewees (e.g. the voluntary participation in the interview study is most likely connected
to some degree of digital affinity). The study is also limited by its focus on the German market and associated
legal and regulatory conditions. For future research, including additional stakeholder groups, especially of
patients and their incentive structures, can generate further insights and perspectives. This could enable
the deriving of strategies for stakeholder alignment. Moreover, considering disease-specific requirements
as well as the different types of sensor-based mobile health business models is promising. We also earmark
for future research, those concepts for developing sustainable business models which are enabled by an
ethical, socially impactful and regulatorily compliant health data cycle which from our perspective deserves
special attention.

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
16
References
Adner, R. 2017. “Ecosystem as structure: An actionable construct for strategy,” Journal of Management
(43:1), pp. 39-58.
Arora, A., Athreye, S., and Huang, C. 2016. “The paradox of openness revisited: Collaborative innovation
and patenting by UK innovators,” Research Policy (45:7), pp. 1352-1361 (doi:
10.1016/j.respol.2016.03.019).
Baldwin, C., and Hippel, E. von 2011. “Modeling a paradigm shift: From producer innovation to user and
open collaborative innovation,” Organization Science (22:6), pp. 1399-1417.
Baur, N., and Blasius, J. 2014. Handbuch Methoden der empirischen Sozialforschung, Wiesbaden:
Springer Fachmedien Wiesbaden.
Benbasat, I., Goldstein, D. K., and Mead, M. 1987. “The Case Research Strategy in Studies of Information
Systems,” MIS Quarterly (11:3), pp. 369-386.
Benedict, M., and Schlieter, H. 2015. “Governance guidelines for digital healthcare ecosystems,” eHealth,
pp. 233-240.
Bertelsmann Stiftung 2018. “#SmartHealthSystems: Digitalisierungsstrategien im internationalen
Vergleich,” pp. 1-400.
Birks, D. F., Fernandez, W., Levina, N., and Nasirin, S. 2013. “Grounded theory method in information
systems research: its nature, diversity and opportunities,” European Journal of Information Systems
(22:1), pp. 1-8.
Busse, R., Schreyögg, J., and Gericke, C. (eds.) 2006. Management im Gesundheitswesen, Springer Berlin
Heidelberg.
Chiauzzi, E., Rodarte, C., and DasMahapatra, P. 2015. “Patient-centered activity monitoring in the self-
management of chronic health conditions,” BMC Medicine (13:1), p. 77.
Constantinides, P., and Barrett, M. 2015. “Information infrastructure development and governance as
collective action,” Information Systems Research (26:1), pp. 40-56.
Corbin, J. M., and Strauss, A. 1990. “Grounded theory research: Procedures, canons, and evaluative
criteria,” Qualitative sociology (13:1), pp. 3-21.
DeSantola, A., and Gulati, R. 2017. “Scaling: Organizing and growth in entrepreneurial ventures,” Academy
of Management Annals (11:2), pp. 640-668.
Eaton, B., Elaluf-Calderwood, S., Sörensen, C., and Yoo, Y. 2015. “Distributed Tuning of Boundary
Resources - The Case of Apples iOS Service System,” MIS Quarterly (39:1), pp. 217-243.
European Connected Health Alliance 2019. ECHAlliance Ecosystems.
https://echalliance.com/ecosystems/. Accessed 2 May 2020.
Gao, Y., Li, H., and Luo, Y. 2015. “An empirical study of wearable technology acceptance in healthcare,”
Industrial Management & Data Systems (115:9), pp. 1704-1723.
Gawer, A. 2014. “Bridging differing perspectives on technological platforms: Toward an integrative
framework,” Research Policy (43:7), pp. 1239-1249 (doi: 10.1016/j.respol.2014.03.006).
Gimpel, H., Nißen, M., and Görlitz, R. 2013. “Quantifying the Quantified Self: A Study on the Motivations
of Patients to Track Their Own Health,” Thirty Fourth International Conference on Information
Systems, pp. 1-16.
Giudici, A., Reinmoeller, P., and Ravasi, D. 2018. “Open-system orchestration as a relational source of
sensing capabilities: Evidence from a venture association,” Academy of Management Journal (61:4),
pp. 1369-1402.
Hein, A., Schreieck, M., Riasanow, T., Setzke, D. S., Wiesche, M., Böhm, M., and Krcmar, H. 2019. “Digital
platform ecosystems,” Electronic Markets (30:1), pp. 1-12.
Helfferich, C. 2011. Die Qualität qualitativer Daten: Manual für die Durchführung qualitativer
Interviews, Wiesbaden: Verlag für Sozialwissenschaften, Springer Fachmedien.
Islind, A. S., Lindroth, T., Lundin, J., and Steineck, G. 2019. “Co-designing a digital platform with boundary
objects: bringing together heterogeneous users in healthcare,” Health and Technology (9:4), pp. 425-
438 (doi: 10.1007/s12553-019-00332-5).
Jacobides, M. G., Cennamo, C., and Gawer, A. 2018. “Towards a theory of ecosystems,” Strategic
Management Journal (39:8), pp. 2255-2276.
Kapoor, R. 2018. “Ecosystems: broadening the locus of value creation,” Journal of Organization Design
(7:1), pp. 1-16.
Klein, H. K., and Myers, M. D. 1999. “A set of principles for conducting and evaluating interpretive field
studies in information systems,” MIS Quarterly (23:1), pp. 67-94.

Digital Health Ecosystems for Sensor Technology Integration
Forty-First International Conference on Information Systems, India 2020
17
Lessig, L. 2001. The Future of Ideas: The Fate of the Commons in a Connected World, Random House Inc.
Mantzana, V., Themistocleous, M., Irani, Z., and Morabito, V. 2007. “Identifying healthcare actors involved
in the adoption of information systems,” European Journal of Information Systems (16:1), pp. 91-102.
Martin, P. Y., and Turner, B. A. 1986. “Grounded theory and organizational research,” The Journal of
Applied Behavioral Science (22:2), pp. 141-157.
Mey, G., and Mruck, K. (eds.) 2010. Handbuch Qualitative Forschung in der Psychologie, Wiesbaden: VS
Verlag für Sozialwissenschaften.
Miles, M. B., and Huberman, A. M. 1994. Qualitative data analysis: An expanded sourcebook, Sage.
Moore, J. F. 1993. “Predators and prey: a new ecology of competition,” Harvard Business Review (71:3),
pp. 75-86.
Myers, M. D. 2019. Qualitative research in business and management, Sage Publications Limited.
Myers, M. D., and Avison, D. 2002. Qualitative research in information systems: a reader, Sage.
Myers, M. D., and Newman, M. 2007. “The qualitative interview in IS research: Examining the craft,”
Information and organization (17:1), pp. 2-26.
Nambisan, S., Wright, M., Feldman, M., and others 2019. “The digital transformation of innovation and
entrepreneurship: Progress, challenges and key themes,” Research Policy (48:8), pp. 1-9.
Olla, P., and Shimskey, C. 2015. “mHealth taxonomy: a literature survey of mobile health applications,”
Health and Technology (4:4), pp. 299-308 (doi: 10.1007/s12553-014-0093-8).
Parker, G., van Alstyne, M. W., and Jiang, X. 2017. “Platform ecosystems: How developers invert the firm,”
MIS Quarterly (41:1), pp. 255-266.
Payne, G., and Payne, J. 2004. Key concepts in social research, Sage.
Raghupathi, W., and Raghupathi, V. 2014. “Big data analytics in healthcare: promise and potential,” Health
Information Science and Systems (2:3), pp. 1-10 (doi: 10.1186/2047-2501-2-3).
Raij, A., Ghosh, A., Kumar, S., and Srivastava, M. 2011. “Privacy risks emerging from the adoption of
innocuous wearable sensors in the mobile environment,” in Proceedings of the SIGCHI Conference on
Human Factors in Computing Systems, pp. 11-20.
Rantala, K., and Karjaluoto, H. 2016. “Value co-creation in health care: insights into the transformation
from value creation to value co-creation through digitization,” Proceedings of the 20th International
Academic Mindtrek Conference, pp. 34-41 (doi: 10.1145/2994310.2994325).
Rubin, H., and Rubin, I. 2005. Qualitative Interviewing (2nd ed.): The Art of Hearing Data, Thousand
Oaks, California.
Saltman, R. B., Busse, R., and Mossialos, E. 2002. Regulating entrepreneurial behaviour in European
health care systems, Buckingham, Philadelphia: Open University Press.
Schilling, M. A. 2000. “Toward a general modular systems theory and its application to interfirm product
modularity,” Academy of Management Review (25:2), pp. 312-334.
Schreieck, M., Wiesche, M., and Krcmar, H. 2016. “Design and Governance of Platform Ecosystems – Key
Concepts and Issues for Future Research,” Twenty-Fourth European Conference on Information
Systems, pp. 1-20.
Song, P., Xue, L., Rai, A., and Zhang, C. 2018. “The ecosystem of software platform: A study of asymmetric
cross-side network effects and platform governance,” MIS Quarterly (42:1), pp. 121-142.
Swan, M. 2012. “Sensor Mania! The Internet of Things, Wearable Computing, Objective Metrics, and the
Quantified Self 2.0,” Journal of Sensor and Actuator Networks (1:3), pp. 217-253.
The Economist 2020. Governments are erecting borders for data: Data sovereignty is rapidly becoming
a big issue. https://www.economist.com/special-report/2020/02/20/governments-are-erecting-
borders-for-data. Accessed 2 May 2020.
Tiwana, A., Konsynski, B., and Bush, A. A. 2010. “Platform evolution: Coevolution of platform architecture,
governance, and environmental dynamics,” Information Systems Research (21:4), pp. 675-687 (doi:
10.1287/isre.1100.0323).
Torkamani, A., Andersen, K. G., Steinhubl, S. R., and Topol, E. J. 2017. “High-definition medicine,” Cell
(170:5), pp. 828-843.
Vesselkov, A., Hämmäinen, H., and Töyli, J. 2019. “Design and Governance of mHealth Data Sharing,”
Communications of the Association for Information Systems (45), pp. 299-321.
Wiesche, M., Jurisch, M. C., Yetton, P. W., and Krcmar, H. 2017. “Grounded theory methodology in
information systems research,” MIS Quarterly (41:3), pp. 685-701.
Zittrain, J. 2008. The future of the internet‐and how to stop it, Yale University Press.
Zuboff, S. 2019. The age of surveillance capitalism: The fight for a human future at the new frontier of
power, Profile Books.