Determinants of physicians’ technology acceptance for e-health in ambulatory care

https://doi.org/10.1016/j.ijmedinf.2012.02.002Get rights and content

Abstract

Background

Germany is introducing a nation-wide telemedicine infrastructure that enables electronic health services. The project is facing massive resistance from German physicians, which has led to a delay of more than five years. Little is known about the actual burdens and drivers for adoption of e-health innovations by physicians.

Objective

Based on a quantitative study of German physicians who participated in the national testbed for telemedicine, this article extends existing technology acceptance models (TAM) for electronic health (e-health) in ambulatory care settings and elaborates on determinants of importance to physicians in their decision to use e-health applications.

Methods

This study explores the opinions, attitudes, and knowledge of physicians in ambulatory care to find drivers for technology acceptance in terms of information technology (IT) utilization, process and security orientation, standardization, communication, documentation and general working patterns. We identified variables within the TAM constructs used in e-health research that have the strongest evidence to determine the intention to use e-health applications.

Results

The partial least squares (PLS) regression model from data of 117 physicians showed that the perceived importance of standardization and the perceived importance of the current IT utilization (p < 0.01) were the most significant drivers for accepting electronic health services (EHS) in their practice. Significant influence (p < 0.05) was shown for the perceived importance of information security and process orientation as well as the documentation intensity and the e-health-related knowledge.

Conclusions

This study extends work gleaned from technology acceptance studies in healthcare by investigating factors which influence perceived usefulness and perceived ease of use of e-health services. Based on these empirical findings, we derive implications for the design and introduction of e-health services including suggestions for introducing the topic to physicians in ambulatory care and incentive structures for using e-health.

Highlights

► Validity of the TAM constructs in the environment of German ambulatory healthcare. ► The factors influencing important constructs of e-health in ambulatory environment. ► Significant burdens and drivers for adoption of e-health in ambulatory care.

Introduction

Health information systems have the potential to significantly improve the quality of healthcare. Information technology (IT) in healthcare has been shown to reduce the incidence of adverse events, the administration of wrong medications, the number of hospitalizations and healthcare spending [1], [2], [3], [4], [5], [6]. With this in mind, German health authorities are currently building a nation-wide telemedicine infrastructure (NTI) to enable innovations in the area of electronic healthcare (e-health) [7]. Even though most physicians in Germany acknowledge the potential advantages of e-health [8], the implementation of the NTI has been delayed for more than five years due to physician resistance [9].

Germany is not an isolated case: the low adoption of e-health is a common pattern throughout the European Union [1] despite the huge potential of electronic health services (EHS). The utilization of electronic health records (EHR), a key e-health application, has not been uniformly adopted in medical practices and hospitals. Smaller ambulatory care practices rarely use advanced EHR systems [10]. There are examples of the use of IT in healthcare that seem promising but are not implemented because of resistance from key stakeholders [11].

The motivation to conduct this research was the resistance to accept and adopt national e-health projects in Germany. The NTI project, which provides a secure infrastructure for the diffusion of e-health, lacks broad support, especially among ambulatory care physicians. Several moratoriums filed by physician associations and politicians have caused changes in the specifications for e-health projects, leading to major downsizing of the projects [12]. The main reasons given for opposition to e-health are privacy concerns, the extensive efforts required to implement the project, and dissatisfaction with the performance of the technology [13], [14].

In order to analyze the acceptance of EHS and improve future adoption, it is necessary to first identify the issues that affect the intention to use e-health. Therefore, in line with research in technology acceptance [15], we investigated the factors that influence the intent to use, perceived usefulness (PU), and perceived ease of use (PEOU) of EHS applications by physicians in the ambulatory care setting. Data security is the main concern raised over EHS in Germany. We, therefore, included this topic in our research to investigate whether it directly impacts the intent to use or if it moderates either the performance or the effort expectation, or, more precisely, is it the major barrier in the adoption attitude. Attitude towards privacy can be positively altered with appropriate message framing [11]. The possible failure of the NTI project should be prevented as the NTI represents the opportunity to achieve secure and pervasive availability of EHS in the future. An extended technology acceptance model (TAM) could help to understand the mental model of physicians and reframe e-health projects in an appropriate manner.

Our research targets the following two main research questions:

  • Can existing healthcare-related TAM approaches be verified for e-health in the ambulatory care setting in Germany?

  • What are the barriers and drivers for intention of use, PU, and PEOU for e-health applications in ambulatory care?

First, existing TAM approaches are verified in the given context. As several publications suggest, we develop the “left side of the model” [15] or the “barriers to technology acceptance” [16], and extend the established model with drivers and barriers which impact the intention to use EHS and test their actual impact. Currently existing TAM models do not consider barriers and drivers for adoption of technologies in e-health. We verify our proposed TAM model in the given context.

There is no precise definition for e-health; it is widely used and has various characteristics according to the area of application. Oh et al. compared 51 definitions of e-health in 2005 [17]. For our study, we refer to a widely used definition by Eysenbach [18], which we consider suitable for the context of EHS in German ambulatory care:

“e-health is an emerging field in the intersection of medical informatics, public health and business, referring to health services and information delivered or enhanced through the Internet and related technologies [..]”.

While we consider e-health as the broader term to reflect the application of health services for collaboration purposes, the services offered through NTI will subsequently be called EHS. Within the German NTI, it is planned to offer communication, cooperation, analysis and documentation features as web services to ensure pervasive availability and integrity of medical data to consumers in the public health system [2]. The major advantages resulting from the implementation of EHS are: (1) pharmaceutical drug safety, (2) insurance data maintenance, (3) electronic healthcare records (EHR), and (4) records of emergency treatment. The proposed infrastructure provides the storage of medical patient data on central servers or on smart cards, called electronic Health Cards (eHC) [19]. A major goal of the introduction of EHS is patient-centered medicine and the standardization, efficiency and transparency of medical treatment processes [20]. A commission composed of representatives from health insurance providers, medical associations, and governmental institutions worked out a proposal to guarantee universal accessibility of standardized data beyond institutional boundaries [21]. The proposal aims to ensure a reduction in healthcare costs by avoiding redundant examination of patients and administrative procedures and improving patient safety [7]. According to calculations, telemedicine services will provide added value of between 7.5 and 29.5 billion Euros by 2020 [22]. Fig. 1 illustrates the overall architecture of the German NTI.

Previously conducted surveys have shown that a large group of physicians has rejected telemedicine services that have been planned to be introduced as part of the implementation of the NTI [8], [13], [14], [23]. Numerous campaigns have been started by medical associations and politicians in Germany calling for a moratorium on NTI initiatives [24]. These actions can be regarded as common reactions to the acceptance of telemedicine services by healthcare personnel [11]. Because of these problems in gaining acceptance by physicians, all planned German NTI services, except the emergency record, have been indefinitely postponed. A nationally conducted study [22] concluded that savings achieved by the implementation of EHS might have had a negative financial impact on pharmacies, dentists, and physicians. While physicians could face a reduction in income, private and public health insurance companies would profit with savings of more than Euro 5 Billion.

Technology acceptance describes an individual's intentional or voluntary adoption of a technology [25]. The rapid diffusion of IT in healthcare during recent years has led to intense involvement of physicians and patients with computer technology. User acceptance is an important factor for successful adoption and utilization of the targeted technology. During the last two decades, various models dealing with technology acceptance have been developed and verified. The effectiveness of several of these models were proven for a broad spectrum of IT systems, such as personal digital assistants or electronic medical records [26]. TAM is the most widely recognized model of behavioral intention of using information systems (IS) [27]. In the original TAM model, acceptance is explained by the influence of a person's attitude towards technology and the PU of the technology. The PEOU of the system positively influences the PU and the attitude towards technology. One of the extensions of TAM, the Unified Theory of Acceptance and Use of Technology (UTAUT) model by Venkatesh et al. [28], was successfully tested in the context of healthcare [29]. Holden and Karsh [15] and Yarbrough and Smith [16] have performed extensive reviews on previous applications of technology acceptance models (i.e., TAM and UTAUT) in healthcare. All quantitative studies reviewed in their articles confirmed that the behavioral intention of adoption of technology is significantly influenced by the PU of a technology. The PEOU also affects either the behavioral intention directly, via the attitude towards the technology or the PU. The UTAUT model appends the construct “social influence” to explain the behavioral intention to use a system. Fig. 2 summarizes the constructs of TAM and UTAUT and indicates significant relationships between the constructs.

We use the construct names from TAM as applicable to our study. We did not consider other models, such as the Extended Technology Acceptance Model (TAM2) [30] or theory of planned behavior [31], as they have not been applied in the context of healthcare. Key constructs of these models are included in TAM [25] and have been verified within the UTAUT model [28].

The utilization of e-health technology can either be enforced by health authorities or be achieved as a result of voluntarily adoption. In the German healthcare setting the electronic processing of treatment data (the management of patient data and the quality control of treatment) is not mandatory [3]. It is thus essential to convince physicians to voluntarily adopt the technology to obtain the medical and economic benefits of EHS. An earlier study in ambulatory care revealed that there are two groups of physicians: those with a high behavioral intention to use EHS and those who strongly reject the technology. There is no significant difference in behavioral intention by age or the field of specialization [23]. However, the utilization of advanced IT is especially low in small medical practices [10].

Our intention is to explore more factors that influence the attitude of physicians toward EHS. The main dependent variable in our research model is the behavioral intention to use EHS. The use intention can be considered as the acceptance of our target technology: e-health. As EHS are not yet been implemented in Germany, actual use could not be measured in this research. Apart from measuring the useage intention for EHS in general, we were keen to find out which factors affect PU and PEOU of e-health applications. Our study population was physicians in ambulatory care. This group cannot be considered to be homogeneous, as there is great variation in their daily routines. The number of patient treatments per day ranges from <10 to 120. Some physicians have a “one-person” practice, while others are part of a group practice with up to 10 physicians. The importance of IT support might be lower in one-person practice due to the smaller volume of patients. In the following section we develop hypotheses on factors influencing the acceptance of technology. Using the well-established TAM constructs as a base; we add the perceived importance of various key e-health concepts to physicians.

Using a review of empirical results of TAM in healthcare, we decided to operationalize the most commonly used constructs in healthcare-related TAM research. We chose PU, PEOU and behavioral intention as these factors had the highest support level in past TAM research in healthcare. To ensure the rigor of the research, we included constructs that were utilized in more than 50% of existing articles on TAM and that were found to be significant in more than 50% of cases. We did not use “Social Influence” and “Attitude” (compare Fig. 1) as they lacked support or relevance. The following sections explain the three hypotheses derived from TAM that we developed for this research. These hypotheses include the underlying drivers und obstacles for EHS adoption.

Ten of 12 studies on TAM research in healthcare found a significant positive influence of PEOU on the PU [15], [16]. If systems are easy to use and suitable for integration into the daily routine, they are considered to deliver adequate performance in most cases. This finding is relevant because it points out that e-health applications can benefit medical professionals and patients. Broad adoption of e-health applications, however, has not yet occurred due to complicated handling issues and insufficient skill sets. Therefore, improving the usability of e-health systems is a key issue. We assumed:

H1

PEOU positively influences the PU of EHS.

PU has been defined as the degree to which users believe that using a system will help them to attain gains in job performance [28]. Previous research has employed constructs from different models, e.g., performance expectancy and PU [25]. The construct testing in UTAUT included aspects such as extrinsic motivation, job-fit, relative advantage and outcome expectations from older models [28]. The usefulness of IT system was found to be significant for user acceptance in all healthcare-related TAM studies [15], [16]. Performance in ambulatory care can be described as higher patient throughput, better quality of treatment or the extension of medical services within an institution. Thus, we assumed:

H2

PU positively influences behavioral intention to use EHS.

According to Venkatesh et al. [28], PEOU is defined as the degree of ease associated with the use of the system. The constructs describe either the complexity or the usability of a system [25]. Physicians, who often have limited IT expertise, can be frustrated by the complexity and limited user friendliness of the software. In contrary, intuitive software design can directly influence the behavioral intention to use EHS. Thus, we assumed:

H3

PEOU positively influences the behavioral intention to use EHS.

In addition to the well-established constructs of technology acceptance, we derived six additional hypotheses from the literature and tested their influence on one previously chosen TAM construct.

Research by Sequist et al. [32] highlights how perceived advantages and disadvantages of EHR affect users’ judgment of their IT experience. The perception of the current IT utilization does not only influence whether the use of EHS like EHRs are regarded as fulfilling, but also whether a person will even start to use EHS. Studies by Sequist et al. [32] and Mitchell [33] indicate that physicians’ PU of EHS positively correlates with increased IT utilization. If a physician is already IT savvy, it is more likely that he or she will have a higher demand for EHS. This aspect is stated as being one of the basic principles of e-health usage [33]. It can be concluded that efficient IT utilization influences EHS usage behavior: the more online experience and IT utilization physicians have, the greater the PU of EHS. In accordance with this and other studies on the role of usage motives and their potential effects [15], [16], [34], we focused on whether different motives for EHS use would influence physicians’ decision to use EHS. Based on the studies conducted by Sequist et al. [32] and Boddy et al. [35], drivers behind EHS use we applied in our study were: (1) current use and perceived benefit of IT in medicine (defined by statements such as “I use the internet to keep up to date” and “I use emails to contact my patients”); (2) attitudes regarding the current use of IT (defined by statements such as “nowadays, it is unthinkable to work without IT” and “we use electronic medical records to deliver treatment results”) and attitudes regarding the impact of IT on the quality of care, patient safety, patient–doctor interactions, rural and underserved healthcare; and (3) perception of delivery of culturally appropriate care (defined by statements such as “I believe patient health diaries are great for recording health problems” and “patients feel safe with our EHS”). These drivers of EHS suggest that high IT utilization positively affects the perceived importance of IT if IT is considered by the user to be dependable. We assumed:

H4

The perceived importance of IT utilization positively influences the PU of EHS.

The concerns regarding implementing NTI in Germany are the security and privacy aspects of the system. Physicians perceive data security as important as shown by the results of a survey of primary care clinicians involved in the implementation of an e-health application [32]. The prospective electronic data processing capability of the NTI can secure patient data in a manner that was impossible using the earlier paper-based documentation methods. Therefore, it is likely that greater perceived importance of data security positively influences the PU of EHS. According to Boddy et al. [35], trust in security and dependability of IT has a positive effect on perceived benefits of EHS. The advantages of EHS as addressed in this study are: access to a wide-range of physicians with different experiences in using e-health services (no experience, little experience and e-health service users); differing opinions by healthcare professionals on the security of patient data [35]; current use and perceived benefit of IT in medicine [32]; and attitudes regarding the implementation and current use of e-health services [32], [35], [36]. Disadvantages of e-health as addressed in this study are: passing of time while using e-health services; liability to document all patient-related activities; initial and ongoing financial investment in IT (IT costs); system availability, adoption, and deployment (IT infusion); clinical efficacy, efficiency, quality, and effectiveness (health performance); negatively perceived behavior of other stakeholders; and delayed feedback [36]. The fifth hypothesis addressed in this study is as follows:

H5

The perceived importance of data security in the provision of healthcare positively influences the PU of EHS.

Documentation of medical activities is one of the daily core duties of physicians and sophisticated documentation techniques require technical assistance. The performance and usefulness of EHS depends on the intensity and the elaborateness of documentation in medical practices. Leonard and Sittig [36] identified the widespread adoption of IT throughout the documentation process in healthcare provision. Gururajan [34] underlines the tremendous positive influence that IT has on documentation quality in healthcare. Proper documentation as well as computer accessibility and consistent implementation of protocols facilitate the successful implementation of practice guidelines in acute care settings [37]. Furthermore, perceived benefits of documented health services facilitate the improvement of healthcare quality [32]. Positive effects of well-structured e-health services (e.g., saving of time by physicians) are also noted [35]. According to Yasnoff et al. [38], physicians understand the importance of documentation in their practices and that standardized documentation is perceived to be an enhancer of EHS use. These studies suggest that EHS provides appropriate features to document the provision of healthcare. Thus, we assumed:

H6

The perceived importance of documentation in physicians’ practices positively influences the PU of EHS.

The German NTI project faced significant resistance from healthcare professionals from the very beginning. Despite the resistance against the project itself, which resulted in two moratorium calls from the leading German medical association, several underlying e-health applications included in the NTI project are considered useful by the majority of physicians. Of interest, the group of physicians rejecting the project had a significantly lower level of e-health-related knowledge [23]. Furthermore, surveys have shown that only one-third of German physicians feel adequately informed about the NTI project [13]. At the same time, physicians consider themselves to be the key contact person for their patients concerning e-health application, and believe that they are responsible for properly informing their patients about available e-health applications [39]. Venkatesh states that in the absence of knowledge about the target system and limited direct behavioral experience with the system, individuals will base their PEOU of the target system on general, abstract criteria [40]. In the case of e-health applications, which are considered to be generally beneficial, better knowledge about the technology will result in higher PEOU because in most instances the details of the applications are judged rather than the entire project. Thus, we assumed:

H7

Knowledge about e-health services improves the PEOU for EHS.

Standardization in healthcare is a major effort in the medical as well as in the technical domain [41], [42], [43]. Technical standards exist for medical communication (Health Level 7 [HL7]) or Digital Imaging and Communications in Medicine (DICOM) which enable transmission of documents between various medical IS. The structure of medical documents can be unified, using semantic standards, such as Clinical Document Architecture (CDA), to create common document specifications which contain information on diagnosis, medication or treatment and ensure interoperability [44]. To ensure that the collaborating entities have the same understanding of a content item in a structured medical document, standardized medical terminologies are used. The International Statistical Classification of Diseases and Related Health Problems (ICD) assigns a unified coding to diagnosis thus ensuring a common understanding of medical contents. Treatment guidelines, developed by medical experts, aim to ensure a standard of treatment for specific disease entities based on the best medical evidence available [45], [46]. Therefore, we assumed: the more a medical practice is (1) compliant with medical guidelines, (2) utilizes communication standards, (3) applies documentation standards, and (4) uses common medical terminologies, the lower the effort will be to adopt EHS. Standardized e-health software will likely include standardized guidelines for medical treatment and communication, as it is not possible to develop software adapted for individual medical practices. Thus, we assumed:

H8

The level of standardization in healthcare practices increases the PEOU of EHS.

Collaboration between facilities or institutions providing healthcare is essential for successful treatment, especially for chronic diseases [47]. According to evidence-based medical guidelines, patients pass through several treatment steps, representing a defined medical process, to ensure that they receive the best available medical treatment for their diagnosis [45]. These medical processes are referred to as clinical pathways. Medical treatment, which is currently based on a combination of intuition, knowledge, and experience of the physician, is being increasingly replaced by structured treatment [48]. Using predefined clinical pathways supports better quality of treatment. According to Picot et al. [49], value can generally not be created by IT directly. IT will only enable the application of processes and can deliver the value afterwards. A requirement for the successful implementation of IT has to be derived from the strategy and the resulting business model. In medical treatment, the processes are clinical pathways derived from evidence-based medicine and enriched by the medical experience of the physician. The daily routines of a medical practice can have several levels of process orientation. The practice of a general practitioner, for example, is likely to have a less structured environment because of a larger variety of the diseases being treated and the shorter time (7 min) of patient contact [50]. Thus, we assumed:

H9

Process orientation positively influences the PEOU of EHS.

Table 1 summarizes our nine hypotheses that form the research model in Fig. 3.

Section snippets

Sample description

The Bavarian testbed for Telemedicine has 452,000 inhabitants spread out over 2847 square kilometers. The region is geographically well-defined because of its heterogeneous structure, making it suitable for field testing [51]. The sample consisted of primary care providers and medical specialists evenly distributed between the two groups. The medical specialist areas in ambulatory care which are certified to offer specialized treatment and bill for their services are, e.g., dermatologists,

Model validation and hypotheses testing

Before testing our hypotheses, we assessed the reliability and validity of the used reflective constructs and our research model. Tests were conducted to show validity of the model constructs for the overall sample to ensure adequate performance of the measurements.

Convergent validity was tested using the composite reliability (CR), and the average variance extracted (AVE) [61]. Estimated indices were above the recommended thresholds of 0.6 [62] respectively 0.7 [63] for CR and 0.5 for AVE [61]

Discussion

The research model we describe explains the difficult situation of the German NTI and the partially low usage intention of e-health applications in Germany. In accordance with previously conducted quantitative TAM studies [15], [16], our results showed that PU and PEOU are highly significant for the usage intention for EHS in German ambulatory care. Hence, our findings confirm those from major TAM research in e-health. In addition to the verification of the TAM model for e-health in German

Author contributions

Sebastian Dünnebeil designed the study. Ali Sunyaev, Jan Marco Leimeister, Ivo Blohm and Helmut Krcmar contributed to the design of the study by review and adaption of the underlying research model.

Conflict of interest

All authors disclose any financial and personal relationships with other people or organizations that inappropriately influence (bias) this work. None of the authors has other employments than listed in the affiliations. No payments where received in relation to this work due to consultancies, stock ownership, honoraria, paid expert testimony or patent applications.

Summary points

What was already known before the initiation of the study?

  • The applicability of established technology acceptance

Acknowledgements

The study was conducted in the context of the research project “Use and Security of the German electronic Health Card System (eHC) in Bavaria”. It is a research project of the Technische Universität München funded by the Bavarian Ministry of Environment and Health. The study was not influenced, designed or reviewed by the founding source in any way. The authors thank Carol Krcmar for critical review and editing of this paper.

References (70)

  • European Commission

    eHealth for Safety Report

    (2007)
  • P. Shekelle et al.

    Costs and Benefits of Health Information Technology

    (2006)
  • P. Whitten et al.

    eHealth: market potential and business strategies

    Journal of Computer-Mediated Communication

    (2001)
  • J.M. Lappé et al.

    Improvements in 1-year cardiovascular clinical outcomes associated with a hospital-based discharge medication program

    Annals of Internal Medicine

    (2004)
  • T. Helms et al.

    Telemedical care of chronic heart failure, examples of the training- and care program “Telemedicine for the heart”

    Herz

    (2007)
  • K. Kempf et al.

    Telemedicine for Diabetes: Higher Therapy Satisfaction, Improved Vital Parameters

    (2008)
  • German Ministry of Health

    The German eHealth Strategy (Target and Strategy, Concept, Legal Framework, Activities/Roll-out Plan, Costs and Return of Investment, European Perspective)

    (2005)
  • Allensbach Institute

    Application of Telematics in the Public Health System – Results of a Representative Study Among Physicians in Ambulatory Care and Hospitals

    (2010)
  • A. Tuffs

    Germany plans to introduce electronic Health Card

    BMJ

    (2008)
  • C. DesRoches et al.

    Electronic Health records in ambulatory care—a national survey of physicians

    The New England Journal of Medicine

    (2008)
  • C.M. Angst et al.

    Adoption of electronic health records in the presence of privacy concerns: the elaboration likelihood model and individual persuasion

    Management Information Systems Quarterly

    (2008)
  • Federal Republic of Germany

    Code of Social Law (SGB) Fifth Book, Public Health Insurance

    (1988)
  • Techniker Krankenkasse
    (2009)
  • O. Kalthoff et al.

    Final Report – Evaluation of the Introduction of the electronic Health Card in the Heilbronn Testbed

    (2008)
  • A.K. Yarbrough et al.

    Technology acceptance among physicians: a new take on TAM

    Medical Care Research and Review: MCRR

    (2007)
  • H. Oh et al.

    What is eHealth (3): a systematic review of published definitions

    Journal of Medical Internet Research

    (2005)
  • G. Eysenbach

    What is e-health?

    Journal of Medical Internet Research

    (2001)
  • Fraunhofer Institut

    Specification of a Solution Architecture to Implement Applications of the electronic Health Card

    (2005)
  • gematik

    Introduction of the electronic Health Card – Overall Architecture

    (2008)
  • R. Bernnat

    Costs and Benefits of a Telematics Infrastructure for the German Public Health System

    (2006)
  • S. Dünnebeil et al.

    Do German physicians want electronic Health Services? A characterization of potential adopters and rejectors in German Ambulatory Care

  • A. Tuffs

    Germany puts universal health e-card on hold

    BMJ

    (2010, January)
  • P.J. Hu et al.

    Examining the technology acceptance model using physician acceptance of telemedicine technology

    Journal of Management Information Systems

    (1999)
  • Q. Ma et al.

    The technology acceptance model: a meta-analysis of empirical findings

    Journal of Organizational and End User Computing

    (2004)
  • V. Venkatesh et al.

    User acceptance of information technology: toward a unified view

    Management Information Systems Quarterly

    (2003)
  • Cited by (215)

    View all citing articles on Scopus
    View full text