Understanding physicians’ adoption of electronic medical records: Healthcare technology self-efficacy, service level and risk perspectives

Highlights

• This study aims to investigate whether the individual characteristics of a physician affect EMR adoption.

• A field survey is conducted with a total of 217 physicians and the results indicate that perceived service level is an important antecedent of perceived usefulness.

• Computer self-efficacy, perceived risk, and perceived service level are also important antecedents of perceived ease of use.

• This study is concluded with implications for academics, hospital managers, governments, and medical information service providers.

Abstract

Most developed countries across the globe are deploying electronic medical record (EMR) as one of the most important initiatives in their healthcare policy. EMR can not only reduce the problems associated with managing paper medical records but also improve the accuracy of medical decisions made by physicians and increase the safety of patients. Considering that physicians are the primary users of EMR, their willingness to use EMR is a critical success factor for EMR implementation in a hospital. This study aims to extend an individual-level information technology adoption model by incorporating three additional variables to investigate whether the individual characteristics of a physician affect EMR adoption. A field survey is conducted with a total of 217 physicians from 15 different academic medical centers and metropolitan hospitals for six weeks. Then, the Structural Equation Modeling (SEM) analysis results indicate that perceived service level is an important antecedent of perceived usefulness. Healthcare technology self-efficacy, perceived risk, and perceived service level are also important antecedents of perceived ease of use. This study is concluded with implications for academics, hospital managers, governments, and medical information service providers.

Introduction

Recent years, many developing countries have actively implemented from the paper-based medical record to electronic medical record (EMR), even to the exchange of electronic medical record plan. This type of change can be seen in most of the countries worldwide including USA from 2004 to 2014, China from 2004 to 2006, Canada from 2001 to 2010, Australia from 2004 to 2007, UK from 2000 to 2010, Hong Kong from 1991 to 2006, and Korea from 2006 to 2010 [21]. To participate in this important medical movement, Taiwan has promoted the adoption of EMR by government since 2010. Further, Taiwan hospitals can be mainly divided into four levels including academic medical centers, metropolitan hospitals, local community hospitals, and physician clinics & dental clinics pending on the completeness of the medical services rendered and the mission to handle the improvement of Taiwan medical research and development. In 2015, a total of 399 hospitals accounting for 80.4% of all hospitals, announced that they have implemented electronic medical records technology to provide the services to their patients [11]. This fact also indicates the popularity that the physicians use the electronic medical records in the medical centers and metropolitan hospitals to deal with the associated healthcare processes [11].

In fact, EMR development progresses in five sequential stages and they include automated medical records, computerized medical records, EMR, electronic patient records, and electronic health records (EHR) [58]. EMR applications are mostly used to support clinical records, take care of patient treatments, make medical decisions, and handle related practical applications [55]. Moreover, these applications can be classified into four types in handling record health information (such as treatment notes and reminders), health record management (such as laboratory or radiology tests), order management (such as templates and/or drag in phrases), and associated electronic communications and connectivity (such as electronic medication lists) [31]. In other words, EMR can be employed to improve medical quality [5], reduce the associated risks of adverse drug events in inpatient and ambulatory settings, enhance the patient safety, facilitate the delivery of medical care [61], and certainly lower the medical costs [44].

In this subject field, many empirical studies have aimed to explore the EMR adoption from different perspectives, such as analyzing or investigating it at the individual level or organizational level [20]; in an informational, technology, or system perspective [8]; to improve the healthcare technology self-efficacy [14]; to study the different personal characteristic such as gender [14]; and to identify the subject differences such as physicians [20], [35], nurses [56], [57], nursing students [32], patients [43], and general public [8], [56].

Physicians are the most important users of EMR in the hospital. In specific, physicians can easily create, access, distribute, and share patient records with physicians, patients, and other parties in the healthcare industry. However, during EMR implementations, many physicians are experiencing technical glitches (e.g., system crash and poor interface design), operational inefficiency, data errors, and system incompatibility issues [18]. Moreover, physicians worry about autonomy loss or changes in power structure [48]. The Centers for Disease Control and Prevention surveyed office-based physicians and determined that about 20% of these physicians used fully functioning EMR by 2011 [24]. Lack of physician support is a major barrier to the widespread adoption of EMR. To increase the physicians' intention of adopting EMR, the first and foremost requirement is to change their attitude and have their continuous support because they are the primary users of these systems [54].

In addition, some empirical studies tried to explore how to promote physicians' willingness to adopt EMR [39], [33]. One approach is to demonstrate the usefulness of EMR for physicians, such as supporting clinical treatment [5] and reduce the medical errors [16]. Other approaches may include involving physicians in the system development process [39,55], offering ease-of-use EMR applications (e.g., clinical support or electronic referral), enhancing patient information security [3], streamlined communication between physicians and patients [46], and good return of investment. Another approach is to understand the individual characteristics of physicians and to assess their potential influence on the acceptance of IS [29]. With the understanding of the individual differences among physicians, EMR can be customized to fit their personal needs. However, the current literature lacks the use of personalized approaches to influence physicians to adopt EMR [33]. Several scholars have suggested that EMR adoption be examined from the characteristics of physicians, such as age, computer sophistication [61], accessibility [29], perceived technical barriers [31], anxiety [1], and perceived threats to professional autonomy [59]. Previous research has shown that IS studies need to carefully evaluate the potential influence of unique contextual issues that exist in the medical industry on system adoption [29].

The goal of this study attempts to build and empirical validate a theoretical model on physicians adopting EMR based on the healthcare technology self-efficacy, perceived service level, and perceived risk perspectives. Further, our study place a focus on individual-level of physician, and this is mainly due to the fact that past studies had presented some interesting findings individual level factors and also the importance to investigate the significant influence on physicians’ intention to use/adopt the EHRs (individual level factors) than organization-level factors proposed by the study of Gagnon et al. [20]. By doing so, deeper insights obtained from analyzing the causal relationships between these factors can provide more valuable lessons for hospitals to increase the acceptance rate of EMR by physicians and most importantly to contribute the CS&I readers to develop additional studies in this subject area.

The remainder of this paper is organized as follows: Section 2 discusses the current and previous studies related to EMR adoption and proposes the research hypotheses of this study. Next section presents our proposed research model and method. The demographic analysis, reliability and validity tests, data analysis results using structural equation modeling are covered accordingly in Section 4. Section 5 provides the summary, highlights the contribution of this research, and discusses the research limitations, future directions and research implications. The last section concludes this study.