Effective utilization of Information Technology (IT) in the health sector can lead to cost cutting and restructuring [1] which can help citizens avail medical services at an affordable price. The main challenges for any technology are the users’ acceptance and regular use of technology in day-to-day operations [2] and successful integration with the entire clinical workspace [3]. The mere provision of technology does not guarantee adoption and usage of clinical IT systems [4]. Investments in clinical IT systems can be beneficial only when the systems are used by the physicians for accurate diagnosis and prescriptions [5]. If the usage rate is low, the technology can no longer be effective for organizations [6]. According to Lowenhaupt [7], physicians’ adoption of clinical IT systems is not encouraging. Many studies have corroborated this point [8], [9], [10], [11], [12]. What is (are) the factor(s) that contribute to this negative behavior?
We argue that professional autonomy (or the lack of it) plays a central role in deciding whether clinical IT systems will be adopted or not by the physicians. Professional autonomy is defined as “professionals having control over the conditions, processes, procedures, or content of their work according to their own collective and, ultimately, individual judgment in the application of their profession's body of knowledge and expertise” (: p. 207). The professional autonomy is granted by the State through the necessary legislation and can vary from one country to another country. The professional bodies ensure that this autonomy is not invalidated. In general, professional autonomy addresses the control over the profession's scientific knowledge–production, conveyance, application and evaluation in practice [13]. Walter and Lopez [12] have introduced a new construct called perceived threat to professional autonomy to study the adoption behavior of physicians and we use the same construct in this research. Esmaeilzadeh et al. [14] have explained some of the challenges and issues regarding adoption of health information technology and have emphasized the importance of perceived threat to professional autonomy in IT adoption in the health care sector.
Clinical IT applications are of two types [15], [16]: (1) Electronic Medical Records (EMR) systems–computer systems that allow users to create, store, and retrieve patient charts on a computer and (2) Clinical Decision Support Systems (CDSS)–computer systems that use patient data to generate case-specific advice. The data for CDSSs predominantly come from the EMR systems. In this research, we address issues related to adoption of CDSS by physicians in hospitals in a developing country. More physicians are comfortable in adopting EMR than CDSS. A study among the outpatient physicians in USA indicates that 17% use CDSS and 30% use EMR systems [17]. Why is there a need for a different framework to study the behavior of CDSS adoption by physicians when there are general frameworks available?
Many theories and frameworks have been developed from different perspectives to study user acceptance of IT systems and this single factor is critical to determine if the adoption has been successful or not [18], [19], [20], [21]. A unified model called UTAUT [22] and DeLone and McLean model [23] are popular among researchers to study the adoption and effectiveness of IT systems. We agree with other researchers that these frameworks are for general users and are not completely suitable for studying the adoption behavior of professionals such as physicians. Each profession has special contextual characteristics that may affect IT adoption behavior. The physicians are different because of their specialized training, autonomous practices and professional work arrangements [24].
Our study differs from earlier studies in the following ways. First, we study the antecedents and outcomes of perceived threat to professional autonomy. Previous studies have studied this construct with a few other constructs. For example, Walter and Lopez [12] have studied this construct with perceived ease of use and perceived usefulness on intention to adopt clinical IT systems. Sambasivan et al. [25] have studied this construct with performance expectancy, effort expectancy and physicians’ involvement in decision making on intention to adopt CDSS. In this study, we have developed an integrated framework that includes the antecedents and outcomes of physicians’ perceived threat to professional autonomy. An integrated framework can help us comprehend the entire mechanism of CDSS adoption. Second, the study has been conducted in a fast developing country of South-east Asia, Malaysia. Developing countries face more challenges than developed countries in implementing IT systems. The usage rate of CDSS in developing countries is very low [26] and some of the pertinent reasons are: (1) problem in implementing and using EMR (EMR provides input to CDSS), (2) poor design of human interface, (3) difficulties in fitting CDSS as a part of routine care process, (4) computer illiteracy of physicians and (5) cost of purchasing and implementing CDSS.
Literature looks at professional autonomy from two perspectives: (1) empirical–sociological and (2) normative. The first perspective deals with the actual power and competences of physicians and argues that professional power is a social reality that cannot be disputed. This enables the professionals to claim that professional autonomy has to serve their own interests. The second perspective considers professional autonomy as a professional ideal and is related to a standard of excellence for
A significant milestone in IT adoption literature is the development of UTAUT model by Venkatesh et al. [22]. As indicated earlier, UTAUT model and other models like Technology Acceptance Model (TAM) are meant for studying general users. Therefore, in this research, we have developed a framework by integrating (1) the characteristics that are unique to physicians–(perceived threat to) professional autonomy, attitude toward knowledge sharing, interactivity perception and computer self-efficacy,
Seven large public hospitals and five private hospitals located in and around Kuala Lumpur, capital city of Malaysia were chosen to conduct the study. The criteria for selecting these hospitals were: (1) the availability of latest IT systems such as EMR and (2) availability of specialty doctors. These hospitals were most likely to use CDSS to improve healthcare delivery. Before commencing, permission to conduct the study was obtained from the Ministry of Health and the questionnaire and the
We conducted Confirmatory Analysis (CFA) on all the constructs. We followed the four steps outlined by Hair et al. [61] to test the convergent and discriminant validity of the constructs. The Average Variance Extracted (AVE) and Composite Reliability (CR) were calculated for each construct and the values are given in Table 2. It can be seen that AVE and CR values satisfy the threshold levels prescribed by Hair et al. [61] and the levels are: (1) above 0.5 for AVE and (2) above 0.7 for CR.
The
This research was undertaken in a developing country, Malaysia, on the assertion that professionals such as physicians were different from general users of IT. The models developed to study general users might not be comprehensive for analyzing the adoption behavior of physicians. The framework developed addressed the factors influencing the adoption of CDSS by physicians in public and private hospitals. We argued the role of perceived threat to professional autonomy as a central construct that
PE and MS designed the study. PE collected and analyzed the data. MS, NK and HN monitored the progress of the study and vetted the results at each stage. MS prepared the draft and vetted by PE, NK and HN.
There is no conflict of interest in this study. Summary points What was already known on this topic? Adoption of technology by physicians is hindered by professional autonomy. An integrated framework to study professional autonomy is missing in literature. What this study added to our knowledge? Decreasing perceived threat can increase adoption of information technology systems. Involvement in decision making and performance expectancy can increase adoption. Attitude to knowledge sharing and perceived
Similarly, Sahay et al. (2010) detail the interplay of conflicting institutional logics shaping the HMIS implementation in a developing country. Healthcare organizational processes such as management support (Ahmadi et al., 2015; Hikmet, Bhattacherjee, Menachemi, Kayhan, & Brooks, 2007; Ingebrigtsen et al., 2014; Lavalle, Hopkins, Lesser, Shockley, & Kruschwitz, 2010; Sun, Cegielski, Jia, & Hall, 2018) and response to technology-resistance (Abdekhoda, Ahmadi, Gohari, & Noruzi, 2015; Boonstra & Broekhuis, 2010; Esmaeilzadeh, Sambasivan, Kumar, & Nezakati, 2015) are important to understand adoption of health information systems; these processes have been studied extensively within the health information systems literature (Davidson, Baird, & Prince, 2018). However, governance arising from the institutional field is also key to understanding how or why HA adoption may proceed within or across organizations within a field.
We identified three factors, including professionals' level of involvement in decision makings [38, 42], institutional trust [35], and organizational attitude toward the AI techniques [46], as organizations' cultural characteristics associated with the adoption of AI in healthcare settings. Two studies found a positive association between involvement in decision-making and the adoption of AI [38, 42]. One study showed that institutional trust is positively associated with the intention to use electronic healthcare record systems [35].
