Modeling the Acceptance and Use of Electronic Medical Records from Patients’ Point of View: Evidence from Saudi Arabia

The employment of IT in healthcare reects a need to meet the increasing demands of patients and to upgrade the quality and productivity of the provided services. However, the literature demonstrates many failures in systems and IT technology implementation in the context of healthcare. One of the reasons behind these failures is users’ resistance to accept such new technology. The ministry of health in Saudi Arabia has prioritized the embracing of new technologies that could level up the healthcare service, such as the adoption of electronic medical record (EMR) systems. On the other hand, studies that address the acceptance and use of hospital information systems and EMRs in Saudi Arabia from patients’ points of view are scarce. The aim of this study was to explore patients’ acceptance of an EMR system by testing a proposed theoretical model adapted from the technology acceptance model (TAM) and the unied theory of acceptance and use of technology (UTAUT). organizational challenge/organizational characteristics, human challenge/individual characteristics, and technological challenge/technological

healthcare institutes in the developing countries, it doesn't assure users' adoption and acceptance (Esmaeilzadeh et al., 2015). At the same time, the HIS adoption literature shows that more than 50% of health information systems are not used because of factors such as the extensive time needed for digitizing patients' records, checking and evaluating HIS decisions, and users' refusal or resistance (Kilsdonk et al., 2010). The usefulness of any great technology can be measured by the number of uses and its employment in achieving its purposes. Thus, various technological barriers appear when users start interacting with the technology. Some of these barriers are issues relating to ease of use, usefulness, complexity, usability, output quality, compatibility, and observability. These technological challenges are studied through human-computer interaction. Also, user acceptance plays a major role in the successful adoption of healthcare IT solutions and systems (Aldosari, 2012; Alsharo et al., 2019). For these reasons, studying users' behavior toward using and accepting the system can improve the system implementation and performance (Kilsdonk et al., 2010). The HIS adoption literature shows that many HIS implementations fail as a result of users' resistance or rejection (Kijsanayotin et al., 2009;Spetz et al., 2014). Consequently, exploring the factors that impact the users' adoption is signi cant to overcome the implementation challenges that may hinder the system success (Cresswell et al, 2013). Based on that, this study aims to explore patients' perceptions of the adoption of the online services of an EMR system. User behavior toward new technology can be illustrated and expected by some theories and models such as the theory of reasoned action (Fishbein & Ajzen &, 1975), the technology acceptance model (TAM) (Davis, 1989), the motivational model (Davis, 1989), the theory of planned behavior (Ajzen, 1991), a combined TAM and TPB (Taylor & Todd, 1995), diffusion of innovation theory (Moore & Benbasat, 1991;Rogers, 1962;Rogers, 2010), and the Uni ed Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh et al., 2003).
In the context of HIS adoption and acceptance in SA, Mohamed & El-Naif (2005) conducted a quantitative study in a military hospital in Riyadh to explore opinions and perceptions on the implementation of an EMR system. The sample included 105 physicians, 109 nurses, and 120 patients, who were surveyed on the current medical records department (MRD) and the quality of MRD services. However, only physicians were surveyed concerning EMRs. The study revealed that physicians at the time of the study had negative perceptions toward conversion from the current paper-based MRD to an EMR system. Only 31.3% of physicians believed it was time to move to EMRs, 68.7% believed that the existing MRD was more credible than EMRs, and more than 90% of the hospital physicians believed EMRs would add a burden on them, as EMRs require signi cant data entry from their side. Also, more than 80% of the sample physicians believed EMRs would decrease their productivity.
Another quantitative study by Bah et al. (2011) was conducted to measure EHR system adoption across the governmental hospitals located in the Eastern Province of SA. The study targeted the hospitals' IT managers. Out of 19 hospitals, only three hospitals used the same EHR system. The implemented system provides ve main functionalities: chart review, decision support, order entry, documentation, and additional tools. For chart review functions, the system can obtain and review lab and radiology results, review progress notes, and monitor current and past medications and medication re lls. The decision support functions of the system include receiving drug interaction and drug-allergy alerts when writing prescriptions, and highlighting test results that are out of normal range. Order entry functions enable the hospital staff to enter the lab, radiology, and pharmacy orders. Also, the EHR systems have some documentation functions that involve the systems' ability to create and maintain patient-related medical problem lists and common medication lists, identify patient-speci c allergies, and document patient discharge instructions. Finally, the additional tools of the systems include managing patient referrals, automating coding of disease conditions, generating health statistics, and performing data backup and disaster recovery (Bah et al., 2011).
The Ministry of Health in SA has set many standards and policies for electronic health services to ensure effective and reliable implementation (Alkabba et al., 2012). Yet Alkabba and colleagues highlighted that the implemented systems were not capable of or not utilized to perform the following functionalities: creation and reviewing of scanned documents, communications, allowing hospitals' physicians to access patient records over the Internet when they are outside the hospital, and allowing patients to access their health records over the Internet (Bah et al., 2011). A similar study was carried about by Aldosari (2014) to examine the status of the EHR system in Riyadh as a sample of SA. Riyadh has a total of 30 hospitals; of these, 22 (16 governmental and 6 private) were surveyed to determine the rate, level, and determinants of EHR adoption. The study targeted project managers, medical directors, heads of IT departments, and EHR development teams. A total of 280 respondents completed the survey across all hospitals. The study found that for the rate of EHR adoption in Riyadh, of 22 hospitals, 19 had fully or partially adopted an EHR system. Of the hospitals included in the sample, 50% (11 hospitals) had implemented a fully functional EHR system, 36% (8 hospitals) were in the process of implementing an EHR system, and 14% (3 hospitals) had not yet implemented a system.
Unlike the Eastern Province of SA (Bah et al., 2011), the 19 adopted hospitals implemented different EHR systems. Regarding the EHR system adoption level, three adoption phases were considered: implementation, maintenance, and improvement. For the implementation phase, the study found there was low preparation for the needed actions for the conversion from the current paperbased record system to an electronic one. For the maintenance phase, the major weakness was centered on software updates and maintaining and updating the CDSS. Concerning the improvement phase of EHR adoption, HIS communication and sharing is the main issue (Aldosari, 2014). With respect to EHR system adoption determinants, the study examined three primary determinants: hospital size, hospital ownership, and the EHR system development team. First, hospital size was reported to be positively related to the level of care complexity. Larger and tertiary hospitals had advanced EHR adoption. For the hospital ownership (public or private), the study ndings showed that public hospitals were more advanced in the system implementation and maintenance phases than were private hospitals, which had better performance in the system improvement phase (Aldosari, 2014).
Additionally, a study on EHR adoption and barriers from nurses' perspective was conducted by El Mahalli (2015), which targeted 185 nurses in three public hospitals where the same EHR systems were implemented in the Eastern Province of SA. The applied systems provided some functionalities analogous to the ones in the study by Bah et al. (2011). The study revealed the underutilization of all EHR functionalities across the three hospitals. Also, similar to the ndings of Bah et al. (2011), there was no utilization of any communication features; there were zero instances of using tools that "[allowed] patients to use the Internet to access parts of their health records" (El Mahalli, 2015).
For other forms of HISs, a study carried by Aldosari (2012) investigated radiology users' acceptance of a PACS in the radiology department at King Abdulaziz Medical City hospital in Riyadh, SA. The study used a modi ed TAM that contained three constructs: perceived usefulness (PU), perceived ease of use (PEOU), and change. In addition, a survey was conducted to validate the proposed model, and the targeted population (89 respondents) was radiology staff: consultants, radiologists, residents, technologists, and others who used the PACS in their work in the radiology department. The study concluded that all constructs in the proposed model (i.e., PU, PEU, and change) had a signi cant effect on radiology staff acceptance and the use of the PACS (Aldosari, 2012).
A recent study carried by Aljarboa et al. (2019) aimed to explore the factors of CDSS adoption in the context of the Saudi healthcare sector and to identify the possible use challenges of such technology. CDSS is a computerized mechanism that "provides clinicians with knowledge, intelligently ltered or presented at appropriate times, to enhance health and healthcare, and can be seen as an effective pathway to improve patient safety, providing, for instance, alerts for error reduction" (Zikos & Delellis, 2018). The study targeted nine physicians from various specializations at different public and private hospitals and used quali ed semi-structured interviews and applied some modi cations to the UTAUT model constructs to t both the healthcare and SA contexts. Two constructs were added to UTAUT: diagnostic accuracy and patient con dence. The study concluded that only ve factors affected the physicians' intention to use CDSS: performance expectancy, effort expectancy, facilitating conditions, diagnostic accuracy, and patient con dence. The social in uence determinant was reported as an insigni cant factor (Aljarboa et al., 2019). Table 2 shows a summary of the studies mentioned above.

Objective
As per the literature, HISs contribute in improving healthcare productivity and cost-effectiveness, empowering patients by involving them in healthcare decision making and reducing medical errors (Zhang, 2002). On the other hand, there have been few investigations and studies regarding patients' adoption and acceptance of health IT solutions that require their inputs. Most adoption and acceptance studies focus on hospital staff, which includes physicians, nurses, medical directors, decision makers, laboratory technicians, and pharmacists (Alshahrani et al., 2019).
Considering this article's scope, which focuses on Saudi Arabia, the Ministry of Health in SA has prioritized the digitizing of health records and the development of electronic solutions in the healthcare context (Bah et al., 2011). The Ministry of Health has also made a variety of agreements to smooth the adoption and implementation of EMRs in primary healthcare centers. Nevertheless, the use of EMRs among Saudis is uncommon (Al-Sobhi et al., 2011). It is di cult to track the adoption of EHRs in SA due to limited publications in that area. A review of the 31 currently available publications on eHealth in Saudi Arabia showed that eHealth implementation and adoption is growing. On the other hand, the number of studies is limited and not growing at the same pace (Alsulame et al., 2016). Very few research studies have been conducted to quantify or measure the adoption of such technology (Aldosari, 2014), and none of these studies, to the author's knowledge, has focused on patients as among the HIS users. On this basis, this study aims to answer the following research question: What are the factors in uencing patients to accept and use the online services of an EMR system?

Methods
This section shows the model proposed for the study and the constructs considered. The research is posited on the theoretical background of the TAM. Compared to other models that examine technology adoption, TAM is the most widely used in the literature (Ahlan & Ahmad, 2014; Ma & Liu, 2004). TAM is a theory that illustrates and helps to predict the impact of a system's usefulness and ease of use on users' intention to accept and use that technology (Davis, 1989). TAM was initially established to investigate the reasons why some users did not use and accept new technologies or systems when they were available to them (Holden & Karsh, 2010). TAM has been broadly used in many HIS studies to depict healthcare providers' behavioral intention to use HISs. Moreover, many researchers have argued that TAM is more applicable than other technology acceptance theories in healthcare settings ( 16 of 20 studies concerning the implementation of health IT for patient care used the TAM or modi ed TAM, which predicts a signi cant portion of health IT acceptance and use. Although some studies used UTAUT model as it is the recent one and integrates constructs from different models, a study conducted recently revealed that applying the UTAUT to study the adoption of HIT in developing countries is inadequate and needs to be modi ed to t the developing countries context (Bawack et al., 2018). In this study, TAM (Davis, 1989) is used as a theoretical foundation with one more construct from the UTAUT model (Venkatesh et al., 2003). Table 3 and Figure 1 demonstrate the proposed model.

Study Setting and Participants
The study was conducted in the Royal Commission Medical Center (RCMC) at Yanbu, SA. It is one of the main hospitals in Yanbu. RCMC is a public (governmental) hospital that was established in 1980 with a capacity of 68 beds; it currently has a capacity of over 400 beds.
RCMC has one main center, which is the Occupational Health Care Center, and seven polyclinics. RCMC received local and international accreditation from the Saudi Central Board for Accreditation of Health Care Institutions and the Joint Commission International. RCMC recently implemented an EMR system for managing patients' information, medical history, clinic visit appointments, examination appointments, and other functions. The system has many users, including patients. The targeted population includes RCMC patients. Random sampling was used, and a total of 116 patients responded to a questionnaire.
Considering the FCs construct of the UTAUT model (Venkatesh et al., 2003), an d based on the four core determinants of TAM (PEOU, PU, ATT, and BI) (Davis, 1989), the following hypotheses were formulated: Hypothesis 1 (H1): Facilitating conditions (e.g., information on the hospital website, skills, knowledge, and availability of technical support) will have a positive effect on the RCMC EMR system perceived ease of use.
Hypothesis 2 (H2): Perceived ease of use of the RCMC EMR system will have a positive effect on the perceived usefulness of the system. Hypothesis 3 (H3): Perceived usefulness of the RCMC EMR system will have a positive effect on patients' attitudes toward the system. Hypothesis 4 (H4): Patients' attitudes towards using the RCMC EMR system will have a positive effect on the intention to use the system.

Data Collection
A structured questionnaire was developed for this study to investigate patients' perceptions of using RCMC's electronic services. The questionnaire was online and written in Arabic and English. It included two sections where a combination of category and scaleranking questions were used. The response scale was a 5-point Likert-type scale, serving to measure the level of patients' agreement with questions and statements (1-Strongly disagree, 2-Disagree, 3-Neutral, 4-Agree, and 5-Strongly agree).
The rst section of the questionnaire aimed to identify the patients' demographic characteristics using three questions inspired by Aldosari et al. (2018). The second part included 18 questions that were based on TAM user acceptance factors: PU, PEOU, ATT, and BI. Also, three questions were based on the FCs factor of the UTAUT model. Table 4 summarizes the items used.
All questions in the online questionnaire were set to be required for submission; therefore, there was no missing data among the 116 responses. For the sample characteristics, there was a higher proportion of female respondents (66%) than male respondents (34%). More than half of the participants belonged to the age group of 20-29 years old (57%). The second-highest age group of the participants was 30-39 (18%), followed by 40-49 (10%), and patients who were 50-59 and over 60 represented 15% of the sample.

Results
For the study model analysis, partial least squares-based structural equation modelling was performed for the data analysis using SmartPLS version 3.3.2. Two-step analysis was applied: measurement model and structural model analysis.
Testing the Measurement Model Measurement model assessment aims to evaluate the reliability and validity of a proposed model's constructs, which include both re ective and formative measures. For the current study, re ective constructs, factor loadings, composite reliability (CR), and average variance extracted (AVE) were applied to evaluate the convergent validity. As Table 6 shows, all item loadings were at more than 0.7, which exceeded the recommended value of 0.5. The item composite reliability ranged from 0.889 to 0.957, which exceeded the recommended value of 0.7. For the AVE, the items' values ranged from 0.668 to 0.882, which exceeded the recommended value of 0.5.
After convergent validity, discriminant validity was tested, which included item cross-loading and the Fornell and Larcker (1981) criterion. For item cross-loading, as Table 7 shows, the loadings on constructs (written in bold) were higher than loadings with other constructs. For the inter-construct correlation test or Fornell and Larcker (1981) criterion, Table 8 shows that the construct square root of the AVE (written in bold) was higher than the correlation with other re ective constructs. Therefore, the re ective measurement model proved the convergent and discriminant validity.
For the formative measurement model, item weights and multicollinearity between indicators were tested to validate the formative constructs. First, the signi cance of item weights was evaluated. As shown in Table 9, there were some items with signi cant weight, such as PU1 and FC3, but they were not deleted. After that, multicollinearity was examined by the variance in ation factor (VIF). All items' VIFs (Table 9) were within the acceptable range, which is below 5. Therefore, no items were deleted. Table 10 and Figure 2 demonstrate the hypothesis testing that was used to assess the structural model. As shown, facilitating conditions (β = 0.791, p < 0.01) was positively related to perceived ease of use; therefore, H1 was supported. Perceived ease of use was signi cantly related to perceived usefulness (β = 0.755, p < 0.01), which supported H2. Next is the relationship between perceived usefulness (β = 0.479, p < 0.01) and attitude, which was also positively signi cant. Thus, H3 was supported. Similarly, the last relationship was veri ed, in which attitude (β = 0.652, p < 0.01) was positively related to behavioral intention. Therefore, H4 was supported.

Discussion
This study was developed with the aim of exploring patients' acceptance of the electronic services provided by the RCMC EMR system in Saudi Arabia. A modi ed TAM was used to investigate patients' acceptance of the EMR system. implies that healthcare technology, such as EMRs, would be perceived as more useful if it were easy and effortless to use. The same relationship was positive and signi cant in Tubaishat's (2017) study, which applied TAM to investigate nurses' use behavior of an EHR system. Similarly, many studies have con rmed this relationship from physicians' perspective, such as studies by Alsharo et al.
(2019), Chen and Hsiao (2012), and Yarbrough and Smith (2007). In addition, a study conducted by Or et al. (2011) to examine home-care patients' acceptance of a web-based and interactive self-management technology revealed that PEOU did not directly affect the BI, but it had a signi cant impact on PU. Another example is a study by Dutta et al. (2018), which examined individuals' intention to use one form of HIS, the personal health record. The ndings showed that the PEOU-PU relationship was signi cant. Third, the current study showed that the relationship between PU and ATT was con rmed ( Last is the relationship between ATT and users' BI. Contrary to the ndings of I nedo (2017), which did not con rm a relationship between nurses' attitudes and their behavioral intention to use HIS at work, in the present study the ATT and BI relationship was con rmed. The effect of physicians' ATT on BI has also been shown to be positively signi cant in other studies (

Conclusion
This study aimed at studying patients' perceptions about accepting and using the online services of an EMR system in Saudi Arabia. The study proposed and tested a model that includes constructs from UTAUT and TAM models. The results revealed that facilitating conditions that involved patients' knowledge, skills, and the provided support by the hospital directly in uenced the degree of system perceived ease of use, which in turn affected the EMR's perceived usefulness and directly affected patients' attitudes toward using the EMR. In addition, a positive effect was found between attitudes and the behavioral intention to use the EMR. The study analysis con rmed the validity of the proposed model in the context of patients' acceptance of healthcare technologies.
This study makes signi cant contributions to HIS research because it considers patients as among healthcare systems' users. Most healthcare adoption research has focused on physicians, nurses, and hospital administers. Patients have received little attention, especially in Saudi Arabia. To the best of the author's knowledge, this is the rst study in the context of HIS/EMR adoption and acceptance from patients' perspectives in Saudi Arabia. On the other hand, the study has several limitations. First, it was limited to 116 patients who were located within the limited geographical context of Saudi Arabia. Different results might be seen in different countries and environments. Second, the proposed and tested model included only ve constructs, which might be insu cient for understanding the use behavior of patients. Third, the patients' demographic characteristics were not considered in drawing the study ndings. Last, the study questionnaire was only available online, which implied that patients who did not have access to the Internet were not included in the sample.
For future research, more studies are required to explore the factors that affect all HIS users' adoption to overcome the implementation challenges that may hinder the system's success. Investigation of patients' attitudes toward the adoption and acceptance of HIT/HIS is recommended (Aljarboa et al., 2019). In addition, according to Alkabba et al. (2012), the con dentiality of patients' information is ranked among the top three ethical issues for healthcare providers, patients, and their families in SA.
Therefore, different constructs that are important to patients could be integrated with the model proposed in this study; these could include data privacy, data quality, system quality, system complexity, and social norms. Also, further research is needed on the impact of patients' demographic characteristics on their adoption behavior-characteristics such as gender, age, computer use experience, and education level. Author's contribution The author was the sole contributor to the entire study.

Funding
The author was the source of funds for this study.

Availability of data and materials
The data sets analyzed during the present study available from the corresponding author on reasonable request.

Ethics approval and consent to participate
Ethical approval was obtained from the research unit of the Yanbu University College for the research process.
The study participants, who are patients, were informed that their participation in the online survey was entirely voluntary, and they may choose to withdraw at any time. Participants answers are kept con dential, and no personally identifying information would be collected or recorded Consent for publication I agree to the publication of this manuscript.

Competing interests
The author declares that he has no competing interests.

Author Details
Department of Management Science, Yanabu University College, Yanbu, Saudi Arabia "The Electronic Health Record (EHR) is a longitudinal electronic record of patient health information generated by one or more encounters in any care delivery setting. Included in this information are patient demographics, progress notes, problems, medications, vital signs, past medical history, immunizations, laboratory data and radiology reports. The EHR automates and streamlines the clinician's work ow. The EHR has the ability to generate a complete record of a clinical patient encounter-as well as supporting other care-related activities directly or indirectly via interfaceincluding evidence-based decision support, quality management, and outcomes reporting" (HIMSS, n.d.).
HIS A hospital information system (HIS) is de ned as "a comprehensive, integrated information system designed to manage the administrative, nancial and clinical aspects of a hospital. It aims to achieve the best possible support of patient care and administration by electronic data processing" (Ismail et al., 2010, p. 16-24). "It is an information system that performs the function of processing data, information and knowledge in the secondary and tertiary healthcare levels" (Lee et al., 2011(Lee et al., , p. 2129(Lee et al., -2140 HIT Health information technology is "the application of information processing involving both computer hardware and software that deals with the storage, retrieval, sharing, and use of health care information, data, and knowledge for communication and decision making. Applications of health IT include the electronic health record (EHR), the personal health record, computerized physician order entry, and clinical decision support. In addition, health information exchanges are being developed to support sharing of information electronically among health care providers" (O ce of the National Coordinator for Health Information Technology, n.d.) eHealth eHealth refers to "all forms of electronic health care delivered via information and communication technology channels, ranging from informational, educational, and commercial, to direct services offered by Healthcare organizations, professionals, and consumers themselves" (Oh et al., 2005). E-health is "the use in the health sector, of digital data for clinical; educational and administrative purposes, both at the local site and at a distance" (Geangu et al., 2014, p. 473-482)

CDSS
A clinical decision support system is "a computerized mechanism developed to assist healthcare providers make quality decisions regarding patient treatment and improve clinical management" (Hunt et al. 1998, p. 1339-1346).  "The degree to which a person believes that using a particular system would enhance his or her job performance." Davis (1989) Perceived ease of use (PEOU) "The degree to which a person believes that using a particular system would be free of effort." Davis (1989) Attitude toward behavior (  I have the knowledge necessary to use the electronic health services offered on the website or the app.

FC1
I think using the electronic health services in the website or the app ts well with the way I like to get health services.

FC2
A speci c person (or group) is available for assistance if I have problems using the electronic health services offered in the website or the app.

FC3
Attitude (ATT) Using electronic health services is a good idea. ATT1 I will be satis ed in using the electronic health services. ATT2 I think it is valuable to use electronic medical services. ATT3 Using electronic health services is favorable to me. ATT4 Behavioral Intention (BI) I intend to use the electronic health services in the next months. BI1 I predict I would use the electronic health services in the next months. BI2 I plan to use the electronic health services in the next months. BI3    Note: Diagonal elements are the square roots of the AVE of the latent variables and indicate the highest in any column and row.    Results of the Structural Model. Note: *p<0.01

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementaryFile1Questionnaire.docx