Developing Minimum Dataset For Electronic Health Record In Iran

Background: Health information system is an integral part of the health system that has a vital role in increasing the eciency of the health system, especially in primary health care settings. This study was conducted to determine the minimum dataset required in the electronic health record within the health system of Iran as a lower middle income country. Method: This study combines qualitative and quantitative methods. It includes three main stages: reviewing the theoretical foundations of research, designing the main framework for interview questions, conducting a qualitative study by interviewing 42 managers of the health system across the country to determine the minimum dataset in the electronic health. Interviews were carried out from 2020 to 2021. The validity of data was assessed by Delphi method using SPSS 15 software. Results: After reviewing the minimum dataset in the electronic health records of seven countries, 7 main concepts and 23 sub-concepts were extracted from the interviews with experts across the country. Accordingly, 159 information elements were surveyed and a two-round Delphi provided 145 information elements in seven categories of children's program, mothers' program, mental health, elderly, paraclinical services, drug program, and vaccination. Conclusion: Health systems in different countries determine the minimum dataset required in health care setting based on their demographic and epidemiological needs, which can facilitate access to accurate and unambiguous information.

proper communication between health care providers as well as timely decision-making for managers (2).
In Iran, attempts were made to create a centralized system that provides all information required at different management levels in a single software, including IEHS (Iranian Health System), the comprehensive software for all Iranian hospitals (Takfab project), an electronic le project )Shahab Health( (9). Despite high costs of this project for various reasons such as centralized management and monopoly, limited work experience of contracting companies in the health sector Failure failed (10) However, it should be noted that the establishment of such a system requires the creation of integrated infrastructure based on the social and cultural conditions of the country as well as the ability to exchange information at different levels of service delivery (11). Therefore, designing and determining the minimum dataset is a major step in developing a national information registration system that allows medical institutions to identify their patients, meets government requirements and internal needs, and ultimately satis es the medical community (12).
Since the current health system collects information from different areas (health, treatment, and insurance) in a dispersed and non-integrated manner, it is necessary to determine the standard data elements to improve the design of EHRs. Given the importance of primary health care information system in increasing e ciency, the rational use of health resources, reduced health care costs, improved quality of primary health care, higher satisfaction of patients and ultimately enhanced community health, the present study aims to determine the minimum dataset required to design a health care information system in Iran. The ndings of this study can offer valuable experiences to other countries that deal with the same problems in their health care systems.

Methods:
This study, which uses a combination of qualitative and quantitative research, was conducted from 2020 to 2021. The study steps are as follows: In the rst step, valid databases such as PubMed, Science Direct, SID, and Google Scholar were searched using various combinations of keywords such as electronic health records, primary health care, outpatient care, health information technology. Then, information on minimum dataset in HIS was obtained from seven leading countries in the design and implementation of e-health records, including the United States, Canada, the United Kingdom, Australia, Turkey, Malaysia, and Iran. Accordingly, the minimum dataset was extracted from in each country.
Accordingly, using the comparative matrix, the similarities and differences of the minimum dataset were identi ed in the selected countries and interview questions were designed accordingly. The interviewees were selected from among experts with at least 5 years of experience in the eld of HIS at national level.
In addition, experts needed to have a research background in technology or be a member of the headquarters or technical units of health information technology and health deputy of the Ministry of Health or medical universities for least two years. The participants were selected using snowball sampling method.
For the interviews, rst the interviewees were reached by phone or received a text containing the letter of introduction for research. The research title, purpose and con dentiality of information were explained to participants in person. Accordingly, 24 face-to-face interviews were conducted, 12 via telephone and 6 via email. All interviews were recorded and then transcribed. On average, each interview lasted between 30 and 60 min. All interviews were conducted personally by the researcher.
The content validity of the interview questions was con rmed based on the opinions of experts in this eld. During the interview, the researcher noted down a summary of the contents and ambiguous items. The notes were then corrected or modi ed, especially their key points. Immediately after each interview, the dialogues were transcribed. For data analysis, a ve-step framework analysis method was used , which included identifying a thematic framework, indexing, drawing tables, drawing maps, and interpreting (13). Data analysis began from the rst interview and continued along with other interviews.
The interview texts were read several times to gain an overview of the materials, and then the text was read line by line to gain a general ideal of the structure of sentences. From the outset, the rst level was coded, which involved extracting the analytical units from answers to the questions, Semantic units were extracted from main concepts in the analysis units and a code was assigned to each semantic unit. After comparing the codes, a list of main and sub-codes was obtained. In the second level of coding, the primary and secondary codes were read and then the primary codes that had the same meaning were classi ed to form classes.
To ensure the accuracy of the qualitative ndings, four criteria (validity or acceptability, reliability or similarity, transferability and veri cation) proposed by GABA and Lincoln were considered. The researcher attempted to ensure that the ndings re ect the real experiences of participants (15).
In the third stage, the validity of the proposed minimum dataset was assessed using the Delphi technique based on the opinions of 76 experts in this eld. Experts had at least one year of work experience in the Ministry of Health, were member of the headquarters or technical units of health information technology or faculty members at the departments of health information management in various universities of medical sciences. Finally, the Likert-type questions were developed based on the results of the qualitative section and sent to the observers in person or by e-mail along with an introduction to the goals of research. The participants were asked to write their comments in the questionnaire or submit their ideas as a corrective proposal in cases they did not agree with one of the concepts or components. In the end, the questionnaires were collected and analyzed. Each data element, relative to its importance, received a value of 0 to 100. Following the assessment of experts and analysis of data, options that gained less than 80% of the total points (out of 159 suggested elements, 19 did not attain the required points) were re-examined based on suggestions, and then the newly developed questions were sent for re-examination. Ultimately, 145 data elements were approved by experts. The results of this step were analyzed by SPSS15 software.

Results:
Based on the ndings of the comparative study, the minimum health information dataset of seven countries including the United States (14), Canada(15), the United Kingdom(16), Australia (17), Turkey(18), Malaysia (19), and Iran are listed in Table 1: Table (1) The minimum dataset of electronic health records information in the selected countries Table (1) The minimum set of electronic health records information data in the countries under study Common items in the electronic health les of the above countries included immunization, mental health, the elderly, maternal and infant care, and demographic information. Generally provided at the primary health care level, these services exhibit the importance of these cases in health promotion. Table 2 shows the general characteristics of the subjects participated in the qualitative phase.  Table 3. Table 3: Primary and sub-concepts of minimum electronic health record dataset based on the opinions of experts in the qualitative phase In the rst round, 19 items that had not obtained the required points were modi ed at the discretion of the research team and nally included in the survey.
17 items related to the source of income -amount of income, and existence of support programs, educational backgrounds in the main subgroup of the elderly -skin and eye color that indicates problems or illness, history of Jaundice in the main subgroup of children -maternal support programs, history of alcohol consumption, educational background in the main subgroup of mothers, the position of the drugprescribing person from the main para clinical subgroup, and the pharmacy in electronic referral, as well as the position of the drug-prescribing person in the main subgroup of medicine -source of income -and educational background were removed from the main mental health subgroup.
Three items of the history of spousal or children disease in the main group of mothers, the history of smoking in the main group of mental health and the serial number of vaccines in the main vaccination group were added to the list based on the participants' feedback. Finally, 145 data elements were approved in the second round of Delphi.

Discussion:
This study was conducted to determine the minimum dataset required in the electronic health record in the health system of Iran. Datasets provide a tool for recording the most relevant and up-to-date facts about a patient's health information. Therefore, such data provides valuable information for policymakers, health care professionals, and stakeholders that should be readily available to care providers (20). A review of the minimum dataset used in the seven countries studied revealed that the information in the e-health le is used on a case-by-case basis, depending on the country's geopolitical and health conditions. In general, the minimum dataset used in these seven countries manifested a strong emphasis on primary health care, which highlights the potentials of this type of care on health promotion, cost-effectiveness, and assurance of its effectiveness (21).
In developed countries (e.g. USA, Canada, UK and Australia), primary health care is concerned with the second and third health care, which is effective in designing a set of minimum datasets required for electronic health records (22). In developing countries (e.g. Turkey, Malaysia and Iran), e-health records are chie y at the primary health care and unrelated to the specialized levels (23). The Sib, Sina and Nab systems in Iran were designed based on primary health care service package provided by the Ministry of Health in 1995, which utilized the minimum standards in the content structure. There is a major difference between these systems in terms of the information they contain (standardized texts such as text elds) and how they interact with the software (2) According to the research ndings, the minimum dataset in different registration systems falls in different categories (39).
Demographic information is of high priority in all information systems under study, which re ects the importance of such information in the e-health record (24) The sources and amount of income in all three groups of mothers, the elderly and mental health were not con rmed, which could be explained in terms of cultural sensitivity of people to expressing their income in Iran.
All recommendations in the vaccination section were nalized, which indicates familiarity with the program and its long history in the health system. Given that children vaccination is one of the primary measures taken by the World Health Organization to eliminate pediatric diseases, it is also strongly emphasized in the Iranian health system (25). In all age groups, one of the main elements in the history of vaccination was that interviewees admitted that the Sib system, which is used nationally, does not contain information required by the vaccination program. Darabi et al. studied the design of a minimum dataset required for children. The vaccination data documented in the e-health records of the United States, Australia, Canada, and Iran was consistent with the data nalized in this study (26).
There were 33 items in the main concept of children of which 91% were con rmed. In all countries studied, children's health program is one of the main components of the electronic health record. From the perspective of interviewees, it is necessary to record all information about children sine their birth.
Childcare is one of crucial aspects of health care and has all the characteristics of priority selection in executive programs (27). Darabi et al. in a minimum dataset for children's appetite provided 146 data elements. All of the data elements approved in their study are consistent with the data obtained in the present study (26) Maternal health programs are a major part of the study program in seven countries. Since mothers and babies are considered as vulnerable groups in the society, maintaining and improving their health is the most basic element of health care in any country and one of the major indicators of health in sustainable development of maternal mortality index (27). Cultural issues have a strong impact on in uence the removal of items such as income and type of maternal support (28,29) 18% of items were removed from the original version of of mental health. As reported by the World Health Organization (WHO) in 1990, of 10 main diseases that caused disability in the world, ve were related to mental illness (30). It is mainly rooted in the misguided view as well as the irrational attitude of people, experts and political and health policymakers in the eld of mental health, inappropriate structure, lack of nancial credits in the Ministry of Health, worn-out primary health care network, and lack of specialized health care and treatment in mental health (31).
The rise of the elderly population and changes in the epidemiological pattern of diseases in middle age and old age to chronic diseases, more attention should be paid to preventive and therapeutic policies to maintain and improve the health of the elderly population (32). In the United States, the United Kingdom, Canada, and Australia, aside from health data, data from other health-related organizations are used to supplement electronic health data. In Iran, this extra-sectorial relationship in electronic health data has not yet been crystalized, and most data are based on people's statements. According to interviewees, information such as insurance type, source and amount of income, and the type of support programs for the elderly should be registered in the elderly health records (33).
93% of items related to the Para clinical section were approved by the participants. Since the family physician program and the referral system have not been implemented in the rural areas of Iran and there is still no connection between level one and level two and three of electronic services, the unfamiliarity of the target group may have affected the elimination of Para clinical information.

Conclusions:
In order to standardize the structure and content for data exchange in the electronic health le of patients in Iran, it is necessary to create a database and a glossary of national health information that explains the professional terms related to health and treatment. The important role of users and their information needs in designing an electronic health record system should be taken itno account. Also, a large dataset with a major gap between understanding and interpreting will be confusing. Hence, a minimum dataset of standards can lead to accurate and unambiguous access to the service recipients.

Declarations:
Ethics approval and consent to participate: this research is approved by Ethics Committee of Mashhad Give complete information about the purpose of re search to interviewees; 3) The comments of interviewees will remain con den tial