AI-Driven Innovations in NCD Management: Challenges and Solutions Based on Expert Perspectives in Iran’s Health System

AI-Driven Innovations in NCD Management: Challenges and Solutions Based on Expert Perspectives in Iran’s Health System | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article AI-Driven Innovations in NCD Management: Challenges and Solutions Based on Expert Perspectives in Iran’s Health System Marziye Hadian, Mohammadreza jabbari khanbebin, Mahan Mohammadi, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7225629/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Background: Artificial intelligence (AI) is widely regarded as a transformative technology in healthcare, particularly in the prevention of non-communicable diseases (NCDs). However, in developing countries like Iran, little is known about the readiness of health systems to effectively integrate AI. This study aimed to examine the challenges and benefits of AI implementation in NCD prevention from the perspective of Iranian health experts. Methods: This qualitative study employed conventional content analysis following the approach proposed by Graneheim and Lundman. Data were collected through semi-structured interviews with 34 experts specializing in medicine, health system management, medical ethics, and information technology. Participants were purposively selected from Type 1, 2, and 3 medical universities across Iran between March 2025 and July 2025. Data analysis was conducted using MAXQDA software (version 20). Findings: The study identified seven main themes and 30 sub-themes through content analysis, categorized into two key domains: (1) the advantages and opportunities of using artificial intelligence (AI) in non-communicable disease (NCD) prevention, and (2) the challenges and barriers to its implementation in Iran’s health system. Key benefits included enhanced primary care effectiveness, personalized interventions, resource optimization , and improved data-driven decision-making . Conversely, major barriers encompassed inadequate technological infrastructure , poorly structured data , ethical and legal concerns , cultural resistance , and a workforce unprepared for adopting new technologies . Conclusion: The effective and sustainable integration of artificial intelligence (AI) into Iran’s health system for non-communicable disease (NCD) prevention requires four key enablers : (1) robust technological infrastructure, (2) enhanced human capital, (3) well-defined legal and ethical frameworks, and (4) an adaptive organizational culture. Only by addressing these prerequisites can AI transition from a theoretical potential to a practical tool for policymaking and preventive interventions. Artificial Intelligence Smart Innovation Non-Communicable Diseases Prevention Health Policy Digital Health Iran Introduction In recent decades, non-communicable diseases (NCDs) - including diabetes, cardiovascular diseases, cancers, and chronic respiratory conditions - have emerged as the foremost global public health challenge. This burden is particularly acute in low- and middle-income countries (LMICs), where health systems often struggle to manage the growing epidemic(1) . According to the World Health Organization (WHO), non-communicable diseases (NCDs) accounted for 43 million global deaths in 2021, representing 75% of all mortality excluding pandemic-related deaths. Notably, 18 million of these deaths (42%) occurred prematurely in individuals aged under 70 years, with low- and middle-income countries (LMICs) bearing disproportionate burden—comprising 82% of these premature fatalities.(2). Iran faces a similarly severe NCD burden, with Ministry of Health reports indicating that these diseases account for approximately 90% of annual mortality and over 76% of the country's total disease burden (measured in disability-adjusted life years, DALYs)(3). The rising prevalence of modifiable risk factors - including obesity, physical inactivity, poor nutrition, tobacco use, and alcohol consumption - coupled with high rates of hypertension and diabetes, presents a growing public health crisis for future generations. These conditions not only threaten population health but also create substantial economic burdens through both direct healthcare costs and indirect productivity losses, straining Iran's health system and national economy.(4, 5) Iran currently faces a substantial burden of metabolic disorders, with over 7 million adults diagnosed with diabetes and an additional 13% of the adult population in the prediabetic stage. Compounding this challenge, hypertension affects more than 30% of Iranian adults, with concerning gaps in awareness as a significant proportion remain undiagnosed.(6, 7). Iran is experiencing a concerning upward trajectory in cancer and chronic respiratory disease rates. Current epidemiological projections suggest that, without intervention, NCDs will constitute over 90% of the nation's total disease burden and mortality by 2030 - representing both a public health emergency and a significant threat to sustainable development. (8).These alarming statistics underscore that Iran's NCD epidemic represents a dual crisis: a dire threat to population health and a formidable challenge to health system sustainability and national economic development. Left unaddressed, this growing burden threatens to undermine decades of public health progress and economic growth(9). Given this escalating public health emergency, conventional NCD prevention and control strategies - predominantly reliant on public education campaigns, periodic screening programs, and population-level interventions - have proven insufficient to manage the growing magnitude and complexity of this crisis. The limitations of these traditional approaches highlight the urgent need for innovative solutions capable of addressing the multifaceted nature of NCDs in contemporary healthcare systems.(10) Empirical evidence demonstrates that while comprehensive national NCD control programs have been implemented, their impact remains limited - disease prevalence continues to escalate and the growing burden shows no signs of abatement. The persistent discrepancy between program objectives and actual outcomes reveals fundamental shortcomings in current methodologies, compelling an urgent paradigm shift in our approach to NCD prevention and management(11). In this challenging landscape, artificial intelligence (AI) has emerged as a transformative solution for addressing the NCD crisis. By leveraging its unparalleled capacity for large-scale data analysis, complex pattern recognition, and predictive modeling, AI enables: (1) early detection of high-risk populations, (2) real-time health monitoring, and (3) development of precision public health interventions at scale. These advanced capabilities position AI as a powerful tool for revolutionizing NCD prevention and management strategies(12). AI-based predictive models demonstrate superior performance in NCD screening and risk assessment, achieving up to 30% greater accuracy than conventional approaches while simultaneously reducing healthcare expenditures. These advanced systems enable more efficient resource allocation through earlier and more precise identification of at-risk populations(13). Artificial intelligence has demonstrated significant potential in transforming NCD management globally. A notable example is the UK's AIRE-DM system, which analyzes electrocardiogram (ECG) data to predict type 2 diabetes risk with a 10–13-year lead time. The National Health Service (NHS) has recognized this innovation's value, with pilot implementation scheduled for 2025 as part of its diabetes prevention strategy(14). A prominent example of successful AI implementation comes from China's Hubei Province, where an AI-assisted cervical cancer screening program was deployed across clinical settings from 2017 to 2021. This large-scale application demonstrated AI's potential to enhance cancer detection in resource-constrained environments while maintaining diagnostic accuracy comparable to conventional methods.(15) These implementations demonstrate AI's transformative potential in preventive healthcare. A particularly successful case is Singapore's national diabetic retinopathy screening program, which incorporated deep learning algorithms to achieve a 20% cost reduction—lowering screening expenses from USD 77 to USD 62 per patient annually while maintaining diagnostic accuracy. This achievement highlights how AI can enhance both the efficiency and affordability of large-scale public health initiatives (16). These AI capabilities hold particular relevance for Iran's healthcare system, which must address a mounting NCD burden amidst significant resource constraints. However, successful AI integration for NCD management requires more than technological sophistication alone. Implementation effectiveness hinges on critical contextual factors including: Health system infrastructure and workflows, Cultural acceptance of AI-driven healthcare, Supportive policy frameworks and governance structures, and Alignment with existing public health priorities. (17) While Iran has made progress in health system digitalization, critical barriers persist—including fragmented data infrastructures, shortages of data science and medical ethics specialists, and inconsistent policymaking—that collectively constrain AI's healthcare potential. Notably, existing domestic research has disproportionately emphasized technical AI development while neglecting key stakeholder perspectives on implementation challenges from healthcare providers, IT specialists, ethicists, and policymakers. This study aims to address this critical knowledge and policy gap by highlighting both the quantitative burden of NCDs and the technological potential of AI, in order to inform the development of actionable strategies and effective policies toward transforming Iran’s health system ultimately contributing to public health promotion and the sustainability of national resources. Methods Study design: This qualitative study employed conventional content analysis following Graneheim and Lundman's methodological framework(18). This approach systematically examines textual data through four key phases: (1) identification of meaning units, (2) coding, (3) categorization, and (4) latent theme development, enabling deep interpretation of participant experiences. We selected this framework for its: Methodological flexibility in capturing complex phenomena Capacity to derive contextually grounded concepts Ability to minimize theoretical presuppositions Suitability for exploring multilayered research questions The study adhered to the COREQ (Consolidated Criteria for Reporting Qualitative Research) guidelines to ensure methodological rigor. This 32-item checklist evaluates three critical domains: Study Design : Research team composition, theoretical framework, and participant selection strategy Data Collection & Analysis : Interview procedures, coding processes, and software utilization Findings : Theme development, data interpretation, and reflexivity measures Key methodological strengths included: Systematic application of content analysis procedures Maintenance of an audit trail for analytical decisions Rigorous attention to contextual factors Comprehensive reporting following COREQ standards (19). Training of the Research Team: Prior to the commencement of data collection, all members of the research team, including interviewers and data analysts, received specialized training in the principles of qualitative interviewing, content analysis, and adherence to ethical considerations. Practice and pilot sessions were held to ensure coordinated interview execution and to minimize individual bias. The interview guide was revised and finalized following its pilot implementation on three participants and the collection of their feedback. Participants and Inclusion/Exclusion Criteria: We employed purposive sampling to recruit key informants representing four critical stakeholder groups: Clinical Specialists : Endocrinologists Cardiologists Social medicine specialists Epidemiologists Public Health Experts : Occupational health specialists Preventive medicine specialists Digital Health Professionals : Health information technology specialists AI implementation practitioners Health System Leadership : Medical ethicists Health policy makers Hospital administrators University health managers Inclusion Criteria : Minimum 3 years' experience in NCD prevention/management Direct involvement in digital health or AI initiatives Current affiliation with Iranian medical universities Willingness to provide in-depth perspectives Geographic representation across Iran's healthcare regions Sampling Rationale : This stratified approach ensured: Multidisciplinary perspective integration Balanced clinical/technical/policy expertise Grounded understanding of implementation challenges National-level generalizability of findings Exclusion criteria included: withdrawal from cooperation, provision of insufficient or low-quality data, and inability to complete the interview process. To ensure geographic and structural diversity, samples were selected from Type 1 universities of medical sciences (policy-making and research centers), Type 2 universities (executive centers with moderate research capacity), and Type 3 universities (institutions in underprivileged regions), in order to reflect the diverse realities of Iran’s health system. Sampling Method and Sample Size: The study employed a combined sampling strategy, beginning with purposive selection of key informants followed by snowball sampling to identify additional qualified participants. Initial recruitment focused on pre-identified specialists with demonstrated expertise in the research domain. Following each interview, participants were invited to recommend colleagues who possessed relevant experience and could provide valuable insights, thereby expanding the pool of informants through professional networks. The sampling process continued until reaching theoretical saturation, a point at which no new conceptual insights emerged from subsequent interviews. This threshold was initially observed following twenty-eight interviews, with six additional confirmatory interviews conducted to verify data completeness. The final sample comprised thirty-four participants, ensuring robust representation across key perspectives. Detailed demographic characteristics of study participants—including age distribution, gender representation, professional specialties, years of experience, and institutional affiliations—are provided in Appendix for comprehensive reference. This documentation enables readers to evaluate the sample composition and its relevance to the research context. Data Collection: Data collection occurred between March and July 2025 through semi-structured interviews. The interview guide, developed through a rigorous process involving literature review, expert consultation with three specialists, and pilot testing, covered five key thematic areas: Potential benefits and applications of AI in NCD prevention and management Technical and infrastructural implementation challenges Organizational and policy barriers Ethical considerations including privacy and public trust Health system readiness regarding professional, cultural, and educational factors Interviews were scheduled in advance and conducted either in-person or through secure digital platforms (Google Meet or Skype) to accommodate geographical and logistical constraints. Each session lasted 45-60 minutes and was audio-recorded following participant consent. All recordings were subsequently transcribed verbatim for analysis. Prior to each interview, researchers thoroughly explained: Study objectives and procedures Recording methods and data handling Confidentiality protections Voluntary participation terms Participants provided written or electronic informed consent before proceeding. All collected data (audio recordings and transcripts) were stored in encrypted formats with restricted access. In accordance with ethical protocols, these materials will be permanently deleted upon study completion. Data Analysis: The data analysis followed the steps of conventional content analysis as outlined by Graneheim and Lundman, utilizing MAXQDA 2020 software. Initially, the interview transcripts were read multiple times to ensure a thorough understanding of the data. Meaningful units relevant to the research objectives were then identified, and initial coding was conducted. These codes were organized into preliminary categories, which were subsequently merged and refined to derive the main themes and subthemes through conceptual inference. To ensure reliability, two researchers independently performed the analysis, with a third researcher acting as an arbitrator in cases of disagreement to reach consensus. To enhance the credibility of the findings, summaries of the results and initial codes were shared with participants, and their feedback was incorporated to refine and improve accuracy. Furthermore, MAXQDA software’s advanced features were employed for code management, data categorization, concept searching, and the creation of conceptual maps. Data Management: All audio data, transcribed texts, and data codes were stored in a secure and encrypted cloud environment, with access restricted only to the core members of the research team. After completion of the analysis and publication of the results, identifying data will be completely deleted to ensure the confidentiality and security of participants’ information. Data Trustworthiness and Validity: To ensure data credibility and quality, the study adhered to Lincoln and Guba’s four criteria: credibility, dependability, confirmability, and transferability. Credibility was established through dual independent data analysis, participant validation (member checking), and collaborative team reviews of codes and themes. Dependability was achieved by thoroughly documenting all data collection and analysis procedures, including any adjustments made during the research process. Confirmability was ensured by maintaining transparent records of analytical decisions, justifications for code and category selections, and active researcher involvement at every stage of coding and analysis. Finally, transferability was supported by presenting contextualized participant quotations and detailed descriptions of the study’s setting and conditions, allowing readers to assess the applicability of the findings to other contexts Addressing Researcher Bias: To reduce bias, all stages of coding and data analysis were carried out independently by two researchers, and any disagreements were resolved through joint discussion and, if necessary, with the opinion of a third person. Also, participants' feedback on preliminary findings was obtained and incorporated into the revisions to enhance the accuracy and validity of interpretations. Methodological Limitations: This study has several limitations. First, the findings may not be generalizable to other countries or health systems with different contexts. Second, potential biases could arise from sample selection and participant responses, while data interpretation may be influenced by the researchers’ expertise. Additionally, access to certain specialists was limited, and some perspectives may not have been fully expressed due to professional or ethical constraints. These factors should be considered when interpreting the results Findings Demographic Characteristics of the Participants: In this study, a total of 34 specialists from various sectors of the health system participated. According to Table 1, the average age of the participants was 45 years (with an age range of 35 to 62 years), and their average length of professional experience was 16 years. The participants included 21 men and 13 women in terms of gender. These individuals were categorized into four main areas of expertise and were selected from Type 1, 2, and 3 medical universities across the country to provide more diverse and comprehensive perspectives from different levels of the health system. Among all participants, 8 individuals were physicians and specialists in non-communicable diseases (5 men, 3 women), including specialists in social medicine, endocrinology, and cardiology; 4 were health information technology specialists (3 men, 1 woman) who had experience participating in artificial intelligence or digital health projects; 4 were professors or researchers in medical ethics (1 man, 3 women) from reputable academic centers in the country with experience in technology ethics and health policymaking; and 8 were managers, decision-makers, or policymakers in the health system (4 men, 4 women), including university administrators, members of strategic committees, and individuals with managerial experience at the central levels. 4 were experts or researchers in public and occupational health (3 men, 1 woman), and 4 were epidemiology specialists (4 men, 0 women) who participated in the study. In terms of university distribution, 12 individuals were from Type 1 universities, 11 from Type 2 universities, and 11 from Type 3 universities. This composition was designed in a way to realistically reflect the differences in infrastructure, resources, policies, and real-world experiences at the three academic levels in the country. Participants were also selected from various regions of the country, including the north, south, center, west, and east, to ensure the geographical richness of the data. Most participants held postgraduate degrees and had experience participating in policymaking projects, digital health, or teaching in fields related to health and technology. This composition strengthened the reliability and analytical depth of the extracted data. Main Findings: After content analysis of the data using the Graneheim & Lundman approach and MAXQDA software, a total of seven main themes and 30 subthemes were extracted from the interview transcripts. These themes cover two major areas: the advantages and opportunities of using artificial intelligence in the prevention of NCDs, and the challenges and barriers to implementing this technology in the country's health system. Table 2 presents the benefits of using artificial intelligence in the prevention of non-communicable diseases and Table 3 illustrates the challenges of using artificial intelligence in the prevention of non-communicable diseases. Theme 1: Enhancing the Effectiveness of Primary Care This theme explores the opportunities and potential capacities of artificial intelligence in strengthening the prevention system for non-communicable diseases (NCDs). Participants from various fields—including medicine, health information technology, and health system management—emphasized that AI, through its ability to analyze vast, complex, and multi-source data, can play a key role in the early identification of at-risk individuals. According to them, the use of advanced machine learning algorithms and data analytics enables the delivery of personalized preventive interventions and significantly improves the effectiveness of screening, monitoring, and management programs for NCDs. These capabilities, particularly within the context of Iran’s health system—which faces challenges such as resource limitations and the growing burden of chronic diseases—can drive a transformative shift in policymaking and the implementation of preventive initiatives. 1-1: Disease Risk Prediction Using Machine Learning Algorithms Participants highlighted the potential of AI in more accurately predicting the risk of non-communicable diseases, emphasizing that machine learning algorithms can detect complex and hidden patterns in data that are difficult for humans to identify. According to the interviewees: “The reality is that using machine learning algorithms, we can find highly complex patterns in data that even doctors might not easily notice. This allows us to identify high-risk patients much earlier and act more quickly.” (Interviewee No. 1) “One of the advantages of these algorithms is that they can continuously update themselves, and the more data we provide, the better their predictions become. If we can collect more local data, these models can function more accurately and be better adapted to our own conditions. Basically, the more data, the more reliable the output.” (Interviewee No. 7) 1-2: Early Identification of Risk Factors in High-Risk Populations Experts repeatedly emphasized the importance of AI's ability to detect early signs and risk factors of diseases. They believed that AI can accurately analyze individuals’ health data to identify early indicators and trends that often go unnoticed in conventional diagnostic approaches. This capability enables timely and targeted prevention, allowing healthcare providers to plan appropriate care and interventions before disease progression. According to participants: “In my opinion, one of the strengths of AI is that it can identify individuals who are usually overlooked in traditional systems. This is especially critical for underserved areas where access to medical services is limited.” (Interviewee No. 10) “To explain it simply, when we analyze community health data correctly, we can identify high-risk groups with much greater accuracy, and then design and implement interventions specifically for them. This improves both resource efficiency and overall effectiveness.” (Interviewee No. 23 ) 1-3: Increasing Access to Data-Driven Preventive Interventions Some participants believed that AI has the potential to significantly expand access to preventive services—especially in underserved and remote areas—through digital solutions and intelligent interventions. They emphasized that AI can help bridge existing gaps in healthcare delivery and provide vulnerable populations with access to preventive care. “AI, through apps and smart systems, can deliver preventive education to people who might never have access to such information in remote areas. It's like having a health educator with them all the time.” (Interviewee No. 12) “When access to preventive services becomes easier especially for those with fewer resources it can reduce health inequalities. AI can really be a powerful tool in closing those gaps.” (Interviewee No. 26) 1-4: Enhancing Efficiency of Primary Care Through Intelligent Support Experts highlighted the role of AI in reducing the workload of primary care staff. Through automation of time-consuming tasks such as symptom monitoring, data entry, and initial screenings, AI allows healthcare workers to focus more on patient care. Additionally, AI’s clinical decision support can improve diagnostic accuracy and guide appropriate interventions. Overall, they believed AI enhances both the quality and effectiveness of primary healthcare. “One major advantage of AI is that it can handle repetitive tasks like test reminders or analyzing patient data, freeing us to focus on clinical interventions and more specialized care.” (Interviewee No. 30) “This intelligent support really helps doctors make better decisions and manage patients more accurately and timely. It noticeably improves the quality of care in daily practice.” (Interviewee No. 5) 1-5: Empowering the Public Through Personalized Technologies Participants noted that AI can raise public awareness about personal health by providing tailored information and recommendations based on individual data. They emphasized that this approach makes the information more relevant and actionable, boosting people’s motivation and engagement in managing their own health. According to them, AI can shift individuals from passive recipients to active participants in their health management. “AI-based technologies can deliver health education exactly tailored to each person’s needs. That way, people get the info that’s actually useful for them and can take a more active role in their own care.” (Interviewee No. 28) “Empowerment is only valuable when it respects privacy and autonomy. As long as personal data is protected and used with consent, AI can truly improve public health and help people become more proactive in managing their well-being.” (Interviewee No. 14) Theme 2: Improving Decision-Making and Personalizing Preventive Interventions This theme specifically focuses on the role of artificial intelligence in improving the quality of clinical decision-making and developing personalized interventions for the prevention of non-communicable diseases. Specialists in the fields of medicine, information technology, and health management presented valuable perspectives on how artificial intelligence can help enhance preventive outcomes. They believe that artificial intelligence, through precise analysis of health data and providing individual insights, can assist physicians in making more accurate decisions and tailoring therapeutic interventions according to each patient's condition. This approach not only increases the effectiveness of prevention, but it can also help reduce the burden of chronic diseases in society and improve the overall quality of healthcare. 2-1: Personalizing Preventive Programs Based on Individual and Environmental Data According to participants’ opinions, artificial intelligence, by combining and analyzing genetic, behavioral, and environmental data, provides the necessary foundation for formulating precise health programs tailored to individual characteristics. This personalization-based approach can significantly increase the effectiveness of preventive interventions and also enables optimal allocation of health resources. According to the participants: “A general prevention program really doesn’t work for everyone. Artificial intelligence has the capability to design a more specific and precise program for each person, based on their medical history, habits, and living environment, which certainly has much greater effectiveness.” (Interviewee No. 20). “I think right now we have the possibility to combine various data from different sources like apps and even medical records, and by analyzing them, we can offer the best preventive strategy for each person. This makes the recommendations much more precise and tailored to each individual’s condition.” (Interviewee No. 3) 2-2: Clinical Decision Support with Intelligent Tools and Analytical Algorithms Participants believed that artificial intelligence–based tools can assist physicians in making faster and more accurate decisions, especially when facing complex conditions or a high number of patients. This technology, by analyzing large volumes of data and offering timely suggestions, can significantly improve the quality of clinical decision-making. According to the participants: “I think one of the good things about intelligent algorithms is that they can precisely show us which patients need urgent intervention and which ones can be managed with simpler programs. This really makes things easier for us and makes decision-making faster and more effective.” (Interviewee No. 8). “These tools are not supposed to replace the doctor’s judgment, but rather act as an assistant beside them and help increase the accuracy of decisions. In fact, the goal is that with the help of these technologies, doctors can make better and more confident decisions.” (Interviewee No. 13) 2-3: Reducing Human Errors in Risk Assessment and Determining Preventive Strategies Participants believed that artificial intelligence, by reducing errors caused by fatigue and high workload, increases the accuracy of clinical evaluations and helps physicians make better decisions in complex situations. According to the participants: “When doctors are under high work pressure, the likelihood of mistakes increases. That’s where algorithms can really be helpful and reduce errors so the patient receives the best care.” (Interviewee No. 30). “In my opinion, one of the main advantages of artificial intelligence is that it can analyze data with much higher accuracy than humans, and that significantly reduces the likelihood of human errors.” (Interviewee No. 9) 2-4: Improving Continuous Remote Monitoring of Patients and Predicting Disease Progression Participants emphasized that artificial intelligence will have the capability to continuously monitor patients’ health status using data obtained from wearable devices and sensors and predict the progression of diseases. This capability enables early diagnosis and the provision of personalized treatments, and ultimately helps improve the quality of medical care. “I think continuous monitoring really helps us intervene before the disease reaches a severe stage. This is especially important in areas where access to healthcare centers is limited and can save many lives.” (Interviewee No. 11). “If artificial intelligence were implemented, it would enable us to have up-to-date data. These live and real-time data really allow us to respond faster and better to changes in the patient’s condition. This makes care more accurate and timelier, and ultimately yields a better outcome for the patient.” (Interviewee No. 15 ) 2-5: Facilitating Communication Among Care Teams Through Shared Data Analysis Participants believed that if artificial intelligence is implemented, especially in complex healthcare systems, it can improve coordination and communication among care team members. This technology, by facilitating fast and accurate information exchange, increases the quality of care and reduces potential errors. Furthermore, artificial intelligence helps in managing treatment plans and patient follow-up and enhances team coordination, which is especially important in challenging environments. According to the participants: “When all team members have access to accurate and integrated information, their collaboration becomes much better, and ultimately the services provided to patients become higher in quality. This really plays an important role in improving the care process.” (Interviewee No. 27). “If artificial intelligence is implemented, one of its good features is that it can summarize complex data in a simple and understandable way for all team members. This reduces the time needed for decision-making and allows the team to act more quickly.” (Interviewee No. 22) 2-6: Enhancing the Quality of Health Decision-Making and Policy-Making The study showed that artificial intelligence, by analyzing large volumes of big data, can assist policymakers in making more accurate and evidence-based decisions. This capability helps improve the quality of decisions and the design of more effective health policies. According to the participants: “When the data is up-to-date and the analysis is done intelligently, we can design more precise and effective prevention programs. This way, the success rate of the programs increases, and resources are used more efficiently.” (Interviewee No. 13). “Artificial intelligence is a really good tool for helping us better identify priorities and allocate resources more intelligently. This helps decisions become more accurate and efficient.” (Interviewee No. 6) Theme 3: Resource Optimization This theme explores the opportunities and benefits of artificial intelligence in improving the prevention and management of non-communicable diseases. Medical professionals, IT experts, and health system administrators emphasized the role of AI in early diagnosis, disease trend prediction, and the delivery of personalized treatments. They also pointed to enhanced healthcare system efficiency, cost reduction, and increased accessibility for patients. By analyzing big data, AI enables the design of more effective preventive programs. 3-1: Improving Diagnostic Accuracy to Optimize Health Resources Participants believed that AI, by leveraging big data analytics, is capable of providing early diagnoses and more accurate predictions. They also emphasized that this capability can significantly improve the quality of healthcare and help reduce treatment costs. According to participants: "Based on the workshops we attended, AI has the ability to identify disease patterns at early stages. This is very important because it enables better prevention and thus reduces disease burden. In my opinion, this is one of the greatest advantages of AI in healthcare that we should not overlook." (Interviewee No.18). "With AI, we can predict the course of a disease and intervene on time before it becomes more serious." (Interviewee No. 12) 3-2: Increasing Access to Healthcare Services and Reducing Costs One of the key advantages of AI is the expansion of preventive services to remote areas with limited access to healthcare. This technology can reduce costs and make services more accessible for underserved populations, thereby promoting health equity. According to participants: "AI enables us to deliver preventive services more easily and affordably, especially in underserved areas." (Interviewee No.19). "AI helps patients receive basic care at home or in closer local centers, reducing the need for hospital visits." (Interviewee No.26) 3-3: Improving Data Management and System Integration Participants believed that AI, by organizing and analyzing large volumes of health data, can enhance coordination between different sectors. According to participants: "AI can gather scattered data and provide important, actionable information to support better decision-making. This means we can make decisions with a broader and more precise view." (Interviewee No.24). "When data is integrated, the system becomes more responsive and the care process runs more smoothly. This greatly contributes to the quality of health services." (Interviewee No.18) Theme 4: Education and Workforce Development This theme emphasizes the importance of training and empowering the workforce to effectively and sustainably utilize smart technologies in the healthcare system. Participants highlighted various educational needs and professional development opportunities, stressing the necessity of enhancing staff knowledge and skills for optimal use of these technologies. 4-1: Need for Specialized Training in AI and Emerging Technologies Participants believed that specialized training and skills development in AI are essential for physicians, nurses, and health administrators. They emphasized that such education plays a critical role in improving performance and productivity in the health system. According to participants: "Those working in the health sector need to be familiar with the basics of AI so they can use it properly." (Interviewee No. 17) "If physicians lack sufficient knowledge, they won't be able to trust or correctly apply data-driven decisions. That's why education and familiarity with these technologies are crucial for effective utilization." (Interviewee No. 24) 4-2: Importance of Interdisciplinary Education and Team Collaboration Participants emphasized that joint training across medicine, information technology, and health management is essential to foster effective and coordinated collaboration among multidisciplinary teams. This approach can enhance performance and improve the quality of healthcare services. According to participants: "When interdisciplinary training is provided, mutual understanding among team members improves, and so does their coordination. This leads to more effective and impactful collaboration." (Interviewee No.22) "Without teamwork and coordination, AI projects usually face serious challenges and are more likely to fail. For success, all members need to work together hand in hand." (Interviewee No. 2) 4-3: Continuous Knowledge Updating Participants stressed the importance of continuously updating knowledge and skills in smart technologies. They believed this is essential for maintaining efficiency and productivity in using such technologies. According to participants: "If AI is going to be integrated into our work, we must recognize that it's evolving rapidly. Therefore, we must always provide opportunities for continuous, up-to-date training so professionals can keep pace with these changes." (Interviewee No. 5) "Participants believed that professional development programs should be ongoing and embedded within the organizational culture to ensure sustainable learning and growth. This approach strengthens collaboration, motivation, and performance improvement within organizations." (Interviewee No. 14) 4-4: Creating Motivation and Acceptance of Technology through Education Participants believed that appropriate education can reduce resistance to technology and increase the motivation to adopt it. According to participants: "When training is done properly, people are more likely to trust AI and be willing to use it. This trust plays a key role in the successful adoption and implementation of technology." (Interviewee No. 23) "When people become familiar with the benefits and applications of AI, their resistance to it decreases and they are more likely to accept it. This awareness fosters trust and acceptance of AI in workplace and care settings." (Interviewee No. 14) Theme 5: Infrastructural and Technical Challenges This theme highlights the key infrastructural and technical barriers to implementing artificial intelligence (AI) in the country's health system, particularly in the area of non-communicable disease (NCD) prevention. IT specialists, physicians, and health system managers emphasized several major challenges, including the lack of high-quality and reliable data which forms the foundation of any successful AI system, inadequate technical infrastructure for integrated data collection, storage, and processing, and serious concerns around data security and patient privacy. These issues significantly limit the widespread implementation and effective use of AI in healthcare. Addressing them will require careful planning, technological investment, and the training of specialized human resources to fully realize the potential of AI. 5-1: Lack of High-Quality and Structured Data for Algorithm Training Participants believed that one of the biggest barriers to AI development in healthcare is the shortage of accurate, complete, and structured data needed for training and optimizing algorithms. Without such data, the performance of intelligent systems decreases, limiting their ability to provide precise and trustworthy decisions. “The reality is that our current data is fragmented, incomplete, and mostly stored in unstructured formats. Without sufficient and accurate data, algorithms can't perform well, and their outcomes won’t be reliable.” (Interviewee No.29) “In our system, data is still mostly kept in traditional ways, making it difficult to extract and use for AI. That’s a major challenge we need to address.” (Interviewee No.4) 5-2: Weak IT Infrastructure in Health Centers Experts pointed out the absence of robust and comprehensive infrastructure—including advanced servers, secure networks, and integrated data management systems. This lack significantly hinders AI implementation and prevents the full utilization of its capabilities. Without such infrastructure, handling large volumes of medical data and ensuring efficient information exchange across sectors becomes difficult, reducing healthcare quality. According to participants: “The fact is, many health centers lack the necessary technological infrastructure for advanced systems. Without servers, secure networks, and integrated platforms, AI efforts are likely to fail.” (Interviewee No.22) “To use AI safely and effectively, we need to upgrade both our hardware and software. Without these updates, full adoption of the technology isn’t possible, and system security could be compromised.” (Interviewee No.9) 5-3: Security and Data Privacy Concerns Experts viewed security and privacy as central issues in AI usage. Without strong security systems, data leaks are more likely and patient trust may be lost slowing down AI adoption. Therefore, ensuring data confidentiality is a top priority. According to participants: “It’s absolutely critical that patient data is fully secured and protected from unauthorized access especially when stored digitally and centrally. If this isn't addressed, major problems could arise and public trust would be undermined.” (Interviewee No.4) “Protecting privacy must always come first. If we fail here, people won’t trust AI systems, and they’ll hesitate to share their data. That could derail the entire effort.” (Interviewee No.15) 5-4: Shortage of Skilled Professionals and Adequate Training Participants emphasized the lack of trained professionals in AI, which slows down implementation. “Most healthcare staff are still unfamiliar with new technologies, and there isn’t enough training. This makes effective use of AI difficult. We need targeted, up-to-date training programs to build staff capabilities.” (Interviewee No.16) “To succeed, teams must be composed of up-to-date experts who receive continuous training. That way, necessary knowledge and skills are preserved, and projects move forward more effectively.” (Interviewee No.2) 5-5: Lack of Clear Legal and Policy Frameworks One of the key challenges is the absence of clear regulations and guidelines regarding AI use and health data governance. This leads to ambiguity around privacy, accountability, and performance standards hindering the technology's growth. “Current laws aren’t adequate for the complexities of AI. We need more updated and comprehensive frameworks to use AI effectively and manage security and privacy risks.” (Interviewee No.10) “We all know that without solid legal frameworks, it’s hard to use intelligent technologies safely and responsibly. These laws must protect data and privacy—otherwise, there’s a high risk of misuse and loss of public trust.” (Interviewee No.2) 5-6: Organizational Culture and Resistance to Change Participants noted that resistance from healthcare staff and managers to adopt new technologies is a major obstacle. This is often due to inadequate training and fear of disrupting established routines. They stressed that regular training and institutional support can ease this resistance. “Some colleagues are really hesitant to let go of traditional methods and trust smart systems. It’s mostly about habits and the fear that the old ways are safer and change is risky.” (Interviewee No.8) “To improve technology adoption, organizational culture must shift, and everyone needs to be prepared for change.” (Interviewee No.13) Theme 6: Ethical and Social Considerations This theme focuses on ethical concerns, social implications, and human consequences of using AI in healthcare. Medical ethicists, physicians, and health managers raised issues such as data privacy, potential bias, accountability, and broader social impacts. 6-1: Protection of Patient Privacy and Data Confidentiality Participants emphasized the importance of protecting patients' personal information and expressed concerns that sensitive data might be misused or exposed. According to participants: “Data confidentiality is a red line for us. If we lose patients’ trust, we can’t use these technologies effectively—everything will fall apart. So, we must ensure their information remains fully protected.” (Interviewee No.11) “Patients need to understand how their information is used and be assured that it’s secure.” (Interviewee No. 16) 6-2: Concerns About Bias and Inequity in AI Algorithms Some participants raised concerns about unintended discrimination in AI algorithms that might exclude certain groups from accessing services or accurate diagnoses. They stressed the need for ongoing monitoring and correction to prevent such bias. According to participants: “If the input data is biased, the algorithm output will also be unfair—and may result in discrimination. We must ensure data quality to protect equity.” (Interviewee No.18) “We have to be very cautious that AI doesn’t widen health disparities. If mismanaged, it could leave already underserved groups even further behind. Ensuring fairness must always be a priority.” (Interviewee No.22) 6-3: Accountability for AI-Based Decisions Participants emphasized the need for clear responsibility in cases of AI decision errors. They acknowledged that, given the technical and legal complexities, assigning accountability is not easy, but necessary. According to participants: “When an algorithm makes a mistake, we need to know exactly who’s accountable—is it the doctor making the final call, the software developer, or the entire health system? Without clear answers, we can’t address errors properly or maintain public trust.” (Interviewee No.5) Theme 7: Technology This theme addresses the role of information technology infrastructure in either enabling or limiting the implementation of artificial intelligence (AI) in the healthcare system. Participants referred to challenges and opportunities related to both hardware and software infrastructure, emphasizing the importance of developing appropriate platforms. 7-1: Lack of adequate and up-to-date technological infrastructure Participants believed that the lack of modern equipment and updated technologies hinders the effective implementation of AI and creates limitations in this area. According to participants: "Our infrastructure isn't sufficient for using AI. We need serious investment and must upgrade our equipment to better utilize this technology." (Participant No. 8) "Without strong and reliable infrastructure, you can’t expect AI to perform well or deliver results. Infrastructure is like a foundation—everything is built on top of it." (Participant No. 27) 7-2: The necessity of data and system standardization Participants emphasized that developing data standards and creating interoperability between various systems plays a crucial role in improving performance and integration of technologies. According to participants: "Data must be collected and stored in a standardized and unified format so that AI can analyze it properly and deliver more accurate results. This is very important for the success of AI projects." (Participant No. 30) "When no standard exists, data becomes inconsistent, and this leads to problems in analysis and unreliable outcomes." (Participant No. 11) 7-3: Effective use of modern technologies and cloud platforms Some participants pointed to the advantages of modern and cloud-based technologies, which can reduce hardware limitations and offer better capabilities for AI implementation. According to participants: "Cloud platforms allow us to access sufficient processing power without having to purchase expensive hardware, making AI project implementation much easier." (Participant No. 21) "When we use new technologies, not only is the speed of delivering preventive services increased, but the quality also improves. This enables us to respond more effectively to people’s needs." (Participant No. 2) Discussion This qualitative study explored the challenges and opportunities of implementing artificial intelligence (AI) in non-communicable disease (NCD) prevention from the perspectives of experts across multiple disciplines, including healthcare, information technology, health system management, and medical ethics. While findings reveal widespread professional optimism about AI's potential benefits, they also highlight significant structural, technical, ethical, and cultural obstacles to its effective implementation within Iran's healthcare system. This research incorporates perspectives from universities across different tiers (Types 1, 2, and 3) nationwide, providing a multidimensional assessment of intelligent technologies' current status, opportunities, and challenges within the healthcare system. The study's significance stems from Iran's healthcare system being at a nascent stage of AI implementation, where emerging digitalization initiatives coexist with the need for coherent strategies and targeted investments. Using conventional content analysis, we systematically examined professional perspectives on AI's potential in healthcare. While many identified challenges align with global discourse, Iran's healthcare system - characterized by its centralized structure, resource constraints, and socio-environmental diversity - presents unique contextual dimensions that demand localized solutions. The analysis examines each major theme in depth, with findings contextualized through comparison with both domestic and international research. This approach highlights both universal concerns in AI healthcare implementation and country-specific considerations that require tailored approaches. Theme 1: Enhancing the Effectiveness of Primary Care Interviewees viewed artificial intelligence as a powerful tool for risk prediction, screening, and designing early interventions. They believed that using AI in prevention could lead to resource optimization, improved planning, and better health outcomes. This perspective aligns with international studies. For example, in a study conducted by Topol in 2019 in the UK, findings showed that AI, through processing large datasets, was able to predict the risk of cardiovascular and metabolic diseases more accurately than traditional models( 20 ). In a similar study conducted in China by Guan and colleagues in 2023, the findings indicated that artificial intelligence, as an emerging technology in the field of digital health, provides a significant opportunity to improve the efficiency of diabetes care. The use of AI can contribute to more effective diabetes prevention by optimizing disease management, enhancing patient self-care, and reducing treatment costs. Moreover, the development of an AI-based digital care ecosystem that includes both prevention and management of diabetes could be an important step toward reducing the disease burden and improving population health outcomes.( 21 ) In Iran, due to limited access to high-quality and structured data, the full potential of these capabilities has not yet been realized. However, participants believed that with the development of data infrastructure, artificial intelligence could play a strategic role in reducing the prevalence of non-communicable diseases. Their emphasis on the early identification of risk factors using intelligent tools also aligns with the WHO's 2021 report, which highlights “smart and personalized prevention” as a key strategy for controlling the epidemic of non-communicable diseases( 22 ). Theme 2: Enhancing Decision-Making and Personalization of Preventive Interventions Participants recognized artificial intelligence's potential to enhance decision-making accuracy for both individual patients and population health management. By analyzing demographic, personal, and biological data, AI could enable more precisely targeted preventive interventions. These findings corroborate research by Beam et al. (2018) in the United States, which demonstrated AI's effectiveness in developing personalized treatment plans and prevention strategies based on individual biological profiles and lifestyle factors. Experts further highlighted AI's particular value in optimizing healthcare resource allocation, especially for high-risk populations and underserved areas with limited access to services. This capability could help address critical disparities in healthcare delivery while improving overall system efficiency ( 23 ). In a study conducted by Patel and colleagues in India in 2022, the findings showed that AI algorithms helped the government accurately identify regions at high risk for cardiovascular diseases and direct preventive resources to those areas. This data-driven approach significantly improved the efficiency of public health programs( 24 ). While recognizing AI's considerable potential, Iranian experts expressed concerns regarding the scarcity of accurate, locally-relevant data needed for developing targeted interventions. Participants emphasized that creating integrated, comprehensive databases represents a fundamental prerequisite for effectively implementing AI in preventive healthcare. Such infrastructure would enable optimal service allocation tailored to local epidemiological patterns and demographic characteristics. Establishing these robust data systems could facilitate the development of intelligent, personalized solutions for diabetes prevention and management while addressing current implementation barriers. Theme 3: Resource Optimization Study participants identified several strategic advantages of artificial intelligence for preventing non-communicable diseases. AI applications can: Expand access to preventive services Enhance health system efficiency Enable early detection of high-risk individuals Reduce healthcare costs Facilitate predictive analytics using local data Experts particularly emphasized AI's ability to improve both the accuracy and speed of risk identification. Machine learning algorithms can effectively analyze multiple risk factors simultaneously, including: Personal health metrics Lifestyle patterns Family medical history Other relevant biomarkers This analytical capability proves especially valuable for conditions like diabetes, hypertension, and prevalent cancers that typically develop through prolonged preclinical stages. These findings align with research by Takahashi et al. (2017), whose Japanese study demonstrated superior sensitivity of AI models in predicting type 2 diabetes risk compared to conventional screening methods ( 25 ). The interviewees also noted that artificial intelligence can help reduce both the direct and indirect costs of the healthcare system, as early detection and timely interventions can prevent the development of costly complications. In a study conducted by Areia in 2022 in the United States, findings showed that the use of smart tools in screening colonoscopy reduced the cost per patient from $ 3,400 to $ 3,343, resulting in approximately $ 290 million in annual savings at the population level( 26 ). This point is also aligned with a report by the World Health Organization (WHO) in 2022, which highlighted the potential of artificial intelligence to reduce the economic burden of non-communicable diseases in low- and middle-income countries( 27 ). Despite the emphasis of this study and other research on the role of artificial intelligence in reducing healthcare system costs, some studies conducted in the United States in 2019 have pointed to increased initial expenses due to the implementation of technology, training of human resources, and the development of infrastructure( 28 , 29 ). However, these initial expenses can be compensated in the long term through disease burden reduction and improved system efficiency. Another significant advantage was the enhancement of access to preventive services in underserved areas. Many experts believed that by utilizing smart tools such as health-focused applications, preventive chatbots, and population-tracking algorithms, it is possible to expand the delivery of educational, screening, and follow-up services in remote regions. In a study conducted by Alaran in Africa in 2025, findings showed that the use of AI-powered chatbots for hypertension risk assessment successfully led to the identification of hundreds of undiagnosed cases of silent hypertension( 30 ). Participants also emphasized that one of the key advantages of artificial intelligence is its ability to assist policymakers by providing interpretable and analytical information. According to them, AI dashboards can be used to analyze disease trends and patterns, which in turn enables the design of more effective preventive policies. This approach is currently being implemented in several European countries, including the Netherlands and Denmark, and has shown positive results in the control of obesity and type 2 diabetes( 31 ). If properly implemented, artificial intelligence can play an unparalleled role in enhancing prevention, optimizing resource use, increasing access, reducing costs, and improving decision-making in the health system. To realize these benefits, it is essential to develop localized algorithms, provide specialized training for healthcare professionals, and establish reliable technological infrastructures. Theme 4: Education and Development Most interviewees believed that the successful implementation of artificial intelligence in the prevention of non-communicable diseases requires serious capacity building of human resources at various levels. They pointed out the lack of formal AI training for healthcare professionals, weak interdisciplinary collaboration, and insufficient awareness among managerial and service-level personnel. The lack of preparedness among human resources is one of the structural barriers to implementing new technologies. The findings of this study are in line with an international study by Mesko et al. (2020), which showed that in most countries, even physicians have limited understanding of the practical applications of AI and generally do not trust it( 32 ). In Iran, although numerous AI-related training workshops have been held at some universities in recent years, interviewees stated that the lack of coherent, systematic, and formal educational content in the curricula of medical sciences remains one of the main challenges in developing AI applications in the country's health system. Some interviewees pointed to a lack of mutual understanding between health professionals and data engineers. They emphasized that, to develop effective systems, physicians need to understand the basics of data and algorithms, and conversely, technology developers should possess at least a minimal knowledge of the health system and preventive interventions. On the other hand, several health system managers stated that even at high-level management positions, there is insufficient understanding of the benefits and risks of AI, which leads to delays in making technology-related decisions. This issue was also examined in a study conducted by Kim et al. in the United States (2024), which found that poor digital literacy among policymakers is one of the critical barriers to implementing digital transformation projects( 33 ). The realization of AI in the prevention of non-communicable diseases requires substantial investment in specialized training and the enhancement of digital literacy among physicians, managers, and health professionals. Designing interdisciplinary courses, establishing new roles within the health system, and integrating technological skills into higher education curricula in health sciences are essential prerequisites for the success of this technological transformation. Theme 5: Infrastructural and Technical Challenges Study participants consistently identified critical technological barriers hindering AI adoption in Iran's healthcare system: Inadequate IT infrastructure Absence of integrated data systems Poor electronic documentation practices Limited availability of structured data Insufficient computational resources The research highlights that structured, integrated, and high-quality data form the essential foundation for effective AI implementation. However, Iran's health information currently exists in fragmented, unstructured formats, with significant portions still paper-based. These data limitations compromise machine learning algorithm performance and analytical accuracy. This challenge mirrors global experiences in healthcare AI adoption. Rajkomar et al. (2019) found that even in advanced systems like the U.S. healthcare sector, inconsistent EHR data standardization remains a major obstacle to achieving optimal AI model performance. ( 34 ). Also, in an article published by Wang in China in 2021, the findings showed that although China is a leader in artificial intelligence, the lack of data integration among hospitals remains one of the country’s major national challenges ( 35 ). Moreover, the lack of cloud infrastructures and the inability of systems to store and process large volumes of data were clearly highlighted by IT experts and health system managers in Iran as serious barriers. This issue is also comparable to a study conducted in low-income African countries by Andigema et al. in 2024, which reported that weak infrastructural capacity has led to the failure of AI-driven projects at the national level( 36 ). In addition, concerns were also raised regarding cybersecurity and the protection of patient information. Participants stated that under current conditions, not only is there a lack of adequate hardware infrastructure, but also the security protocols needed to safeguard sensitive data have not been properly designed. These concerns were similarly highlighted in a study conducted by Price and colleagues in the United States, where the need for clear regulations governing the use of artificial intelligence in healthcare was strongly emphasized( 37 ). The lack of adequate information technology infrastructure, the absence of comprehensive and structured health databases, limited capacity for big data processing, and security concerns are among the most significant challenges hindering the effective use of artificial intelligence in the prevention of non-communicable diseases in Iran. Addressing these challenges requires national-level action in policymaking, investment in technology, the development of data standards, and the enhancement of information security within the country’s health system. Theme 6: Ethical and Social Considerations Interviewees pointed to a range of ethical and social challenges that could seriously hinder the safe adoption and acceptance of AI technologies within the health system. These challenges include concerns about patient privacy, lack of algorithmic transparency (explainability), the risk of algorithmic discrimination, and the absence of clear legal accountability in case of errors. One of the main ethical concerns, repeatedly emphasized in most interviews, was the issue of patient privacy and data confidentiality. Given the lack of appropriate data protection structures in Iran, specialists expressed concern that deploying AI without clear regulatory frameworks could lead to the exposure of patients' sensitive information. This concern closely mirrors warnings raised in official documents such as the WHO report (2021)( 22 ) And also the study by Price et al. in 2019. And they were also addressed in the study by Price et al. in 2019.( 37 ) It has been emphasized that implementing AI without adherence to the principles of trust, informed consent, and transparency can lead to public distrust. The issue of algorithmic discrimination was also a major concern. Many participants believed that algorithms trained on incomplete, biased, or non-representative data may overlook certain population groups or treat them unfairly. This perspective has also been supported by international studies. For instance, in a study conducted by Obermeyer in 2019, findings showed that AI algorithms in the United States were less likely to recommend Black patients than White patients for preventive interventions, due to biases embedded in the training data( 38 ). The lack of algorithmic transparency, or explainability, was also identified as a key challenge. Medical ethics and IT experts interviewed emphasized that decisions made by AI must be explainable; otherwise, not only patients but even healthcare professionals may distrust AI-generated recommendations. This issue is particularly critical in the health domain, as errors in decision-making could lead to serious harm or even death. The European Union's 2019 guidelines also stress the importance of the understandability of decisions( 39 ). Another important issue is the lack of clear legal accountability in cases of harm or error resulting from AI-based decision-making. Health law experts pointed out that it is still unclear whether, in the event of an error, responsibility lies with the algorithm’s developer, the decision-making physician, or the service-providing organization. This concern echoes the debates highlighted in the study by Wendehorst (2020) ( 40 ). The implementation of artificial intelligence in the healthcare system, without careful consideration of ethical and social issues, can not only lead to irreparable harm but also significantly increase the likelihood of complete project failure. To overcome these challenges, comprehensive ethical frameworks, clear policies on data confidentiality, standards to reduce algorithmic bias, and mechanisms for determining legal accountability must be established. Only under such conditions can social acceptance and the long-term effectiveness of AI in the healthcare system be expected. Theme 7: Technology Participants emphasized that the absence of legal frameworks, national data standards, and high-level supportive policies are among the major obstacles to operationalizing artificial intelligence in the field of non-communicable disease prevention in Iran. They believed that without clear regulatory systems, the implementation of advanced technologies would not only be ineffective but also high-risk. One of the key issues raised by most interviewees was the lack of transparent legal frameworks for the use of AI in the healthcare system. This concern was especially highlighted by medical ethics experts, high-level health administrators, and some legal specialists. They warned that without well-defined legal and regulatory provisions, there would be serious gaps in accountability, protection of patients' rights, and management of technological risks. This issue is precisely aligned with the WHO (2021) report, which identifies "ethical and legal governance" as one of the six foundational principles for the responsible development of AI in health( 41 ). Additionally, some interviewees pointed to the lack of high-level supportive policies and dedicated budgets for the development of health-related artificial intelligence. They believed that AI is currently not a priority within Iran’s healthcare system, which has led to fragmented projects and the absence of national coordination in this area. This concern is also observed in countries with similar structural contexts. For example, in a study conducted by Bortolini et al. (2024), findings revealed that weak national policies and the lack of a strategic roadmap were among the primary reasons behind the failure of AI projects in healthcare( 42 ). In addition to these issues, some information technology specialists identified the lack of ethical and legal guidelines for the testing, evaluation, and validation of artificial intelligence algorithms as a major challenge. They emphasized that without accepted criteria for testing and approving algorithms, it is impossible to ensure their safe and reliable performance. For the successful implementation of artificial intelligence in the prevention of non-communicable diseases, legal, ethical, data, and policy infrastructures must be systematically and comprehensively designed and executed. The presence of national standards, clear guidelines, accountability mechanisms, and targeted budgeting are essential prerequisites for establishing a secure and effective environment for the use of intelligent technologies in Iran’s healthcare system. Conclusion This qualitative study engaged 34 experts across medicine, health IT, data science, medical ethics, and health system management from Iran's tiered medical universities. Participants viewed AI applications for NCD prevention as both transformative and challenging, with findings organized into 7 main themes and 30 subthemes. While recognizing AI's potential benefits—including early risk detection, system efficiency gains, personalized prevention, and enhanced decision-making—experts identified substantial structural, technical, ethical, and regulatory obstacles. Critical implementation barriers include : Fragmented data infrastructure lacking standardization Inadequate workforce training and readiness Unresolved ethical and legal considerations Absence of comprehensive policy frameworks These findings position Iran's health system in a pre-implementation phase for AI integration, requiring: Strategic long-term planning Sustained investment Multi-stakeholder collaboration Policy Implications and Recommendations Successful AI adoption necessitates: Developing transparent national policies on health AI Creating unified data standards and governance frameworks Establishing ethical guidelines for patient data protection Building workforce capacity across health system levels Implementation pathways should include : Interdisciplinary training programs bridging clinical, public health, and technology domains Targeted pilot projects to evaluate real-world feasibility Strengthened institutional partnerships between healthcare and technology sectors When systematically implemented, AI could become a vital tool for addressing Iran's growing NCD burden while advancing health equity and preventive care effectiveness. Study Limitations : This study has several limitations that should be considered when interpreting the results. First, the data were collected solely based on the perspectives of selected experts within Iran’s health system; therefore, the generalizability of the findings to other settings or countries is limited. Second, despite efforts to ensure diversity in participant selection, certain professional or geographical groups may not have been adequately represented. Third, the qualitative nature of the data and the potential for cognitive or experiential biases in responses may also influence the interpretation of the results. Declarations Ethics approval and consent to participate This study was approved by the Research Ethics Committee of Shiraz University of Medical Sciences under the code IR.SUMS.REC.1404.073. Informed consent was obtained from all participants. All procedures were carried out in accordance with relevant guidelines and regulations. Consent for publication: Not applicable. This study does not contain any individual person’s data in any form (including individual details, images, or videos). Aavailability of data and materials: The data generated and analyzed during this study are not publicly available due to confidentiality agreements with the expert participants. The study was based on interviews with experts, and although interview excerpts are presented in the manuscript, it is not possible to publish the names of the interviewees. Competing interest: The authors declare that there is no scientific, financial, or personal conflict of interest related to the subject of this article. Funding: This research was funded by Deputy of Research and Technology, Shiraz University of Medical Sciences, grant number 140473. Authors' contributions: M.H., M.J.K., and M.M. contributed to the study design and development of the interview guide. A.J. coordinated data collection and supervised the research process. M.H. and M.M. conducted the interviews and performed the initial data analysis. A.J. and M.J.K. reviewed and refined the codes and themes. M.H. and M.J.K. drafted the main manuscript text. A.J. contributed substantially to the interpretation of findings and critical revision of the manuscript. All authors reviewed and approved the final version of the manuscript. Acknowledgments: The authors sincerely appreciate the kind cooperation of all the experts in the fields of medicine, health information technology, data science, medical ethics, and health system management from Iranian medical universities, who generously dedicated their time and shared their valuable experiences during the interviews. The authors also express their gratitude to the Vice-Chancellery for Research and Technology at Shiraz University of Medical Sciences for their scientific and administrative support in conducting this study. References Budreviciute A, Damiati S, Sabir DK, Onder K, Schuller-Goetzburg P, Plakys G, et al. Management and Prevention Strategies for Non-communicable Diseases (NCDs) and Their Risk Factors. Front Public Health. 2020;8:574111. Organization WH. Noncommunicable diseases country profiles 2021. 2021. NADERIMAGHAM SHOHREH AZ, Torabi Parisa, MAHDAVIHEZAVEH ALIREZA, Moradi Mohamad, VALIZADEH BEHZAD, Ezati Elahe, OSTOVAR AFSHIN. Elucidation of Current Status, Implemented Policies and Interventions, Achieved Results, and Future Plans of Iran to Control the Risk Factors of Non-Communicable Diseases: A Review Article. JOURNAL OF DIABETES AND METABOLIC DISORDERS[Internet]. 2023;22(5 ):265-281. Islam SMS, Purnat TD, Phuong NTA, Mwingira U, Schacht K, Fröschl G. Non‐Communicable Diseases (NCDs) in developing countries: a symposium report. Globalization and Health. 2014;10(1):81. WHO GS. Global status report on noncommunicable diseases 2010. 2014. Esteghamati A, Larijani B, Aghajani MH, Ghaemi F, Kermanchi J, Shahrami A, et al. Diabetes in Iran: prospective analysis from first nationwide diabetes report of National Program for Prevention and Control of Diabetes (NPPCD-2016). Scientific reports. 2017;7(1):13461. Torabi Z, Shakibazadeh E, Tajvar M, Rezaei N. Non-communicable diseases challenges and opportunities in Iran: a qualitative study. Scientific Reports. 2025;15(1):8975. Farzadfar F, Yousefi M, Jafari-Khounigh A, Khorrami Z, Haghdoost A, Shadmani FK. Trend and projection of non-communicable diseases risk factors in Iran from 2001 to 2030. Scientific Reports. 2024;14(1):8092. WHO. National status, policies and interventions for the prevention and control of non-communicable diseases in the Islamic Republic of Iran. Global Burden of Desease. 2011. Turon H, Wolfenden L, Finch M, McCrabb S, Naughton S, O’Connor SR, et al. Dissemination of public health research to prevent non-communicable diseases: a scoping review. BMC Public Health. 2023;23(1):757. Jayanna K, Swaroop N, Kar A, Ramanaik S, Pati MK, Pujar A, et al. Designing a comprehensive Non-Communicable Diseases (NCD) programme for hypertension and diabetes at primary health care level: evidence and experience from urban Karnataka, South India. BMC Public Health. 2019;19(1):409. Al Kuwaiti A, Nazer K, Al-Reedy A, Al-Shehri S, Al-Muhanna A, Subbarayalu AV, et al. A review of the role of artificial intelligence in healthcare. Journal of personalized medicine. 2023;13(6):951. Fisher S, Rosella LC. Priorities for successful use of artificial intelligence by public health organizations: a literature review. BMC Public Health. 2022;22(1):2146. Guardian. “The NHS in England is set to commence a pioneering trial of an AI tool capable of predicting the risk of type 2 diabetes up to 13 years before its onset… The trial, starting in 2025 at specific NHS trusts…” Imperial College Healthcare NHS Trust. 2024;14:54-87. Zhu X, Yao Q, Dai W, Ji L, Yao Y, Pang B, et al. Cervical cancer screening aided by artificial intelligence, China. Bulletin of the World Health Organization. 2023;101(6):381. Xie Y, Nguyen QD, Hamzah H, Lim G, Bellemo V, Gunasekeran DV, et al. Artificial intelligence for teleophthalmology-based diabetic retinopathy screening in a national programme: an economic analysis modelling study. The Lancet Digital Health. 2020;2(5):e240-e9. Ramezani M, Takian A, Bakhtiari A, Rabiee HR, Ghazanfari S, Mostafavi H. The application of artificial intelligence in health policy: a scoping review. BMC Health Services Research. 2023;23(1):1416. Graneheim UH, Lundman B. Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness. Nurse education today. 2004;24(2):105-12. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. International journal for quality in health care. 2007;19(6):349-57. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56. Guan Z, Li H, Liu R, Cai C, Liu Y, Li J, et al. Artificial intelligence in diabetes management: Advancements, opportunities, and challenges. Cell Rep Med. 2023;4(10):101213. Organization WH. Assessing national capacity for the prevention and control of noncommunicable diseases: report of the 2021 global survey: World Health Organization; 2023. Beam AL, Kohane IS. Big Data and Machine Learning in Health Care. JAMA. 2018;319(13):1317-8. Patel K, Mistry C, Mehta D, Thakker U, Tanwar S, Gupta R, et al. A survey on artificial intelligence techniques for chronic diseases: open issues and challenges. Artificial Intelligence Review. 2022:1-54. Takahashi H TH, Arai Y, Inoue Y, Kawashima H Applying artificial intelligence to disease staging: Deep learning for improved staging of diabetic retinopathy. PLoS ONE 12(6): e0179790 2017. Areia M, Mori Y, Correale L, Repici A, Bretthauer M, Sharma P, et al. Cost-effectiveness of artificial intelligence for screening colonoscopy: a modelling study. The Lancet Digital Health. 2022;4(6):e436-e44. ORGANIZATION WH. "Going digital for noncommunicable diseases: the case for action". 2022. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nature medicine. 2019;25(1):44-56. Davenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future healthcare journal. 2019;6(2):94-8. Alaran MA, Lawal SK, Jiya MH, Egya SA, Ahmed MM, Abdulsalam A, et al. Challenges and opportunities of artificial intelligence in African health space. DIGITAL HEALTH. 2025;11:20552076241305915. Burki T. European Commission classifies obesity as a chronic disease. The Lancet Diabetes & Endocrinology. 2021;9(7):418. Meskó B, Görög M. A short guide for medical professionals in the era of artificial intelligence. npj Digital Medicine. 2020;3(1):126. Kim KK, Backonja U. Perspectives of community-based organizations on digital health equity interventions: a key informant interview study. Journal of the American Medical Informatics Association. 2024;31(4):929-39. Rajkomar A, Dean J, Kohane I. Machine Learning in Medicine. New England Journal of Medicine. 2019;380(14):1347-58. Wang L, Zhang Y, Wang D, Tong X, Liu T, Zhang S, et al. Artificial intelligence for COVID-19: a systematic review. Frontiers in medicine. 2021;8:704256. Andigema A, Cyrielle NNT, Danaëlle MKL, Ekwelle E. Transforming african healthcare with AI: paving the way for improved health outcomes. Journal of Translational Medicine & Epidemiology. 2024;7(1):1-8. Price WN, Cohen IG. Privacy in the age of medical big data. Nature Medicine. 2019;25(1):37-43. Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. 2019;366(6464):447-53. Hleg A. Ethics guidelines for trustworthy AI. 2019. Wendehorst C. Strict liability for AI and other emerging technologies. Journal of European Tort Law. 2020;11(2):150-80. Guidance W. Ethics and governance of artificial intelligence for health. World Health Organization. 2021. Bortolini VS, Colombo C. Artificial Intelligence in Medicine: the need to see beyond. Brazilian Journal of Law, Technology and Innovation. 2024;2(1):71-89. Tables Tables 1 to 3 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files InterviewChecklist.pdf Tables.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 27 Sep, 2025 Reviewers agreed at journal 16 Sep, 2025 Reviews received at journal 15 Sep, 2025 Reviewers agreed at journal 12 Sep, 2025 Reviewers agreed at journal 10 Sep, 2025 Reviewers invited by journal 03 Sep, 2025 Editor invited by journal 12 Aug, 2025 Editor assigned by journal 04 Aug, 2025 Submission checks completed at journal 01 Aug, 2025 First submitted to journal 01 Aug, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7225629","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":512041497,"identity":"c91673c0-5b73-456f-9f02-854e464256b2","order_by":0,"name":"Marziye Hadian","email":"","orcid":"","institution":"Iran University of Medical Sciences","correspondingAuthor":false,"prefix":"","firstName":"Marziye","middleName":"","lastName":"Hadian","suffix":""},{"id":512041498,"identity":"0f71715e-e7ff-447e-97ed-4fe5b447cb1d","order_by":1,"name":"Mohammadreza jabbari khanbebin","email":"","orcid":"","institution":"Iran University of Medical Sciences","correspondingAuthor":false,"prefix":"","firstName":"Mohammadreza","middleName":"jabbari","lastName":"khanbebin","suffix":""},{"id":512041499,"identity":"e583792e-8da5-4fe0-9090-cd6fa26e1c59","order_by":2,"name":"Mahan Mohammadi","email":"","orcid":"","institution":"Griffith University","correspondingAuthor":false,"prefix":"","firstName":"Mahan","middleName":"","lastName":"Mohammadi","suffix":""},{"id":512041500,"identity":"cbd80d15-bd6e-4212-941a-1030638ba47a","order_by":3,"name":"abdosaleh Jafari","email":"data:image/png;base64,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","orcid":"","institution":"Shiraz University of Medical Sciences","correspondingAuthor":true,"prefix":"","firstName":"abdosaleh","middleName":"","lastName":"Jafari","suffix":""}],"badges":[],"createdAt":"2025-07-27 10:23:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7225629/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7225629/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90980403,"identity":"e27b5dae-15e9-448a-8b4f-0d2686e6391d","added_by":"auto","created_at":"2025-09-10 09:20:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2052831,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7225629/v1/25a6d6c9-3953-4d07-8773-9b634101a9c4.pdf"},{"id":90980101,"identity":"70a33f74-214c-4f4b-b27c-74ade2fe0c9d","added_by":"auto","created_at":"2025-09-10 09:12:33","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":134273,"visible":true,"origin":"","legend":"","description":"","filename":"InterviewChecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7225629/v1/ef29793e39bb59ac3a0a6cfe.pdf"},{"id":90978967,"identity":"e3ee48ea-119e-449b-bb7b-f568408e7999","added_by":"auto","created_at":"2025-09-10 09:04:33","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":25396,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7225629/v1/3b0b302a261b939b0a395d1b.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"AI-Driven Innovations in NCD Management: Challenges and Solutions Based on Expert Perspectives in Iran’s Health System","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn recent decades, non-communicable diseases (NCDs) - including diabetes, cardiovascular diseases, cancers, and chronic respiratory conditions - have emerged as the foremost global public health challenge. This burden is particularly acute in low- and middle-income countries (LMICs), where health systems often struggle to manage the growing epidemic(1) . According to the World Health Organization (WHO), non-communicable diseases (NCDs) accounted for 43 million global deaths in 2021, representing 75% of all mortality excluding pandemic-related deaths. Notably, 18 million of these deaths (42%) occurred prematurely in individuals aged under 70 years, with low- and middle-income countries (LMICs) bearing disproportionate burden—comprising 82% of these premature fatalities.(2). Iran faces a similarly severe NCD burden, with Ministry of Health reports indicating that these diseases account for approximately 90% of annual mortality and over 76% of the country's total disease burden (measured in disability-adjusted life years, DALYs)(3). The rising prevalence of modifiable risk factors - including obesity, physical inactivity, poor nutrition, tobacco use, and alcohol consumption - coupled with high rates of hypertension and diabetes, presents a growing public health crisis for future generations. These conditions not only threaten population health but also create substantial economic burdens through both direct healthcare costs and indirect productivity losses, straining Iran's health system and national economy.(4, 5) Iran currently faces a substantial burden of metabolic disorders, with over 7 million adults diagnosed with diabetes and an additional 13% of the adult population in the prediabetic stage. Compounding this challenge, hypertension affects more than 30% of Iranian adults, with concerning gaps in awareness as a significant proportion remain undiagnosed.(6, 7). Iran is experiencing a concerning upward trajectory in cancer and chronic respiratory disease rates. Current epidemiological projections suggest that, without intervention, NCDs will constitute over 90% of the nation's total disease burden and mortality by 2030 - representing both a public health emergency and a significant threat to sustainable development.\u0026nbsp;(8).These alarming statistics underscore that Iran's NCD epidemic represents a dual crisis: a dire threat to population health and a formidable challenge to health system sustainability and national economic development. Left unaddressed, this growing burden threatens to undermine decades of public health progress and economic growth(9). Given this escalating public health emergency, conventional NCD prevention and control strategies - predominantly reliant on public education campaigns, periodic screening programs, and population-level interventions - have proven insufficient to manage the growing magnitude and complexity of this crisis. The limitations of these traditional approaches highlight the urgent need for innovative solutions capable of addressing the multifaceted nature of NCDs in contemporary healthcare systems.(10) Empirical evidence demonstrates that while comprehensive national NCD control programs have been implemented, their impact remains limited - disease prevalence continues to escalate and the growing burden shows no signs of abatement. The persistent discrepancy between program objectives and actual outcomes reveals fundamental shortcomings in current methodologies, compelling an urgent paradigm shift in our approach to NCD prevention and management(11). In this challenging landscape, artificial intelligence (AI) has emerged as a transformative solution for addressing the NCD crisis. By leveraging its unparalleled capacity for large-scale data analysis, complex pattern recognition, and predictive modeling, AI enables: (1) early detection of high-risk populations, (2) real-time health monitoring, and (3) development of precision public health interventions at scale. These advanced capabilities position AI as a powerful tool for revolutionizing NCD prevention and management strategies(12). AI-based predictive models demonstrate superior performance in NCD screening and risk assessment, achieving up to 30% greater accuracy than conventional approaches while simultaneously reducing healthcare expenditures. These advanced systems enable more efficient resource allocation through earlier and more precise identification of at-risk populations(13).\u0026nbsp;Artificial intelligence has demonstrated significant potential in transforming NCD management globally. A notable example is the UK's AIRE-DM system, which analyzes electrocardiogram (ECG) data to predict type 2 diabetes risk with a 10–13-year lead time. The National Health Service (NHS) has recognized this innovation's value, with pilot implementation scheduled for 2025 as part of its diabetes prevention strategy(14). \u0026nbsp;A prominent example of successful AI implementation comes from China's Hubei Province, where an AI-assisted cervical cancer screening program was deployed across clinical settings from 2017 to 2021. This large-scale application demonstrated AI's potential to enhance cancer detection in resource-constrained environments while maintaining diagnostic accuracy comparable to conventional methods.(15) These implementations demonstrate AI's transformative potential in preventive healthcare. A particularly successful case is Singapore's national diabetic retinopathy screening program, which incorporated deep learning algorithms to achieve a 20% cost reduction—lowering screening expenses from USD 77 to USD 62 per patient annually while maintaining diagnostic accuracy. This achievement highlights how AI can enhance both the efficiency and affordability of large-scale public health initiatives (16).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThese AI capabilities hold particular relevance for Iran's healthcare system, which must address a mounting NCD burden amidst significant resource constraints. However, successful AI integration for NCD management requires more than technological sophistication alone. Implementation effectiveness hinges on critical contextual factors including: Health system infrastructure and workflows, Cultural acceptance of AI-driven healthcare, Supportive policy frameworks and governance structures, and Alignment with existing public health priorities. (17)\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhile Iran has made progress in health system digitalization, critical barriers persist—including fragmented data infrastructures, shortages of data science and medical ethics specialists, and inconsistent policymaking—that collectively constrain AI's healthcare potential. Notably, existing domestic research has disproportionately emphasized technical AI development while neglecting key stakeholder perspectives on implementation challenges from healthcare providers, IT specialists, ethicists, and policymakers. This study aims to address this critical knowledge and policy gap by highlighting both the quantitative burden of NCDs and the technological potential of AI, in order to inform the development of actionable strategies and effective policies toward transforming Iran’s health system ultimately contributing to public health promotion and the sustainability of national resources.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy design:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis qualitative study employed conventional content analysis following Graneheim and Lundman\u0026apos;s methodological framework(18).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis approach systematically examines textual data through four key phases: (1) identification of meaning units, (2) coding, (3) categorization, and (4) latent theme development, enabling deep interpretation of participant experiences. We selected this framework for its:\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003eMethodological flexibility in capturing complex phenomena\u003c/li\u003e\n \u003cli\u003eCapacity to derive contextually grounded concepts\u003c/li\u003e\n \u003cli\u003eAbility to minimize theoretical presuppositions\u003c/li\u003e\n \u003cli\u003eSuitability for exploring multilayered research questions\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThe study adhered to the COREQ (Consolidated Criteria for Reporting Qualitative Research) guidelines to ensure methodological rigor. This 32-item checklist evaluates three critical domains:\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003e\u003cstrong\u003eStudy Design\u003c/strong\u003e: Research team composition, theoretical framework, and participant selection strategy\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eData Collection \u0026amp; Analysis\u003c/strong\u003e: Interview procedures, coding processes, and software utilization\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eFindings\u003c/strong\u003e: Theme development, data interpretation, and reflexivity measures\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003eKey methodological strengths included:\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eSystematic application of content analysis procedures\u003c/li\u003e\n \u003cli\u003eMaintenance of an audit trail for analytical decisions\u003c/li\u003e\n \u003cli\u003eRigorous attention to contextual factors\u003c/li\u003e\n \u003cli\u003eComprehensive reporting following COREQ standards\u003cspan dir=\"RTL\"\u003e(19).\u003c/span\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eTraining of the Research Team:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePrior to the commencement of data collection, all members of the research team, including interviewers and data analysts, received specialized training in the principles of qualitative interviewing, content analysis, and adherence to ethical considerations. Practice and pilot sessions were held to ensure coordinated interview execution and to minimize individual bias. The interview guide was revised and finalized following its pilot implementation on three participants and the collection of their feedback.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParticipants and Inclusion/Exclusion Criteria:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe employed purposive sampling to recruit key informants representing four critical stakeholder groups:\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003e\u003cstrong\u003eClinical Specialists\u003c/strong\u003e:\u003cul\u003e\n \u003cli\u003eEndocrinologists\u003c/li\u003e\n \u003cli\u003eCardiologists\u003c/li\u003e\n \u003cli\u003eSocial medicine specialists\u003c/li\u003e\n \u003cli\u003eEpidemiologists\u003c/li\u003e\n \u003c/ul\u003e\n \u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003ePublic Health Experts\u003c/strong\u003e:\u003cul\u003e\n \u003cli\u003eOccupational health specialists\u003c/li\u003e\n \u003cli\u003ePreventive medicine specialists\u003c/li\u003e\n \u003c/ul\u003e\n \u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eDigital Health Professionals\u003c/strong\u003e:\u003cul\u003e\n \u003cli\u003eHealth information technology specialists\u003c/li\u003e\n \u003cli\u003eAI implementation practitioners\u003c/li\u003e\n \u003c/ul\u003e\n \u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eHealth System Leadership\u003c/strong\u003e:\u003cul\u003e\n \u003cli\u003eMedical ethicists\u003c/li\u003e\n \u003cli\u003eHealth policy makers\u003c/li\u003e\n \u003cli\u003eHospital administrators\u003c/li\u003e\n \u003cli\u003eUniversity health managers\u003c/li\u003e\n \u003c/ul\u003e\n \u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003eInclusion Criteria\u003c/strong\u003e:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eMinimum 3 years\u0026apos; experience in NCD prevention/management\u003c/li\u003e\n \u003cli\u003eDirect involvement in digital health or AI initiatives\u003c/li\u003e\n \u003cli\u003eCurrent affiliation with Iranian medical universities\u003c/li\u003e\n \u003cli\u003eWillingness to provide in-depth perspectives\u003c/li\u003e\n \u003cli\u003eGeographic representation across Iran\u0026apos;s healthcare regions\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eSampling Rationale\u003c/strong\u003e:\u003c/p\u003e\n\u003cp\u003eThis stratified approach ensured:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eMultidisciplinary perspective integration\u003c/li\u003e\n \u003cli\u003eBalanced clinical/technical/policy expertise\u003c/li\u003e\n \u003cli\u003eGrounded understanding of implementation challenges\u003c/li\u003e\n \u003cli\u003eNational-level generalizability of findings\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eExclusion criteria included: withdrawal from cooperation, provision of insufficient or low-quality data, and inability to complete the interview process.\u003cspan dir=\"RTL\"\u003e\u0026nbsp;\u003c/span\u003eTo ensure geographic and structural diversity, samples were selected from Type 1 universities of medical sciences (policy-making and research centers), Type 2 universities (executive centers with moderate research capacity), and Type 3 universities (institutions in underprivileged regions), in order to reflect the diverse realities of Iran\u0026rsquo;s health system.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSampling Method and Sample Size:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study employed a combined sampling strategy, beginning with purposive selection of key informants followed by snowball sampling to identify additional qualified participants. Initial recruitment focused on pre-identified specialists with demonstrated expertise in the research domain. Following each interview, participants were invited to recommend colleagues who possessed relevant experience and could provide valuable insights, thereby expanding the pool of informants through professional networks. The sampling process continued until reaching theoretical saturation, a point at which no new conceptual insights emerged from subsequent interviews. This threshold was initially observed following twenty-eight interviews, with six additional confirmatory interviews conducted to verify data completeness. The final sample comprised thirty-four participants, ensuring robust representation across key perspectives. Detailed demographic characteristics of study participants\u0026mdash;including age distribution, gender representation, professional specialties, years of experience, and institutional affiliations\u0026mdash;are provided in Appendix for comprehensive reference. This documentation enables readers to evaluate the sample composition and its relevance to the research context.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Collection:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData collection occurred between March and July 2025 through semi-structured interviews. The interview guide, developed through a rigorous process involving literature review, expert consultation with three specialists, and pilot testing, covered five key thematic areas:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003ePotential benefits and applications of AI in NCD prevention and management\u003c/li\u003e\n \u003cli\u003eTechnical and infrastructural implementation challenges\u003c/li\u003e\n \u003cli\u003eOrganizational and policy barriers\u003c/li\u003e\n \u003cli\u003eEthical considerations including privacy and public trust\u003c/li\u003e\n \u003cli\u003eHealth system readiness regarding professional, cultural, and educational factors\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eInterviews were scheduled in advance and conducted either in-person or through secure digital platforms (Google Meet or Skype) to accommodate geographical and logistical constraints. Each session lasted 45-60 minutes and was audio-recorded following participant consent. All recordings were subsequently transcribed verbatim for analysis.\u003c/p\u003e\n\u003cp\u003ePrior to each interview, researchers thoroughly explained:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eStudy objectives and procedures\u003c/li\u003e\n \u003cli\u003eRecording methods and data handling\u003c/li\u003e\n \u003cli\u003eConfidentiality protections\u003c/li\u003e\n \u003cli\u003eVoluntary participation terms\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eParticipants provided written or electronic informed consent before proceeding.\u003c/p\u003e\n\u003cp\u003eAll collected data (audio recordings and transcripts) were stored in encrypted formats with restricted access. In accordance with ethical protocols, these materials will be permanently deleted upon study completion.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Analysis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data analysis followed the steps of conventional content analysis as outlined by Graneheim and Lundman, utilizing MAXQDA 2020 software. Initially, the interview transcripts were read multiple times to ensure a thorough understanding of the data. Meaningful units relevant to the research objectives were then identified, and initial coding was conducted. These codes were organized into preliminary categories, which were subsequently merged and refined to derive the main themes and subthemes through conceptual inference.\u003c/p\u003e\n\u003cp\u003eTo ensure reliability, two researchers independently performed the analysis, with a third researcher acting as an arbitrator in cases of disagreement to reach consensus. To enhance the credibility of the findings, summaries of the results and initial codes were shared with participants, and their feedback was incorporated to refine and improve accuracy. Furthermore, MAXQDA software\u0026rsquo;s advanced features were employed for code management, data categorization, concept searching, and the creation of conceptual maps.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Management:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll audio data, transcribed texts, and data codes were stored in a secure and encrypted cloud environment, with access restricted only to the core members of the research team. After completion of the analysis and publication of the results, identifying data will be completely deleted to ensure the confidentiality and security of participants\u0026rsquo; information.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Trustworthiness and Validity:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo ensure data credibility and quality, the study adhered to Lincoln and Guba\u0026rsquo;s four criteria: credibility, dependability, confirmability, and transferability.\u003c/p\u003e\n\u003cp\u003eCredibility was established through dual independent data analysis, participant validation (member checking), and collaborative team reviews of codes and themes. Dependability was achieved by thoroughly documenting all data collection and analysis procedures, including any adjustments made during the research process. Confirmability was ensured by maintaining transparent records of analytical decisions, justifications for code and category selections, and active researcher involvement at every stage of coding and analysis. Finally, transferability was supported by presenting contextualized participant quotations and detailed descriptions of the study\u0026rsquo;s setting and conditions, allowing readers to assess the applicability of the findings to other contexts\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAddressing Researcher Bias:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo reduce bias, all stages of coding and data analysis were carried out independently by two researchers, and any disagreements were resolved through joint discussion and, if necessary, with the opinion of a third person. Also, participants\u0026apos; feedback on preliminary findings was obtained and incorporated into the revisions to enhance the accuracy and validity of interpretations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethodological Limitations:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study has several limitations. First, the findings may not be generalizable to other countries or health systems with different contexts. Second, potential biases could arise from sample selection and participant responses, while data interpretation may be influenced by the researchers\u0026rsquo; expertise. Additionally, access to certain specialists was limited, and some perspectives may not have been fully expressed due to professional or ethical constraints. These factors should be considered when interpreting the results\u003c/p\u003e"},{"header":"Findings","content":"\u003cp\u003e\u003cstrong\u003eDemographic Characteristics of the Participants:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, a total of 34 specialists from various sectors of the health system participated. According to Table 1, the average age of the participants was 45 years (with an age range of 35 to 62 years), and their average length of professional experience was 16 years. The participants included 21 men and 13 women in terms of gender. These individuals were categorized into four main areas of expertise and were selected from Type 1, 2, and 3 medical universities across the country to provide more diverse and comprehensive perspectives from different levels of the health system.\u003c/p\u003e\n\u003cp\u003eAmong all participants, 8 individuals were physicians and specialists in non-communicable diseases (5 men, 3 women), including specialists in social medicine, endocrinology, and cardiology; 4 were health information technology specialists (3 men, 1 woman) who had experience participating in artificial intelligence or digital health projects; 4 were professors or researchers in medical ethics (1 man, 3 women) from reputable academic centers in the country with experience in technology ethics and health policymaking; and 8 were managers, decision-makers, or policymakers in the health system (4 men, 4 women), including university administrators, members of strategic committees, and individuals with managerial experience at the central levels. 4 were experts or researchers in public and occupational health (3 men, 1 woman), and 4 were epidemiology specialists (4 men, 0 women) who participated in the study.\u003c/p\u003e\n\u003cp\u003eIn terms of university distribution, 12 individuals were from Type 1 universities, 11 from Type 2 universities, and 11 from Type 3 universities. This composition was designed in a way to realistically reflect the differences in infrastructure, resources, policies, and real-world experiences at the three academic levels in the country. Participants were also selected from various regions of the country, including the north, south, center, west, and east, to ensure the geographical richness of the data.\u003c/p\u003e\n\u003cp\u003eMost participants held postgraduate degrees and had experience participating in policymaking projects, digital health, or teaching in fields related to health and technology. This composition strengthened the reliability and analytical depth of the extracted data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMain Findings:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAfter content analysis of the data using the Graneheim \u0026amp; Lundman approach and MAXQDA software, a total of seven main themes and 30 subthemes were extracted from the interview transcripts. These themes cover two major areas: the advantages and opportunities of using artificial intelligence in the prevention of NCDs, and the challenges and barriers to implementing this technology in the country's health system. Table 2 presents the benefits of using artificial intelligence in the prevention of non-communicable diseases and Table 3 illustrates the challenges of using artificial intelligence in the prevention of non-communicable diseases.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 1: Enhancing the Effectiveness of Primary Care\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme explores the opportunities and potential capacities of artificial intelligence in strengthening the prevention system for non-communicable diseases (NCDs). Participants from various fields—including medicine, health information technology, and health system management—emphasized that AI, through its ability to analyze vast, complex, and multi-source data, can play a key role in the early identification of at-risk individuals. According to them, the use of advanced machine learning algorithms and data analytics enables the delivery of personalized preventive interventions and significantly improves the effectiveness of screening, monitoring, and management programs for NCDs. These capabilities, particularly within the context of Iran’s health system—which faces challenges such as resource limitations and the growing burden of chronic diseases—can drive a transformative shift in policymaking and the implementation of preventive initiatives.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1-1: Disease Risk Prediction Using Machine Learning Algorithms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants highlighted the potential of AI in more accurately predicting the risk of non-communicable diseases, emphasizing that machine learning algorithms can detect complex and hidden patterns in data that are difficult for humans to identify. According to the interviewees:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“The reality is that using machine learning algorithms, we can find highly complex patterns in data that even doctors might not easily notice. This allows us to identify high-risk patients much earlier and act more quickly.” (Interviewee No. 1)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“One of the advantages of these algorithms is that they can continuously update themselves, and the more data we provide, the better their predictions become. If we can collect more local data, these models can function more accurately and be better adapted to our own conditions. Basically, the more data, the more reliable the output.” (Interviewee No. 7)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1-2: Early Identification of Risk Factors in High-Risk Populations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExperts repeatedly emphasized the importance of AI's ability to detect early signs and risk factors of diseases. They believed that AI can accurately analyze individuals’ health data to identify early indicators and trends that often go unnoticed in conventional diagnostic approaches. This capability enables timely and targeted prevention, allowing healthcare providers to plan appropriate care and interventions before disease progression. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“In my opinion, one of the strengths of AI is that it can identify individuals who are usually overlooked in traditional systems. This is especially critical for underserved areas where access to medical services is limited.” (Interviewee No. 10)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“To explain it simply, when we analyze community health data correctly, we can identify high-risk groups with much greater accuracy, and then design and implement interventions specifically for them. This improves both resource efficiency and overall effectiveness.” (Interviewee No. 23\u003c/em\u003e)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1-3: Increasing Access to Data-Driven Preventive Interventions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSome participants believed that AI has the potential to significantly expand access to preventive services—especially in underserved and remote areas—through digital solutions and intelligent interventions. They emphasized that AI can help bridge existing gaps in healthcare delivery and provide vulnerable populations with access to preventive care.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“AI, through apps and smart systems, can deliver preventive education to people who might never have access to such information in remote areas. It's like having a health educator with them all the time.” (Interviewee No. 12)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“When access to preventive services becomes easier\u0026nbsp;especially for those with fewer resources\u0026nbsp;it can reduce health inequalities. AI can really be a powerful tool in closing those gaps.” (Interviewee No. 26)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1-4: Enhancing Efficiency of Primary Care Through Intelligent Support\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExperts highlighted the role of AI in reducing the workload of primary care staff. Through automation of time-consuming tasks such as symptom monitoring, data entry, and initial screenings, AI allows healthcare workers to focus more on patient care. Additionally, AI’s clinical decision support can improve diagnostic accuracy and guide appropriate interventions. Overall, they believed AI enhances both the quality and effectiveness of primary healthcare.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“One major advantage of AI is that it can handle repetitive tasks like test reminders or analyzing patient data, freeing us to focus on clinical interventions and more specialized care.” (Interviewee No. 30)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“This intelligent support really helps doctors make better decisions and manage patients more accurately and timely. It noticeably improves the quality of care in daily practice.” (Interviewee No. 5)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1-5: Empowering the Public Through Personalized Technologies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants noted that AI can raise public awareness about personal health by providing tailored information and recommendations based on individual data. They emphasized that this approach makes the information more relevant and actionable, boosting people’s motivation and engagement in managing their own health. According to them, AI can shift individuals from passive recipients to active participants in their health management.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“AI-based technologies can deliver health education exactly tailored to each person’s needs. That way, people get the info that’s actually useful for them and can take a more active role in their own care.” (Interviewee No. 28)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Empowerment is only valuable when it respects privacy and autonomy. As long as personal data is protected and used with consent, AI can truly improve public health and help people become more proactive in managing their well-being.” (Interviewee No. 14)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 2: Improving Decision-Making and Personalizing Preventive Interventions\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme specifically focuses on the role of artificial intelligence in improving the quality of clinical decision-making and developing personalized interventions for the prevention of non-communicable diseases. Specialists in the fields of medicine, information technology, and health management presented valuable perspectives on how artificial intelligence can help enhance preventive outcomes. They believe that artificial intelligence, through precise analysis of health data and providing individual insights, can assist physicians in making more accurate decisions and tailoring therapeutic interventions according to each patient's condition. This approach not only increases the effectiveness of prevention, but it can also help reduce the burden of chronic diseases in society and improve the overall quality of healthcare.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-1: Personalizing Preventive Programs Based on Individual and Environmental Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to participants’ opinions, artificial intelligence, by combining and analyzing genetic, behavioral, and environmental data, provides the necessary foundation for formulating precise health programs tailored to individual characteristics. This personalization-based approach can significantly increase the effectiveness of preventive interventions and also enables optimal allocation of health resources. According to the participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“A general prevention program really doesn’t work for everyone. Artificial intelligence has the capability to design a more specific and precise program for each person, based on their medical history, habits, and living environment, which certainly has much greater effectiveness.” (Interviewee No. 20).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“I think right now we have the possibility to combine various data from different sources like apps and even medical records, and by analyzing them, we can offer the best preventive strategy for each person. This makes the recommendations much more precise and tailored to each individual’s condition.” (Interviewee No. 3)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-2: Clinical Decision Support with Intelligent Tools and Analytical Algorithms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that artificial intelligence–based tools can assist physicians in making faster and more accurate decisions, especially when facing complex conditions or a high number of patients. This technology, by analyzing large volumes of data and offering timely suggestions, can significantly improve the quality of clinical decision-making. According to the participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“I think one of the good things about intelligent algorithms is that they can precisely show us which patients need urgent intervention and which ones can be managed with simpler programs. This really makes things easier for us and makes decision-making faster and more effective.” (Interviewee No. 8).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“These tools are not supposed to replace the doctor’s judgment, but rather act as an assistant beside them and help increase the accuracy of decisions. In fact, the goal is that with the help of these technologies, doctors can make better and more confident decisions.” (Interviewee No. 13)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-3: Reducing Human Errors in Risk Assessment and Determining Preventive Strategies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that artificial intelligence, by reducing errors caused by fatigue and high workload, increases the accuracy of clinical evaluations and helps physicians make better decisions in complex situations. According to the participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“When doctors are under high work pressure, the likelihood of mistakes increases. That’s where algorithms can really be helpful and reduce errors so the patient receives the best care.” (Interviewee No. 30).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“In my opinion, one of the main advantages of artificial intelligence is that it can analyze data with much higher accuracy than humans, and that significantly reduces the likelihood of human errors.” (Interviewee No. 9)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-4: Improving Continuous Remote Monitoring of Patients and Predicting Disease Progression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized that artificial intelligence will have the capability to continuously monitor patients’ health status using data obtained from wearable devices and sensors and predict the progression of diseases. This capability enables early diagnosis and the provision of personalized treatments, and ultimately helps improve the quality of medical care.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“I think continuous monitoring really helps us intervene before the disease reaches a severe stage. This is especially important in areas where access to healthcare centers is limited and can save many lives.” (Interviewee No. 11).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“If artificial intelligence were implemented, it would enable us to have up-to-date data. These live and real-time data really allow us to respond faster and better to changes in the patient’s condition. This makes care more accurate and timelier, and ultimately yields a better outcome for the patient.” (Interviewee No. 15\u003c/em\u003e)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-5: Facilitating Communication Among Care Teams Through Shared Data Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that if artificial intelligence is implemented, especially in complex healthcare systems, it can improve coordination and communication among care team members. This technology, by facilitating fast and accurate information exchange, increases the quality of care and reduces potential errors. Furthermore, artificial intelligence helps in managing treatment plans and patient follow-up and enhances team coordination, which is especially important in challenging environments. According to the participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“When all team members have access to accurate and integrated information, their collaboration becomes much better, and ultimately the services provided to patients become higher in quality. This really plays an important role in improving the care process.” (Interviewee No. 27).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“If artificial intelligence is implemented, one of its good features is that it can summarize complex data in a simple and understandable way for all team members. This reduces the time needed for decision-making and allows the team to act more quickly.” (Interviewee No. 22)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2-6: Enhancing the Quality of Health Decision-Making and Policy-Making\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study showed that artificial intelligence, by analyzing large volumes of big data, can assist policymakers in making more accurate and evidence-based decisions. This capability helps improve the quality of decisions and the design of more effective health policies. According to the participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“When the data is up-to-date and the analysis is done intelligently, we can design more precise and effective prevention programs. This way, the success rate of the programs increases, and resources are used more efficiently.” (Interviewee No. 13).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Artificial intelligence is a really good tool for helping us better identify priorities and allocate resources more intelligently. This helps decisions become more accurate and efficient.” (Interviewee No. 6)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 3: Resource Optimization\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme explores the opportunities and benefits of artificial intelligence in improving the prevention and management of non-communicable diseases. Medical professionals, IT experts, and health system administrators emphasized the role of AI in early diagnosis, disease trend prediction, and the delivery of personalized treatments. They also pointed to enhanced healthcare system efficiency, cost reduction, and increased accessibility for patients. By analyzing big data, AI enables the design of more effective preventive programs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003e3-1: Improving Diagnostic Accuracy to Optimize Health Resources\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that AI, by leveraging big data analytics, is capable of providing early diagnoses and more accurate predictions. They also emphasized that this capability can significantly improve the quality of healthcare and help reduce treatment costs. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Based on the workshops we attended, AI has the ability to identify disease patterns at early stages. This is very important because it enables better prevention and thus reduces disease burden. In my opinion, this is one of the greatest advantages of AI in healthcare that we should not overlook.\" (Interviewee No.18).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"With AI, we can predict the course of a disease and intervene on time before it becomes more serious.\" (Interviewee\u0026nbsp;No. 12)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3-2: Increasing Access to Healthcare Services and Reducing Costs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOne of the key advantages of AI is the expansion of preventive services to remote areas with limited access to healthcare. This technology can reduce costs and make services more accessible for underserved populations, thereby promoting health equity. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"AI enables us to deliver preventive services more easily and affordably, especially in underserved areas.\" (Interviewee No.19).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"AI helps patients receive basic care at home or in closer local centers, reducing the need for hospital visits.\" (Interviewee No.26)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3-3: Improving Data Management and System Integration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that AI, by organizing and analyzing large volumes of health data, can enhance coordination between different sectors. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"AI can gather scattered data and provide important, actionable information to support better decision-making. This means we can make decisions with a broader and more precise view.\" (Interviewee No.24).\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"When data is integrated, the system becomes more responsive and the care process runs more smoothly. This greatly contributes to the quality of health services.\" (Interviewee No.18)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 4: Education and Workforce Development\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme emphasizes the importance of training and empowering the workforce to effectively and sustainably utilize smart technologies in the healthcare system. Participants highlighted various educational needs and professional development opportunities, stressing the necessity of enhancing staff knowledge and skills for optimal use of these technologies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4-1: Need for Specialized Training in AI and Emerging Technologies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that specialized training and skills development in AI are essential for physicians, nurses, and health administrators. They emphasized that such education plays a critical role in improving performance and productivity in the health system. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Those working in the health sector need to be familiar with the basics of AI so they can use it properly.\" (Interviewee\u0026nbsp;No. 17)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"If physicians lack sufficient knowledge, they won't be able to trust or correctly apply data-driven decisions. That's why education and familiarity with these technologies are crucial for effective utilization.\" (Interviewee\u0026nbsp;No. 24)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4-2: Importance of Interdisciplinary Education and Team Collaboration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized that joint training across medicine, information technology, and health management is essential to foster effective and coordinated collaboration among multidisciplinary teams. This approach can enhance performance and improve the quality of healthcare services. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"When interdisciplinary training is provided, mutual understanding among team members improves, and so does their coordination. This leads to more effective and impactful collaboration.\" (Interviewee No.22)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Without teamwork and coordination, AI projects usually face serious challenges and are more likely to fail. For success, all members need to work together hand in hand.\" (Interviewee\u0026nbsp;No. 2)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4-3: Continuous Knowledge Updating\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants stressed the importance of continuously updating knowledge and skills in smart technologies. They believed this is essential for maintaining efficiency and productivity in using such technologies. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"If AI is going to be integrated into our work, we must recognize that it's evolving rapidly. Therefore, we must always provide opportunities for continuous, up-to-date training so professionals can keep pace with these changes.\" (Interviewee\u0026nbsp;No. 5)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Participants believed that professional development programs should be ongoing and embedded within the organizational culture to ensure sustainable learning and growth. This approach strengthens collaboration, motivation, and performance improvement within organizations.\" (Interviewee\u0026nbsp;No. 14)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4-4: Creating Motivation and Acceptance of Technology through Education\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that appropriate education can reduce resistance to technology and increase the motivation to adopt it. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"When training is done properly, people are more likely to trust AI and be willing to use it. This trust plays a key role in the successful adoption and implementation of technology.\" (Interviewee\u0026nbsp;No. 23)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"When people become familiar with the benefits and applications of AI, their resistance to it decreases and they are more likely to accept it. This awareness fosters trust and acceptance of AI in workplace and care settings.\" (Interviewee\u0026nbsp;No. 14)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 5: Infrastructural and Technical Challenges\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme highlights the key infrastructural and technical barriers to implementing artificial intelligence (AI) in the country's health system, particularly in the area of non-communicable disease (NCD) prevention. IT specialists, physicians, and health system managers emphasized several major challenges, including the lack of high-quality and reliable data which forms the foundation of any successful AI system, inadequate technical infrastructure for integrated data collection, storage, and processing, and serious concerns around data security and patient privacy. These issues significantly limit the widespread implementation and effective use of AI in healthcare. Addressing them will require careful planning, technological investment, and the training of specialized human resources to fully realize the potential of AI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-1: Lack of High-Quality and Structured Data for Algorithm Training\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that one of the biggest barriers to AI development in healthcare is the shortage of accurate, complete, and structured data needed for training and optimizing algorithms. Without such data, the performance of intelligent systems decreases, limiting their ability to provide precise and trustworthy decisions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“The reality is that our current data is fragmented, incomplete, and mostly stored in unstructured formats. Without sufficient and accurate data, algorithms can't perform well, and their outcomes won’t be reliable.” (Interviewee No.29)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“In our system, data is still mostly kept in traditional ways, making it difficult to extract and use for AI. That’s a major challenge we need to address.” (Interviewee No.4)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-2: Weak IT Infrastructure in Health Centers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExperts pointed out the absence of robust and comprehensive infrastructure—including advanced servers, secure networks, and integrated data management systems. This lack significantly hinders AI implementation and prevents the full utilization of its capabilities. Without such infrastructure, handling large volumes of medical data and ensuring efficient information exchange across sectors becomes difficult, reducing healthcare quality.\u0026nbsp;According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“The fact is, many health centers lack the necessary technological infrastructure for advanced systems. Without servers, secure networks, and integrated platforms, AI efforts are likely to fail.” (Interviewee No.22)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“To use AI safely and effectively, we need to upgrade both our hardware and software. Without these updates, full adoption of the technology isn’t possible, and system security could be compromised.” (Interviewee No.9)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-3: Security and Data Privacy Concerns\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExperts viewed security and privacy as central issues in AI usage. Without strong security systems, data leaks are more likely and patient trust may be lost\u0026nbsp;slowing down AI adoption. Therefore, ensuring data confidentiality is a top priority.\u0026nbsp;According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“It’s absolutely critical that patient data is fully secured and protected from unauthorized access\u0026nbsp;especially when stored digitally and centrally. If this isn't addressed, major problems could arise and public trust would be undermined.” (Interviewee No.4)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Protecting privacy must always come first. If we fail here, people won’t trust AI systems, and they’ll hesitate to share their data. That could derail the entire effort.” (Interviewee No.15)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-4: Shortage of Skilled Professionals and Adequate Training\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized the lack of trained professionals in AI, which slows down implementation.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Most healthcare staff are still unfamiliar with new technologies, and there isn’t enough training. This makes effective use of AI difficult. We need targeted, up-to-date training programs to build staff capabilities.” (Interviewee No.16)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“To succeed, teams must be composed of up-to-date experts who receive continuous training. That way, necessary knowledge and skills are preserved, and projects move forward more effectively.” (Interviewee No.2)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-5: Lack of Clear Legal and Policy Frameworks\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOne of the key challenges is the absence of clear regulations and guidelines regarding AI use and health data governance. This leads to ambiguity around privacy, accountability, and performance standards\u0026nbsp;hindering the technology's growth.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Current laws aren’t adequate for the complexities of AI. We need more updated and comprehensive frameworks to use AI effectively and manage security and privacy risks.” (Interviewee No.10)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“We all know that without solid legal frameworks, it’s hard to use intelligent technologies safely and responsibly. These laws must protect data and privacy—otherwise, there’s a high risk of misuse and loss of public trust.” (Interviewee No.2)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5-6: Organizational Culture and Resistance to Change\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants noted that resistance from healthcare staff and managers to adopt new technologies is a major obstacle. This is often due to inadequate training and fear of disrupting established routines. They stressed that regular training and institutional support can ease this resistance.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Some colleagues are really hesitant to let go of traditional methods and trust smart systems. It’s mostly about habits and the fear that the old ways are safer and change is risky.” (Interviewee No.8)\u003cbr\u003e\u0026nbsp;“To improve technology adoption, organizational culture must shift, and everyone needs to be prepared for change.” (Interviewee No.13)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTheme 6: Ethical and Social Considerations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme focuses on ethical concerns, social implications, and human consequences of using AI in healthcare. Medical ethicists, physicians, and health managers raised issues such as data privacy, potential bias, accountability, and broader social impacts.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6-1: Protection of Patient Privacy and Data Confidentiality\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized the importance of protecting patients' personal information and expressed concerns that sensitive data might be misused or exposed.\u0026nbsp;According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Data confidentiality is a red line for us. If we lose patients’ trust, we can’t use these technologies effectively—everything will fall apart. So, we must ensure their information remains fully protected.” (Interviewee No.11)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“Patients need to understand how their information is used and be assured that it’s secure.” (Interviewee No. 16)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6-2: Concerns About Bias and Inequity in AI Algorithms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSome participants raised concerns about unintended discrimination in AI algorithms that might exclude certain groups from accessing services or accurate diagnoses. They stressed the need for ongoing monitoring and correction to prevent such bias.\u0026nbsp;According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“If the input data is biased, the algorithm output will also be unfair—and may result in discrimination. We must ensure data quality to protect equity.” (Interviewee No.18)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“We have to be very cautious that AI doesn’t widen health disparities. If mismanaged, it could leave already underserved groups even further behind. Ensuring fairness must always be a priority.” (Interviewee No.22)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6-3: Accountability for AI-Based Decisions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized the need for clear responsibility in cases of AI decision errors. They acknowledged that, given the technical and legal complexities, assigning accountability is not easy, but necessary.\u0026nbsp;According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e“When an algorithm makes a mistake, we need to know exactly who’s accountable—is it the doctor making the final call, the software developer, or the entire health system? Without clear answers, we can’t address errors properly or maintain public trust.” (Interviewee No.5)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eTheme 7: Technology\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis theme addresses the role of information technology infrastructure in either enabling or limiting the implementation of artificial intelligence (AI) in the healthcare system. Participants referred to challenges and opportunities related to both hardware and software infrastructure, emphasizing the importance of developing appropriate platforms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e7-1: Lack of adequate and up-to-date technological infrastructure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants believed that the lack of modern equipment and updated technologies hinders the effective implementation of AI and creates limitations in this area. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Our infrastructure isn't sufficient for using AI. We need serious investment and must upgrade our equipment to better utilize this technology.\" (Participant No. 8)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Without strong and reliable infrastructure, you can’t expect AI to perform well or deliver results. Infrastructure is like a foundation—everything is built on top of it.\" (Participant No.\u003c/em\u003e 27)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e7-2: The necessity of data and system standardization\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants emphasized that developing data standards and creating interoperability between various systems plays a crucial role in improving performance and integration of technologies. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Data must be collected and stored in a standardized and unified format so that AI can analyze it properly and deliver more accurate results. This is very important for the success of AI projects.\" (Participant No. 30)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"When no standard exists, data becomes inconsistent, and this leads to problems in analysis and unreliable outcomes.\" (Participant No. 11)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e7-3: Effective use of modern technologies and cloud platforms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSome participants pointed to the advantages of modern and cloud-based technologies, which can reduce hardware limitations and offer better capabilities for AI implementation. According to participants:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\"Cloud platforms allow us to access sufficient processing power without having to purchase expensive hardware, making AI project implementation much easier.\" (Participant No. 21)\u003cbr\u003e\u0026nbsp;\"When we use new technologies, not only is the speed of delivering preventive services increased, but the quality also improves. This enables us to respond more effectively to people’s needs.\" (Participant No. 2)\u003c/em\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis qualitative study explored the challenges and opportunities of implementing artificial intelligence (AI) in non-communicable disease (NCD) prevention from the perspectives of experts across multiple disciplines, including healthcare, information technology, health system management, and medical ethics. While findings reveal widespread professional optimism about AI's potential benefits, they also highlight significant structural, technical, ethical, and cultural obstacles to its effective implementation within Iran's healthcare system.\u003c/p\u003e\u003cp\u003eThis research incorporates perspectives from universities across different tiers (Types 1, 2, and 3) nationwide, providing a multidimensional assessment of intelligent technologies' current status, opportunities, and challenges within the healthcare system. The study's significance stems from Iran's healthcare system being at a nascent stage of AI implementation, where emerging digitalization initiatives coexist with the need for coherent strategies and targeted investments.\u003c/p\u003e\u003cp\u003eUsing conventional content analysis, we systematically examined professional perspectives on AI's potential in healthcare. While many identified challenges align with global discourse, Iran's healthcare system - characterized by its centralized structure, resource constraints, and socio-environmental diversity - presents unique contextual dimensions that demand localized solutions.\u003c/p\u003e\u003cp\u003eThe analysis examines each major theme in depth, with findings contextualized through comparison with both domestic and international research. This approach highlights both universal concerns in AI healthcare implementation and country-specific considerations that require tailored approaches.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 1: Enhancing the Effectiveness of Primary Care\u003c/b\u003e\u003c/p\u003e\u003cp\u003eInterviewees viewed artificial intelligence as a powerful tool for risk prediction, screening, and designing early interventions. They believed that using AI in prevention could lead to resource optimization, improved planning, and better health outcomes.\u003c/p\u003e\u003cp\u003eThis perspective aligns with international studies. For example, in a study conducted by Topol in 2019 in the UK, findings showed that AI, through processing large datasets, was able to predict the risk of cardiovascular and metabolic diseases more accurately than traditional models(\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn a similar study conducted in China by Guan and colleagues in 2023, the findings indicated that artificial intelligence, as an emerging technology in the field of digital health, provides a significant opportunity to improve the efficiency of diabetes care. The use of AI can contribute to more effective diabetes prevention by optimizing disease management, enhancing patient self-care, and reducing treatment costs. Moreover, the development of an AI-based digital care ecosystem that includes both prevention and management of diabetes could be an important step toward reducing the disease burden and improving population health outcomes.(\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e)\u003c/p\u003e\u003cp\u003eIn Iran, due to limited access to high-quality and structured data, the full potential of these capabilities has not yet been realized. However, participants believed that with the development of data infrastructure, artificial intelligence could play a strategic role in reducing the prevalence of non-communicable diseases. Their emphasis on the early identification of risk factors using intelligent tools also aligns with the WHO's 2021 report, which highlights \u0026ldquo;smart and personalized prevention\u0026rdquo; as a key strategy for controlling the epidemic of non-communicable diseases(\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 2: Enhancing Decision-Making and Personalization of Preventive Interventions\u003c/b\u003e\u003c/p\u003e\u003cp\u003eParticipants recognized artificial intelligence's potential to enhance decision-making accuracy for both individual patients and population health management. By analyzing demographic, personal, and biological data, AI could enable more precisely targeted preventive interventions. These findings corroborate research by Beam et al. (2018) in the United States, which demonstrated AI's effectiveness in developing personalized treatment plans and prevention strategies based on individual biological profiles and lifestyle factors.\u003c/p\u003e\u003cp\u003eExperts further highlighted AI's particular value in optimizing healthcare resource allocation, especially for high-risk populations and underserved areas with limited access to services. This capability could help address critical disparities in healthcare delivery while improving overall system efficiency (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn a study conducted by Patel and colleagues in India in 2022, the findings showed that AI algorithms helped the government accurately identify regions at high risk for cardiovascular diseases and direct preventive resources to those areas. This data-driven approach significantly improved the efficiency of public health programs(\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWhile recognizing AI's considerable potential, Iranian experts expressed concerns regarding the scarcity of accurate, locally-relevant data needed for developing targeted interventions. Participants emphasized that creating integrated, comprehensive databases represents a fundamental prerequisite for effectively implementing AI in preventive healthcare. Such infrastructure would enable optimal service allocation tailored to local epidemiological patterns and demographic characteristics. Establishing these robust data systems could facilitate the development of intelligent, personalized solutions for diabetes prevention and management while addressing current implementation barriers.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 3: Resource Optimization\u003c/b\u003e\u003c/p\u003e\u003cp\u003eStudy participants identified several strategic advantages of artificial intelligence for preventing non-communicable diseases. AI applications can:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eExpand access to preventive services\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eEnhance health system efficiency\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eEnable early detection of high-risk individuals\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eReduce healthcare costs\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFacilitate predictive analytics using local data\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eExperts particularly emphasized AI's ability to improve both the accuracy and speed of risk identification. Machine learning algorithms can effectively analyze multiple risk factors simultaneously, including:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003ePersonal health metrics\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLifestyle patterns\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFamily medical history\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eOther relevant biomarkers\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThis analytical capability proves especially valuable for conditions like diabetes, hypertension, and prevalent cancers that typically develop through prolonged preclinical stages.\u003c/p\u003e\u003cp\u003eThese findings align with research by Takahashi et al. (2017), whose Japanese study demonstrated superior sensitivity of AI models in predicting type 2 diabetes risk compared to conventional screening methods (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe interviewees also noted that artificial intelligence can help reduce both the direct and indirect costs of the healthcare system, as early detection and timely interventions can prevent the development of costly complications.\u003c/p\u003e\u003cp\u003eIn a study conducted by Areia in 2022 in the United States, findings showed that the use of smart tools in screening colonoscopy reduced the cost per patient from \u003cspan\u003e$\u003c/span\u003e3,400 to \u003cspan\u003e$\u003c/span\u003e3,343, resulting in approximately \u003cspan\u003e$\u003c/span\u003e290\u0026nbsp;million in annual savings at the population level(\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). This point is also aligned with a report by the World Health Organization (WHO) in 2022, which highlighted the potential of artificial intelligence to reduce the economic burden of non-communicable diseases in low- and middle-income countries(\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDespite the emphasis of this study and other research on the role of artificial intelligence in reducing healthcare system costs, some studies conducted in the United States in 2019 have pointed to increased initial expenses due to the implementation of technology, training of human resources, and the development of infrastructure(\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). However, these initial expenses can be compensated in the long term through disease burden reduction and improved system efficiency.\u003c/p\u003e\u003cp\u003eAnother significant advantage was the enhancement of access to preventive services in underserved areas. Many experts believed that by utilizing smart tools such as health-focused applications, preventive chatbots, and population-tracking algorithms, it is possible to expand the delivery of educational, screening, and follow-up services in remote regions.\u003c/p\u003e\u003cp\u003eIn a study conducted by Alaran in Africa in 2025, findings showed that the use of AI-powered chatbots for hypertension risk assessment successfully led to the identification of hundreds of undiagnosed cases of silent hypertension(\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). Participants also emphasized that one of the key advantages of artificial intelligence is its ability to assist policymakers by providing interpretable and analytical information. According to them, AI dashboards can be used to analyze disease trends and patterns, which in turn enables the design of more effective preventive policies.\u003c/p\u003e\u003cp\u003eThis approach is currently being implemented in several European countries, including the Netherlands and Denmark, and has shown positive results in the control of obesity and type 2 diabetes(\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIf properly implemented, artificial intelligence can play an unparalleled role in enhancing prevention, optimizing resource use, increasing access, reducing costs, and improving decision-making in the health system. To realize these benefits, it is essential to develop localized algorithms, provide specialized training for healthcare professionals, and establish reliable technological infrastructures.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 4: Education and Development\u003c/b\u003e\u003c/p\u003e\u003cp\u003eMost interviewees believed that the successful implementation of artificial intelligence in the prevention of non-communicable diseases requires serious capacity building of human resources at various levels. They pointed out the lack of formal AI training for healthcare professionals, weak interdisciplinary collaboration, and insufficient awareness among managerial and service-level personnel.\u003c/p\u003e\u003cp\u003eThe lack of preparedness among human resources is one of the structural barriers to implementing new technologies. The findings of this study are in line with an international study by Mesko et al. (2020), which showed that in most countries, even physicians have limited understanding of the practical applications of AI and generally do not trust it(\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn Iran, although numerous AI-related training workshops have been held at some universities in recent years, interviewees stated that the lack of coherent, systematic, and formal educational content in the curricula of medical sciences remains one of the main challenges in developing AI applications in the country's health system. Some interviewees pointed to a lack of mutual understanding between health professionals and data engineers. They emphasized that, to develop effective systems, physicians need to understand the basics of data and algorithms, and conversely, technology developers should possess at least a minimal knowledge of the health system and preventive interventions.\u003c/p\u003e\u003cp\u003eOn the other hand, several health system managers stated that even at high-level management positions, there is insufficient understanding of the benefits and risks of AI, which leads to delays in making technology-related decisions. This issue was also examined in a study conducted by Kim et al. in the United States (2024), which found that poor digital literacy among policymakers is one of the critical barriers to implementing digital transformation projects(\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe realization of AI in the prevention of non-communicable diseases requires substantial investment in specialized training and the enhancement of digital literacy among physicians, managers, and health professionals. Designing interdisciplinary courses, establishing new roles within the health system, and integrating technological skills into higher education curricula in health sciences are essential prerequisites for the success of this technological transformation.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 5: Infrastructural and Technical Challenges\u003c/b\u003e\u003c/p\u003e\u003cp\u003eStudy participants consistently identified critical technological barriers hindering AI adoption in Iran's healthcare system:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eInadequate IT infrastructure\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eAbsence of integrated data systems\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003ePoor electronic documentation practices\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLimited availability of structured data\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eInsufficient computational resources\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThe research highlights that structured, integrated, and high-quality data form the essential foundation for effective AI implementation. However, Iran's health information currently exists in fragmented, unstructured formats, with significant portions still paper-based. These data limitations compromise machine learning algorithm performance and analytical accuracy.\u003c/p\u003e\u003cp\u003eThis challenge mirrors global experiences in healthcare AI adoption. Rajkomar et al. (2019) found that even in advanced systems like the U.S. healthcare sector, inconsistent EHR data standardization remains a major obstacle to achieving optimal AI model performance. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlso, in an article published by Wang in China in 2021, the findings showed that although China is a leader in artificial intelligence, the lack of data integration among hospitals remains one of the country\u0026rsquo;s major national challenges (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eMoreover, the lack of cloud infrastructures and the inability of systems to store and process large volumes of data were clearly highlighted by IT experts and health system managers in Iran as serious barriers. This issue is also comparable to a study conducted in low-income African countries by Andigema et al. in 2024, which reported that weak infrastructural capacity has led to the failure of AI-driven projects at the national level(\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn addition, concerns were also raised regarding cybersecurity and the protection of patient information. Participants stated that under current conditions, not only is there a lack of adequate hardware infrastructure, but also the security protocols needed to safeguard sensitive data have not been properly designed. These concerns were similarly highlighted in a study conducted by Price and colleagues in the United States, where the need for clear regulations governing the use of artificial intelligence in healthcare was strongly emphasized(\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). The lack of adequate information technology infrastructure, the absence of comprehensive and structured health databases, limited capacity for big data processing, and security concerns are among the most significant challenges hindering the effective use of artificial intelligence in the prevention of non-communicable diseases in Iran. Addressing these challenges requires national-level action in policymaking, investment in technology, the development of data standards, and the enhancement of information security within the country\u0026rsquo;s health system.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 6: Ethical and Social Considerations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eInterviewees pointed to a range of ethical and social challenges that could seriously hinder the safe adoption and acceptance of AI technologies within the health system. These challenges include concerns about patient privacy, lack of algorithmic transparency (explainability), the risk of algorithmic discrimination, and the absence of clear legal accountability in case of errors.\u003c/p\u003e\u003cp\u003eOne of the main ethical concerns, repeatedly emphasized in most interviews, was the issue of patient privacy and data confidentiality. Given the lack of appropriate data protection structures in Iran, specialists expressed concern that deploying AI without clear regulatory frameworks could lead to the exposure of patients' sensitive information. This concern closely mirrors warnings raised in official documents such as the WHO report (2021)(\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e) And also the study by Price et al. in 2019. And they were also addressed in the study by Price et al. in 2019.(\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e)\u003c/p\u003e\u003cp\u003eIt has been emphasized that implementing AI without adherence to the principles of trust, informed consent, and transparency can lead to public distrust. The issue of algorithmic discrimination was also a major concern. Many participants believed that algorithms trained on incomplete, biased, or non-representative data may overlook certain population groups or treat them unfairly. This perspective has also been supported by international studies. For instance, in a study conducted by Obermeyer in 2019, findings showed that AI algorithms in the United States were less likely to recommend Black patients than White patients for preventive interventions, due to biases embedded in the training data(\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e). The lack of algorithmic transparency, or explainability, was also identified as a key challenge. Medical ethics and IT experts interviewed emphasized that decisions made by AI must be explainable; otherwise, not only patients but even healthcare professionals may distrust AI-generated recommendations. This issue is particularly critical in the health domain, as errors in decision-making could lead to serious harm or even death. The European Union's 2019 guidelines also stress the importance of the understandability of decisions(\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). Another important issue is the lack of clear legal accountability in cases of harm or error resulting from AI-based decision-making. Health law experts pointed out that it is still unclear whether, in the event of an error, responsibility lies with the algorithm\u0026rsquo;s developer, the decision-making physician, or the service-providing organization. This concern echoes the debates highlighted in the study by Wendehorst (2020) (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe implementation of artificial intelligence in the healthcare system, without careful consideration of ethical and social issues, can not only lead to irreparable harm but also significantly increase the likelihood of complete project failure. To overcome these challenges, comprehensive ethical frameworks, clear policies on data confidentiality, standards to reduce algorithmic bias, and mechanisms for determining legal accountability must be established. Only under such conditions can social acceptance and the long-term effectiveness of AI in the healthcare system be expected.\u003c/p\u003e\u003cp\u003e\u003cb\u003eTheme 7: Technology\u003c/b\u003e\u003c/p\u003e\u003cp\u003eParticipants emphasized that the absence of legal frameworks, national data standards, and high-level supportive policies are among the major obstacles to operationalizing artificial intelligence in the field of non-communicable disease prevention in Iran. They believed that without clear regulatory systems, the implementation of advanced technologies would not only be ineffective but also high-risk.\u003c/p\u003e\u003cp\u003eOne of the key issues raised by most interviewees was the lack of transparent legal frameworks for the use of AI in the healthcare system. This concern was especially highlighted by medical ethics experts, high-level health administrators, and some legal specialists. They warned that without well-defined legal and regulatory provisions, there would be serious gaps in accountability, protection of patients' rights, and management of technological risks. This issue is precisely aligned with the WHO (2021) report, which identifies \"ethical and legal governance\" as one of the six foundational principles for the responsible development of AI in health(\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAdditionally, some interviewees pointed to the lack of high-level supportive policies and dedicated budgets for the development of health-related artificial intelligence. They believed that AI is currently not a priority within Iran\u0026rsquo;s healthcare system, which has led to fragmented projects and the absence of national coordination in this area.\u003c/p\u003e\u003cp\u003eThis concern is also observed in countries with similar structural contexts. For example, in a study conducted by Bortolini et al. (2024), findings revealed that weak national policies and the lack of a strategic roadmap were among the primary reasons behind the failure of AI projects in healthcare(\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e In addition to these issues, some information technology specialists identified the lack of ethical and legal guidelines for the testing, evaluation, and validation of artificial intelligence algorithms as a major challenge. They emphasized that without accepted criteria for testing and approving algorithms, it is impossible to ensure their safe and reliable performance.\u003c/p\u003e\u003cp\u003eFor the successful implementation of artificial intelligence in the prevention of non-communicable diseases, legal, ethical, data, and policy infrastructures must be systematically and comprehensively designed and executed. The presence of national standards, clear guidelines, accountability mechanisms, and targeted budgeting are essential prerequisites for establishing a secure and effective environment for the use of intelligent technologies in Iran\u0026rsquo;s healthcare system.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003e This qualitative study engaged 34 experts across medicine, health IT, data science, medical ethics, and health system management from Iran's tiered medical universities. Participants viewed AI applications for NCD prevention as both transformative and challenging, with findings organized into 7 main themes and 30 subthemes. While recognizing AI's potential benefits\u0026mdash;including early risk detection, system efficiency gains, personalized prevention, and enhanced decision-making\u0026mdash;experts identified substantial structural, technical, ethical, and regulatory obstacles.\u003c/p\u003e\u003cp\u003e\u003cb\u003eCritical implementation barriers include\u003c/b\u003e:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eFragmented data infrastructure lacking standardization\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eInadequate workforce training and readiness\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eUnresolved ethical and legal considerations\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eAbsence of comprehensive policy frameworks\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThese findings position Iran's health system in a pre-implementation phase for AI integration, requiring:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eStrategic long-term planning\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSustained investment\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eMulti-stakeholder collaboration\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003ePolicy Implications and Recommendations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eSuccessful AI adoption necessitates:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eDeveloping transparent national policies on health AI\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eCreating unified data standards and governance frameworks\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eEstablishing ethical guidelines for patient data protection\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eBuilding workforce capacity across health system levels\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eImplementation pathways should include\u003c/b\u003e:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eInterdisciplinary training programs bridging clinical, public health, and technology domains\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTargeted pilot projects to evaluate real-world feasibility\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eStrengthened institutional partnerships between healthcare and technology sectors\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eWhen systematically implemented, AI could become a vital tool for addressing Iran's growing NCD burden while advancing health equity and preventive care effectiveness.\u003c/p\u003e\u003cp\u003e\u003cb\u003eStudy Limitations\u003c/b\u003e:\u003c/p\u003e\u003cp\u003eThis study has several limitations that should be considered when interpreting the results. First, the data were collected solely based on the perspectives of selected experts within Iran\u0026rsquo;s health system; therefore, the generalizability of the findings to other settings or countries is limited. Second, despite efforts to ensure diversity in participant selection, certain professional or geographical groups may not have been adequately represented. Third, the qualitative nature of the data and the potential for cognitive or experiential biases in responses may also influence the interpretation of the results.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Research Ethics Committee of Shiraz University of Medical Sciences under the code IR.SUMS.REC.1404.073. Informed consent was obtained from all participants. All procedures were carried out in accordance with relevant guidelines and regulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. This study does not contain any individual person\u0026rsquo;s data in any form (including individual details, images, or videos).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAavailability of data and materials:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data generated and analyzed during this study are not publicly available due to confidentiality agreements with the expert participants. The study was based on interviews with experts, and although interview excerpts are presented in the manuscript, it is not possible to publish the names of the interviewees.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interest:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there is no scientific, financial, or personal conflict of interest related to the subject of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was funded by Deputy of Research and Technology, Shiraz University of Medical Sciences, grant number 140473.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.H., M.J.K., and M.M. contributed to the study design and development of the interview guide. A.J. coordinated data collection and supervised the research process. M.H. and M.M. conducted the interviews and performed the initial data analysis. A.J. and M.J.K. reviewed and refined the codes and themes. M.H. and M.J.K. drafted the main manuscript text. A.J. contributed substantially to the interpretation of findings and critical revision of the manuscript. All authors reviewed and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors sincerely appreciate the kind cooperation of all the experts in the fields of medicine, health information technology, data science, medical ethics, and health system management from Iranian medical universities, who generously dedicated their time and shared their valuable experiences during the interviews. The authors also express their gratitude to the Vice-Chancellery for Research and Technology at Shiraz University of Medical Sciences for their scientific and administrative support in conducting this study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBudreviciute A, Damiati S, Sabir DK, Onder K, Schuller-Goetzburg P, Plakys G, et al. Management and Prevention Strategies for Non-communicable Diseases (NCDs) and Their Risk Factors. Front Public Health. 2020;8:574111.\u003c/li\u003e\n\u003cli\u003eOrganization WH. Noncommunicable diseases country profiles 2021. 2021.\u003c/li\u003e\n\u003cli\u003eNADERIMAGHAM SHOHREH AZ, Torabi Parisa, MAHDAVIHEZAVEH ALIREZA, Moradi Mohamad, VALIZADEH BEHZAD, Ezati Elahe, OSTOVAR AFSHIN. Elucidation of Current Status, Implemented Policies and Interventions, Achieved Results, and Future Plans of Iran to Control the Risk Factors of Non-Communicable Diseases: A Review Article. JOURNAL OF DIABETES AND METABOLIC DISORDERS[Internet]. 2023;22(5 ):265-281.\u003c/li\u003e\n\u003cli\u003eIslam SMS, Purnat TD, Phuong NTA, Mwingira U, Schacht K, Fr\u0026ouml;schl G. Non‐Communicable Diseases (NCDs) in developing countries: a symposium report. Globalization and Health. 2014;10(1):81.\u003c/li\u003e\n\u003cli\u003eWHO GS. Global status report on noncommunicable diseases 2010. 2014.\u003c/li\u003e\n\u003cli\u003eEsteghamati A, Larijani B, Aghajani MH, Ghaemi F, Kermanchi J, Shahrami A, et al. Diabetes in Iran: prospective analysis from first nationwide diabetes report of National Program for Prevention and Control of Diabetes (NPPCD-2016). Scientific reports. 2017;7(1):13461.\u003c/li\u003e\n\u003cli\u003eTorabi Z, Shakibazadeh E, Tajvar M, Rezaei N. Non-communicable diseases challenges and opportunities in Iran: a qualitative study. Scientific Reports. 2025;15(1):8975.\u003c/li\u003e\n\u003cli\u003eFarzadfar F, Yousefi M, Jafari-Khounigh A, Khorrami Z, Haghdoost A, Shadmani FK. Trend and projection of non-communicable diseases risk factors in Iran from 2001 to 2030. Scientific Reports. 2024;14(1):8092.\u003c/li\u003e\n\u003cli\u003eWHO. National status, policies and interventions for the prevention and control of non-communicable diseases in the Islamic Republic of Iran. Global Burden of Desease. 2011.\u003c/li\u003e\n\u003cli\u003eTuron H, Wolfenden L, Finch M, McCrabb S, Naughton S, O\u0026rsquo;Connor SR, et al. Dissemination of public health research to prevent non-communicable diseases: a scoping review. BMC Public Health. 2023;23(1):757.\u003c/li\u003e\n\u003cli\u003eJayanna K, Swaroop N, Kar A, Ramanaik S, Pati MK, Pujar A, et al. Designing a comprehensive Non-Communicable Diseases (NCD) programme for hypertension and diabetes at primary health care level: evidence and experience from urban Karnataka, South India. BMC Public Health. 2019;19(1):409.\u003c/li\u003e\n\u003cli\u003eAl Kuwaiti A, Nazer K, Al-Reedy A, Al-Shehri S, Al-Muhanna A, Subbarayalu AV, et al. A review of the role of artificial intelligence in healthcare. Journal of personalized medicine. 2023;13(6):951.\u003c/li\u003e\n\u003cli\u003eFisher S, Rosella LC. Priorities for successful use of artificial intelligence by public health organizations: a literature review. BMC Public Health. 2022;22(1):2146.\u003c/li\u003e\n\u003cli\u003eGuardian. \u0026ldquo;The NHS in England is set to commence a pioneering trial of an AI tool capable of predicting the risk of type 2 diabetes up to 13 years before its onset\u0026hellip; The trial, starting in 2025 at specific NHS trusts\u0026hellip;\u0026rdquo; Imperial College Healthcare NHS Trust. 2024;14:54-87.\u003c/li\u003e\n\u003cli\u003eZhu X, Yao Q, Dai W, Ji L, Yao Y, Pang B, et al. Cervical cancer screening aided by artificial intelligence, China. Bulletin of the World Health Organization. 2023;101(6):381.\u003c/li\u003e\n\u003cli\u003eXie Y, Nguyen QD, Hamzah H, Lim G, Bellemo V, Gunasekeran DV, et al. Artificial intelligence for teleophthalmology-based diabetic retinopathy screening in a national programme: an economic analysis modelling study. The Lancet Digital Health. 2020;2(5):e240-e9.\u003c/li\u003e\n\u003cli\u003eRamezani M, Takian A, Bakhtiari A, Rabiee HR, Ghazanfari S, Mostafavi H. The application of artificial intelligence in health policy: a scoping review. BMC Health Services Research. 2023;23(1):1416.\u003c/li\u003e\n\u003cli\u003eGraneheim UH, Lundman B. Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness. Nurse education today. 2004;24(2):105-12.\u003c/li\u003e\n\u003cli\u003eTong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. International journal for quality in health care. 2007;19(6):349-57.\u003c/li\u003e\n\u003cli\u003eTopol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56.\u003c/li\u003e\n\u003cli\u003eGuan Z, Li H, Liu R, Cai C, Liu Y, Li J, et al. Artificial intelligence in diabetes management: Advancements, opportunities, and challenges. Cell Rep Med. 2023;4(10):101213.\u003c/li\u003e\n\u003cli\u003eOrganization WH. Assessing national capacity for the prevention and control of noncommunicable diseases: report of the 2021 global survey: World Health Organization; 2023.\u003c/li\u003e\n\u003cli\u003eBeam AL, Kohane IS. Big Data and Machine Learning in Health Care. JAMA. 2018;319(13):1317-8.\u003c/li\u003e\n\u003cli\u003ePatel K, Mistry C, Mehta D, Thakker U, Tanwar S, Gupta R, et al. A survey on artificial intelligence techniques for chronic diseases: open issues and challenges. Artificial Intelligence Review. 2022:1-54.\u003c/li\u003e\n\u003cli\u003eTakahashi H TH, Arai Y, Inoue Y, Kawashima H Applying artificial intelligence to disease staging: Deep learning for improved staging of diabetic retinopathy. PLoS ONE 12(6): e0179790 2017.\u003c/li\u003e\n\u003cli\u003eAreia M, Mori Y, Correale L, Repici A, Bretthauer M, Sharma P, et al. Cost-effectiveness of artificial intelligence for screening colonoscopy: a modelling study. The Lancet Digital Health. 2022;4(6):e436-e44.\u003c/li\u003e\n\u003cli\u003eORGANIZATION WH. \u0026quot;Going digital for noncommunicable diseases: the case for action\u0026quot;. 2022.\u003c/li\u003e\n\u003cli\u003eTopol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nature medicine. 2019;25(1):44-56.\u003c/li\u003e\n\u003cli\u003eDavenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future healthcare journal. 2019;6(2):94-8.\u003c/li\u003e\n\u003cli\u003eAlaran MA, Lawal SK, Jiya MH, Egya SA, Ahmed MM, Abdulsalam A, et al. Challenges and opportunities of artificial intelligence in African health space. DIGITAL HEALTH. 2025;11:20552076241305915.\u003c/li\u003e\n\u003cli\u003eBurki T. European Commission classifies obesity as a chronic disease. The Lancet Diabetes \u0026amp; Endocrinology. 2021;9(7):418.\u003c/li\u003e\n\u003cli\u003eMesk\u0026oacute; B, G\u0026ouml;r\u0026ouml;g M. A short guide for medical professionals in the era of artificial intelligence. npj Digital Medicine. 2020;3(1):126.\u003c/li\u003e\n\u003cli\u003eKim KK, Backonja U. Perspectives of community-based organizations on digital health equity interventions: a key informant interview study. Journal of the American Medical Informatics Association. 2024;31(4):929-39.\u003c/li\u003e\n\u003cli\u003eRajkomar A, Dean J, Kohane I. Machine Learning in Medicine. New England Journal of Medicine. 2019;380(14):1347-58.\u003c/li\u003e\n\u003cli\u003eWang L, Zhang Y, Wang D, Tong X, Liu T, Zhang S, et al. Artificial intelligence for COVID-19: a systematic review. Frontiers in medicine. 2021;8:704256.\u003c/li\u003e\n\u003cli\u003eAndigema A, Cyrielle NNT, Dana\u0026euml;lle MKL, Ekwelle E. Transforming african healthcare with AI: paving the way for improved health outcomes. Journal of Translational Medicine \u0026amp; Epidemiology. 2024;7(1):1-8.\u003c/li\u003e\n\u003cli\u003ePrice WN, Cohen IG. Privacy in the age of medical big data. Nature Medicine. 2019;25(1):37-43.\u003c/li\u003e\n\u003cli\u003eObermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. 2019;366(6464):447-53.\u003c/li\u003e\n\u003cli\u003eHleg A. Ethics guidelines for trustworthy AI. 2019.\u003c/li\u003e\n\u003cli\u003eWendehorst C. Strict liability for AI and other emerging technologies. Journal of European Tort Law. 2020;11(2):150-80.\u003c/li\u003e\n\u003cli\u003eGuidance W. Ethics and governance of artificial intelligence for health. World Health Organization. 2021.\u003c/li\u003e\n\u003cli\u003eBortolini VS, Colombo C. Artificial Intelligence in Medicine: the need to see beyond. Brazilian Journal of Law, Technology and Innovation. 2024;2(1):71-89.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 3 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-health-services-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bhsr","sideBox":"Learn more about [BMC Health Services Research](http://bmchealthservres.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/BHSR/default.aspx","title":"BMC Health Services Research","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Artificial Intelligence, Smart Innovation, Non-Communicable Diseases, Prevention, Health Policy, Digital Health, Iran","lastPublishedDoi":"10.21203/rs.3.rs-7225629/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7225629/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Artificial intelligence (AI) is widely regarded as a transformative technology in healthcare, particularly in the prevention of non-communicable diseases (NCDs). However, in developing countries like Iran, little is known about the readiness of health systems to effectively integrate AI. This study aimed to examine the challenges and benefits of AI implementation in NCD prevention from the perspective of Iranian health experts.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e This qualitative study employed conventional content analysis following the approach proposed by Graneheim and Lundman. Data were collected through semi-structured interviews with 34 experts specializing in medicine, health system management, medical ethics, and information technology. Participants were purposively selected from Type 1, 2, and 3 medical universities across Iran between March 2025 and July 2025. Data analysis was conducted using MAXQDA software (version 20).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFindings:\u003c/strong\u003e The study identified seven main themes and 30 sub-themes through content analysis, categorized into two key domains: (1) the \u003cstrong\u003eadvantages and opportunities\u003c/strong\u003e of using artificial intelligence (AI) in non-communicable disease (NCD) prevention, and (2) the \u003cstrong\u003echallenges and barriers\u003c/strong\u003e to its implementation in Iran’s health system.\u003c/p\u003e\n\u003cp\u003eKey benefits included enhanced \u003cstrong\u003eprimary care effectiveness, personalized interventions, resource optimization\u003c/strong\u003e, and improved \u003cstrong\u003edata-driven decision-making\u003c/strong\u003e. Conversely, major barriers encompassed \u003cstrong\u003einadequate technological infrastructure\u003c/strong\u003e, \u003cstrong\u003epoorly structured data\u003c/strong\u003e, \u003cstrong\u003eethical and legal concerns\u003c/strong\u003e, \u003cstrong\u003ecultural resistance\u003c/strong\u003e, and a \u003cstrong\u003eworkforce unprepared for adopting new technologies\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion: \u003c/strong\u003eThe effective and sustainable integration of artificial intelligence (AI) into Iran’s health system for non-communicable disease (NCD) prevention requires \u003cstrong\u003efour key enablers\u003c/strong\u003e: (1) robust technological infrastructure, (2) enhanced human capital, (3) well-defined legal and ethical frameworks, and (4) an adaptive organizational culture. Only by addressing these prerequisites can AI transition from a theoretical potential to a practical tool for policymaking and preventive interventions.\u003c/p\u003e","manuscriptTitle":"AI-Driven Innovations in NCD Management: Challenges and Solutions Based on Expert Perspectives in Iran’s Health System","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-10 09:04:28","doi":"10.21203/rs.3.rs-7225629/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2025-09-27T18:11:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"92234586211283859905827346584193563992","date":"2025-09-16T19:35:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-16T02:00:48+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"73313827641337020007567978298299863680","date":"2025-09-12T09:14:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"78152446183919502522747131918647669824","date":"2025-09-10T10:22:24+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-03T12:15:53+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-12T10:28:55+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-04T12:03:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-01T15:14:08+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Health Services Research","date":"2025-08-01T15:09:06+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-health-services-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bhsr","sideBox":"Learn more about [BMC Health Services Research](http://bmchealthservres.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/BHSR/default.aspx","title":"BMC Health Services Research","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1dbead8f-d70e-472b-bc25-38661b66ffe8","owner":[],"postedDate":"September 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-09-10T09:04:28+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-10 09:04:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7225629","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7225629","identity":"rs-7225629","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}