Over the past century, the use of technology has significantly increased across various sectors, including healthcare. Information technology (IT), which facilitates the management, processing, and storage of data in multiple formats, has become integral to daily practices [1]. Artificial intelligence (AI), a subset of IT, involves digital algorithms designed to learn from past experiences and generate outcomes. AI has been recognized as a valuable tool for handling complex tasks [2]. The rapid development of information technology, encompassing advancements in AI, big data processing, and computing, has dramatically altered the structure and efficiency of the healthcare industry. It has also improved the creation and maintenance of medical management information systems [3]. Machine learning, a branch of AI, has been developed to analyze and identify patterns within large datasets, enabling the generation of more accurate responses and decisions [2].

Recent advancements in AI technology are enhancing the potential for improved healthcare services. The application of AI in managing medical data, including electronic medical records, imaging technology, drug design, cancer drug selection, and health management systems, enhances the accuracy and standardization of clinical decision-making. It also facilitates broader data accumulation for knowledge-based medical systems [3]. These developments assist medical practitioners and researchers in selecting optimal treatments and refining treatment plans. However, the implementation of AI technology raises ethical and practical concerns, especially in pediatric care [4]. In pediatrics, treatment is customized for each patient based on age and weight, leading to variations in treatment for patients with the same condition. Relying solely on AI for treatment decisions in pediatric care could introduce bias. This bias might stem from the lack of age-specific considerations in algorithm development or from the inappropriate application of adult data to the pediatric population. Moreover, pediatric patients, due to their diverse developmental stages, may face unique ethical challenges at each stage of their development [5].

To promote student research and aid in the evaluation of study program participants, medical education frequently utilizes AI-enabled information technology. IT and AI are becoming increasingly important in the medical and scientific domains, to the extent that their use can no longer be avoided. Therefore, the purpose of this study was to assess the readiness of pediatric residents and teaching staff to integrate AI into the curriculum.

A cross-sectional study was conducted through an online survey by the Indonesian Collegium of Pediatrics in November 2022. Subjects were recruited from all fifteen nationally accredited pediatric residency programs across Indonesia. The department of child health at each relevant university received an invitation from the Indonesian Collegium of Pediatrics. The first section of the questionnaire provided information about the study and the principal investigator’s contact details. If the subjects agreed to participate, they were directed to the questionnaire section of the study. To respect participant privacy and encourage honest responses, the questionnaire did not collect the names or email addresses of the respondents (Supplement A).

The questionnaire was divided into two sections: demographic information and questions related to IT and AI in child health education. The IT/AI-related questions comprised 16 items designed to evaluate participants’ understanding of IT and AI, the advantages and disadvantages of using these technologies in their educational practices, their views on the potential risks of AI violating medical ethics, and their willingness to incorporate IT and AI technologies into pediatric medical education. Specifically, the advantages and disadvantages of IT and AI were explored through various questions. These included inquiries about whether participants perceived benefits from using IT/AI technology in learning medical science theories and skills, such as easier access to data and procedural tutorials, and improved comprehension of complex theories through enhanced visualization. The questionnaire also aimed to identify any difficulties encountered while using IT technology, including technical errors or challenges in operating the system. Participants could respond to the IT and AI questions with one of five options: strongly disagree, disagree, neutral, agree, or highly agree. The collected data were consolidated into a single digital database and analyzed using SPSS version 26 (IBM, Armonk, NY, USA). The Kolmogorov-Smirnov test was employed to assess the distribution of numerical data in the study. Given that the results indicated an abnormal distribution, the data were presented using the median (interquartile range). Categorical data, in contrast, were presented as number (%). Further analysis involved the use of the Mann-Whitney and Kruskal-Wallis tests to explore any relationships between demographic characteristics and questionnaire results across different groups. The study received ethical approval from the Ethics Committee of the Faculty of Medicine, Universitas Indonesia, under approval number KET-334/UN2.FI/ETIK/PPM.00.02/2023.

A total of 196 teaching staff and 728 pediatric residents’ responses were included in the analysis. The overall median ages were 49.0 years (range, 31.0–80.0 years) for the teaching staff and 31.0 years (range, 23.0–45.0 years) for the pediatric residents. The demographic and clinical characteristics of the participants are summarized in Table 1.

Most of the teaching staff and pediatric residents understood the definitions of IT (91.8%; n = 180 and n = 597, respectively) in relation to questions about their perception of IT. The majority of the teaching staff reported that IT facilitated their teaching of medical theory, including basic biomedical science, anatomy, and physiology (93.9%, n = 184). Additionally, over half of the teaching staff found that IT simplified the teaching of skill sets such as physical examinations and medical procedures (77.6%, n = 152). Consequently, the teaching staff largely disagreed that IT complicated their jobs (65.8%, n = 129). Regarding medical ethics violations, the teaching staff predominantly remained neutral (41.8%, n = 82). The perspectives of pediatric residents were generally supportive of IT and aligned with those of the teaching staff concerning the increased risk of medical ethics violations associated with IT (48.9%, n = 356) (Table 2, Figures 1 and 2).

Teaching staff participants (76.0%, n = 149) and pediatric residents (69.2%, n = 504) both demonstrated an understanding of the definitions of AI. However, only 36.7% of the teaching staff and 44.2% of the pediatric residents had received training on how to use AI for learning and teaching purposes. Both groups—49.6% of pediatric residents (n = 361) and 51.6% of teaching staff (n = 101)—agreed that AI encourages pediatricians to be more confident. Additionally, only seven teaching staff members (3.57%) believed that AI disrupts the learning process, in contrast to 14.7% of pediatric residents (n = 107). The majority of teaching staff (68.9%, n = 135) disagreed with the notion that learning with AI technology diminishes the need for clinical teaching doctors. Furthermore, 54.6% of teaching staff disagreed that AI-powered learning evaluations are preferable to direct human evaluations. Pediatric residents were neutral on whether AI evaluations are superior to direct human evaluations (43.1%, n=314) and on whether learning with AI technology reduces the need for clinical teaching doctors (40.8%, n = 297) (Table 2, Figures 1 and 2).

Both teaching staff and pediatric residents expressed their willingness to incorporate IT and AI into the educational environment. Additionally, to facilitate the development of IT and AI, a significant majority of both pediatric residents and teaching staff were in favor of sharing data between institutions, with 65.4% (n = 449) and 74.5% (n = 140), respectively supporting this initiative. Furthermore, nearly half of the teaching staff (48.4%, n = 91) and a substantial number of pediatric residents (36.97%, n = 254) believed that IT and AI technologies could standardize medical education across all universities (Table 2, Figures 1 and 2).

Gender has emerged as a significant factor influencing responses, with notable differences observed between genders in half of the questions answered by pediatric residents (p < 0.05). Similarly, responses to 11 questions from the teaching staff also showed significant gender differences (p < 0.05). In contrast, the level of residency was significantly correlated only with the willingness to share data across institutions to facilitate the development of IT and AI technologies (p = 0.008). Significant variations were also noted in the responses of the teaching staff, depending on their work experience duration, to specific questions (p = 0.006). Furthermore, the educational level significantly influenced the responses of teaching staff regarding their prior experience in teaching pediatric education using AI technology (p = 0.041).

This study is the first to illustrate the readiness of teaching staff and pediatric residents in Indonesia for AI implementation in medical education. Our findings are more encouraging than those of previous studies, which showed that fewer than 50% of medical students understood the basic definitions of AI [6,7]. This discrepancy may be due to a lack of knowledge, particularly in the early stages of medical education. Most medical students are initially focused on learning about medical health, and their understanding deepens as they progress through their studies [8]. In contrast, the majority of our study’s participants were pediatric residents who had already been introduced to basic AI concepts through their medical school curriculum and clinical experiences. Additionally, they continue to engage daily with technologies such as telemedicine and electronic health records, which have been primarily utilized during the COVID-19 pandemic.

The proportion of women in both the teaching staff and pediatric residents was twice that of men. This disparity may be linked to the higher ratio of female clinicians in pediatrics compared to other disciplines. Additionally, gender significantly influenced the responses to multiple questions in both study groups. This difference may stem from varying levels of enthusiasm and understanding of AI and IT technology between men and women, with men generally showing greater interest and comprehension in technological advancements. Previous studies have demonstrated a higher level of readiness and understanding of AI among male medical students compared to their female counterparts, which supports our findings [9,10]. The median age was 49 years for teaching staff and 32 years for pediatric residents. The demographic data also indicated that the teaching staff possessed a high level of education. The majority of participants held doctoral or subspecialist degrees, and one-third had more than 20 years of professional experience. Consequently, the expertise of the panel was assured, and their knowledgeable responses positively impacted the study outcomes. Furthermore, the high educational level of the pediatric residents expanded the scope of the work and increased the likelihood that the study’s results could be effectively applied to medical education globally, especially within residency programs.

Teaching faculty around the world recognize the need for medical education and curriculum reform to address the challenges presented by the increasing integration of big data and AI in professional practice [11]. AI applications in learning assessment include core surgical skills evaluation, case study grading, attendance tracking, and assignment grading [12]. Future medical professionals must learn to effectively integrate and utilize data from multiple sources [13].

Both teaching personnel and students must possess a basic understanding of AI before it can be effectively integrated into medical education. The majority of participants in our study demonstrated a grasp of basic concepts and exhibited positive attitudes toward IT and AI. Integrating this knowledge into the field of medicine necessitates the incorporation of these concepts early in medical training [14]. Participants need to fully comprehend how AI technology impacts their decision-making processes in medical practice [15]. In our study, both teaching staff and pediatric residents concurred that IT and AI facilitate their learning of theory and skills.

Regarding the question of medical ethics and violence, both groups predominantly chose a neutral stance. This contrasts with a study by Jha et al. [8], where the majority of participants agree that ethical issues associated with AI are likely to emerge. Various patient data raise concerns, particularly regarding privacy and control. In cases where a patient withdraws consent, the European Union mandates the deletion of their personal data. When patient data are scarce, algorithms are trained using fictitious or hypothetical data, increasing the risk of generating harmful and inaccurate treatment recommendations. Consequently, cybersecurity breaches in AI systems could lead to the misclassification of medical data [14]. AI training for medical education should include programs that reduce medical errors, courses on addressing and preventing ethical issues that may arise with AI applications, and mechanisms to prevent data leakage [16]. Additionally, over half of our participants have undergone AI training, which has enhanced their ability to integrate AI throughout their educational journey. Recommended methods for teaching students about AI fundamentals and improving their understanding of AI ethics include hands-on learning opportunities to work directly with AI tools, interactive case-based workshops, and student site visits to explore AI product development [17,18].

An earlier study conducted in Nepal revealed that the majority of participants believed AI might reduce the number of clinical doctors by automating the integration of digital data, radiography, and pathology, which influences future specialty decisions [8]. In contrast to our study, the teaching staff disagreed, while pediatric residents predominantly maintained neutral positions regarding the assertion that AI surpasses direct human evaluation and could eventually diminish the need for clinical teaching physicians. AI technology offers potential benefits such as improving patient access to healthcare, enhancing clinical judgment, reducing medical errors, and increasing clinicians’ access to information [16]. Additionally, integrating AI into the medical curriculum is expected to enhance the learning process and clinical objectivity. However, it will not reduce the necessity for teaching staff, as standardized AI deployment in the curriculum is essential. Standardized AI refers to systems that have undergone rigorous testing and validation to closely replicate the objective evaluations performed by staff on residents. Nevertheless, AI systems are limited by the databases they rely on; therefore, teaching staff remain crucial for assisting residents in managing rare or unusual cases and addressing ethical dilemmas in pediatric patients. Caliskan et al. published a consensus list of skills that medical graduates need to acquire to prepare for the implications of AI [19].

In addition, our survey found that providing basic training before introducing IT and AI technology in medical education enhances proficiency. Students who were already familiar with AI exhibited less apprehension about using technology [20]. Furthermore, when clinicians are equipped to incorporate AI technology into their learning processes, they are expected to function effectively in a healthcare environment that will have experienced significant AI-related transformations [12].

The primary barrier limiting the use of AI in evaluations is the lack of data transformation to digital format, which impacts the ability to meet the data pool requirements necessary for building AI-based systems. This limitation is particularly evident in contemporary educational settings where teaching, assessment, curriculum creation, and evaluation processes predominantly utilize analog technologies [12]. Generally, the development of AI systems requires a substantial amount of data [14]. Accessing previous digital records of the curriculum and converting items from hard copy to soft copy also takes time. Moreover, not all educational institutions can digitally store their data due to financial constraints [21]. To enhance data sources, collaboration among all institutions is essential. Our survey results indicated that teaching staff and pediatric residents from 15 national universities largely support data exchange to facilitate AI development and favor an integrated and standardized curriculum across all universities. Therefore, based on these responses, it is crucial for the Indonesian College of Pediatrics to develop a major database as a fundamental step in developing AI technology, which can then be integrated into the pediatric curriculum.

Currently, the development of medical informatics, including AI, in Indonesia is significantly underdeveloped. Among the 4,523 universities in the country, fewer than one percent offer courses in medical informatics. Additionally, when these courses are available, they are often merely a minor component of other non-health-related majors, such as information systems or computer science. Moreover, research publications that focus on medical or health informatics in Indonesia are exceedingly scarce, especially when compared to other major fields such as medicine. Thus, there is a pressing need for the development of a medical informatics curriculum that can be integrated into major courses. Without a workforce highly trained in medical informatics, integrating AI into health education will pose a substantial challenge [22].

This study represents the first attempt to evaluate the perspectives of pediatric residents and teaching staff in Indonesia regarding their readiness for the integration of AI into the educational process. However, the focus on only these groups may not accurately represent the broader population involved in medical education. Additionally, the lack of follow-up responses to the survey questions may have limited our understanding and hindered the identification of key factors influencing their readiness. Therefore, further prospective research is necessary to identify and validate additional factors that could affect both groups perceptions of their readiness to incorporate AI in both classroom and clinical settings.

According to our study, both pediatric residents and teaching staff held positive views regarding the incorporation of AI into medical education. Nevertheless, significant barriers remain in the implementation of AI technology within pediatric education. Consequently, the creation of a dependable database, coupled with the integration of medical informatics courses into the curriculum, are essential steps to promote the advancement of AI within Indonesia’s health system.