Doctoral Promotion at FPH UI Highlights Innovation in Early Detection of High-Risk Pregnancy Using Machine Learning

On Tuesday, April 22, 2025, the Faculty of Public Health (FPH) at Universitas Indonesia (UI) once again held a Public Doctoral Promotion Session for a student of the Doctoral Program in Public Health. Held at the Doctoral Promotion Room, Building G, FPH UI, the session examined the dissertation of promovendus Eka Santy, titled “Modeling Prediction of Adverse Pregnancy Outcomes Using Machine Learning to Develop a Prototype Electronic Personal Health Record as an Effort to Improve Midwives’ Performance in Early Detection at Primary Health Care Facilities.”

Eka Santy presented the results of her research, which integrates technological and public health approaches to improve maternal and child health services. The open session was chaired by Prof. Dr. Mondastri Korib Sudaryo, M.S., D.Sc., with Prof. Dr. Kemal N. Siregar, M.A., Ph.D. serving as the main supervisor.

In her presentation, Eka explained that Adverse Pregnancy Outcomes (APOs)—such as low birth weight (LBW), preterm birth, and stillbirth—remain leading causes of neonatal mortality. Indonesia continues to record the highest neonatal mortality rate in the ASEAN region and ranks seventh globally.

The aim of Eka’s research was to develop a machine learning-based predictive model for APOs that could be integrated into the development of an electronic Personal Health Record (ePHR), supporting early detection efforts at primary health care facilities. The study employed a sequential explanatory mixed methods design involving four main stages: a literature review, the creation of a machine learning-based APO prediction model, development of the ePHR prototype, and field testing of its acceptance and efficacy.

The results showed that the random forest algorithm was the best-performing model for multiclass classification, achieving an AUC of 98.4%, sensitivity of 95.1%, and F1 score of 94.3%. This model was then implemented into an ePHR prototype named e-bayiKusehAt, featuring 10 key functions to support integrated early APO detection during antenatal care at primary healthcare facilities. The application was designed for use by both midwives and pregnant women, and efficacy testing showed a significant improvement in midwives’ performance, particularly in the area of integrated early detection during prenatal care. “The use of this prototype was also effective in increasing pregnant women’s awareness and readiness regarding potential pregnancy risks that may lead to APOs,” Eka explained in her presentation.

Eka also offered several policy recommendations based on her findings. First, she emphasized the need to integrate APO prediction features into antenatal care services as part of national Standard Operating Procedures (SOP) to reduce rates of LBW, premature births, and neonatal deaths. Second, she highlighted the importance of broadly disseminating the research findings to policymakers and healthcare workers to foster synergy and collaboration in efforts to reduce stunting and neonatal mortality.

Eka further encouraged advancing the ePHR prototype’s Technology Readiness Level (TRL) from level 6–7 to level 8–9, so it can be practically implemented in real healthcare settings. She also recommended conducting further studies in different geographical areas to ensure broader and more accurate generalization of the APO prediction model.

Eka Santy is a seasoned professional with over 26 years of experience in midwifery and maternal and child health. Her educational background includes midwifery and reproductive health. Her commitment is evident in every aspect of her research and innovations, all aimed at enhancing the quality of maternal and child health services and empowering midwives through ongoing education and training.

As part of her commitment to pushing her research findings toward national policy implementation, Eka has disseminated her findings to professional organizations, including the Indonesian Midwives Association (IBI). “Support has also come from the Chair of the IBI Central Board, who has explicitly expressed commitment to supporting the integration of this model and prediction features into existing maternal and child health service applications,” Eka stated.

Successfully defending her dissertation before the examination committee, Eka Santy was declared to have passed with distinction (cum laude), achieving a GPA of 3.91. She became the 8th graduate of the 2025 Doctoral Program in Public Health at FPH UI, the 387th doctoral graduate in the program overall, and the 453rd doctoral graduate of FPH UI. The promotion session was also attended by Prof. Dr. Besral, S.K.M., M.Sc., and Narila Mutia Nasir, S.K.M., M.K.M., Ph.D., as co-supervisors. The examination board included Prof. Dr. Tris Eryando, M.A.; Prof. Dr. Anhari Achadi, S.K.M., Sc.D.; Dr. Artha Prabawa, S.Kom., S.K.M., M.Si.; Dr. Ir. Indrajani, S.Kom., M.M.; and Dr. Ir. Eka Budiarto, S.T., M.Sc.

Eka Santy’s successful completion of her doctoral studies not only marks a personal achievement but also makes a meaningful contribution to the advancement of public health science, especially in enhancing the early detection of Adverse Pregnancy Outcomes. It is hoped that her research will serve as a stepping stone toward better policies for maternal and child health care in Indonesia. (DFD)