<![CDATA[Anemia during pregnancy remains a major public health challenge in Indonesia due to its contribution to maternal morbidity, adverse birth outcomes, and neonatal mortality. This study aimed to develop and validate a multilevel predictive model to identify determinants of anemia among pregnant women using healthcare big data collected from primary healthcare centers in South Sulawesi Province. Secondary data were obtained from the national Maternal and Child Health information system, the e-Cohort application, and electronic medical records from 2020 to 2023. The dataset included 12,847 antenatal visits from 4,312 pregnant women across 87 primary healthcare centers. Data analysis employed multilevel logistic regression and machine learning algorithms, including Random Forest and Gradient Boosting. The findings revealed that non-adherence to iron supplementation, chronic energy deficiency, and high parity were the strongest individual determinants of anemia during pregnancy. In addition, higher quality antenatal care programs at the healthcare facility level demonstrated a protective effect against anemia. Intraclass correlation analysis confirmed significant clustering of anemia prevalence at the facility level. Among all predictive approaches, the Gradient Boosting model achieved the highest predictive performance. These findings highlight the importance of integrating individual and contextual healthcare system factors into predictive modeling to support more targeted and effective anemia prevention strategies for pregnant women in Indonesia.]]>
Copyrights © 2025
