Article Sidebar

PDF
Issue: Vol. 51 No. 5 (2026): JOURNAL OF MILITARY PHARMACO-MEDICINE No.5 - 2026
Section: Articles
DOI: 10.56535/jmpm.v51i5.1967
Date Published: 28/05/2026
Views 176
Downloads 88
How to Cite
Dao, N. H., Le , T. T. N., Do, M. T., Nguyen, H. D., & Doan, M. H. (2026). PRELIMINARY RESULTS OF CHATGPT-BASED LABOR OUTCOME PREDICTION AT MILITARY HOSPITAL 103. Journal of Military Pharmaco-medicine, 51(5), 74-83. https://doi.org/10.56535/jmpm.v51i5.1967

PRELIMINARY RESULTS OF CHATGPT-BASED LABOR OUTCOME PREDICTION AT MILITARY HOSPITAL 103

Nguyen Hung Dao1, , Thi Thanh Nga Le 1, Minh Trung Do1, Hoang Dat Nguyen1, Manh Hung Doan1
1 Vietnam Military Medical University

Main Article Content

Abstract

Objectives: To analyze the clinical and paraclinical characteristics of maternal data provided to ChatGPT and evaluate the preliminary performance of ChatGPT in predicting labor outcomes. Methods: A retrospective, cross-sectional descriptive study was conducted on 500 pregnant women who were monitored and managed during labor at Military Hospital 103, from June to December 2025. Data were collected at the time of initial admission using a standardized form and dataset, then entered into ChatGPT via a unified prompt to predict the mode of delivery. Results: Vaginal delivery rate was 77.6%, while cesarean section accounted for 22.4%, with no statistically significant difference between nulliparous and multiparous women (p > 0.05). ChatGPT achieved an overall prediction accuracy of 90.2%, with a sensitivity of 58.9% and a specificity of 99.2% for predicting cesarean section. Obstetricians demonstrated a higher overall accuracy of 94.4%, with a sensitivity of 75.0% and a specificity of 100%. While obstetricians outperformed ChatGPT in identifying cesarean cases, both approaches showed high precision in predicting vaginal delivery. Conclusion: ChatGPT demonstrated relatively high accuracy and potential as a supportive tool for obstetricians in delivery outcomes prediction.

Keywords

Labor prediction, ChatGPT, Cesarean section

Article Details

References

1. Bộ môn Phụ sản, Trường Đại học Y Hà Nội. Giáo trình Sản khoa. Nhà xuất bản Y học Hà Nội. 2022:172-190.
2. OpenAI. Introducing ChatGPT. 2022.
3. Saida T, Gu W, Hoshiai S, Ishiguro T, Sakai M, Amano T, et al. Artificial intelligence in obstetric and gynecological MR imaging. Magn Reson Med Sci. 2025; 24:354-365.
4. Mai Trọng Dũng và CS. Phân tích các yếu tố liên quan đến chỉ định phẫu thuật lấy thai tại Bệnh viện Phụ sản Trung ương năm 2021. Tạp chí Phụ sản. 2022; 20(3):46-49.
5. Nguyen AD, Duong GTT, Do DT, et al. Primary cesarean section rate among full-term pregnant women with non-previous uterine scar in a hospital of Vietnam. Heliyon. 2022; 8(12):e12222.
6. Phạm Văn Oánh. Nghiên cứu tình hình mổ lấy thai tại Viện Bảo vệ Bà mẹ và Trẻ sơ sinh năm 2000. Luận văn Tốt nghiệp chuyên khoa cấp II. Trường Đại học Y Hà Nội. 2002.
7. Yeniocak AS, Tercan C, Dağdeviren E, Akay E, Aras D, Maş S, et al. ChatGPT for clinical use in labor management: A prospective cohort study. Zeynep Kamil Med J. 2025; 56(3):119-126.
8. AlSaad R, Farrell T, Dela Cruz J, Abd-Alrazaq A, Thomas R, Sheikh J. Artificial intelligence models for predicting the mode of delivery in maternal care. J Gynecol Obstet Hum Reprod. 2025; 54(7):102976. DOI:10.1016/j.jogoh.2025. 102976.
9. Gakuba C, Le Barbey C, Sar A, Bonnet G, Cerasuolo D, Giabicani M, Moyer JD. Evaluation of ChatGPT in predicting 6-month outcomes after traumatic brain injury. Crit Care Med. 2024; 52(6):942-950.