Diagnostic agreement between digital systems and clinicians in severe preeclampsia
Indonesia Journal of Biomedical Science · 30.06.2026 · 20 (1) · 135-138
DOI: 10.15562/ijbs.v20i1.686 ↗Kurz gefasst
Fachartikel in der Zeitschrift Indonesia Journal of Biomedical Science, online erschienen am 30.06.2026. Verfasst von Diah Arumsari Sanrisa Putri und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Introduction
Severe preeclampsia remains a frequent and life‑threatening complication in obstetric practice, particularly in low and middle‑income countries. The aim of the study is to assess the diagnostic agreement between clinicians and three digital systems—ChatGPT, DeepSeek, and Gemini—in classifying severe preeclampsia using standardized clinical case data.
Methods
A cross‑sectional analytic study of 133 de‑identified suspected cases of severe preeclampsia from a regional public hospital. Each case was independently classified by clinicians and then by the three systems using a standardized input protocol. Diagnostic agreement was evaluated with Cohen’s Kappa, while McNemar’s test was used to compare paired diagnostic proportions. Sensitivity and specificity were calculated to describe diagnostic performance.
Results
ChatGPT showed moderate agreement with clinicians (Kappa = 0.60) and the highest diagnostic accuracy (sensitivity 82.1%, specificity 77.3%). DeepSeek and Gemini had lower or negative agreement and reduced sensitivity and specificity compared with ChatGPT.
Conclusion
Under the conditions of this study, ChatGPT more closely approximated clinician diagnoses of severe preeclampsia than the other evaluated systems, suggesting a potential complementary role in obstetric decision‑making.
Zitieren
APA
Putri, D. A. S., Firdaus, F. A., & Pramatirta, A. Y. (2026). Diagnostic agreement between digital systems and clinicians in severe preeclampsia. Indonesia Journal of Biomedical Science, 20(1), 135-138. https://doi.org/10.15562/ijbs.v20i1.686
Vancouver
Putri DAS, Firdaus FA, Pramatirta AY. Diagnostic agreement between digital systems and clinicians in severe preeclampsia. Indonesia Journal of Biomedical Science. 2026;20(1):135-138. doi:10.15562/ijbs.v20i1.686
BibTeX
@article{Putri2026,
author = {Putri, Diah Arumsari Sanrisa and Firdaus, Ferry Achmad and Pramatirta, Akhmad Yogi},
title = {Diagnostic agreement between digital systems and clinicians in severe preeclampsia},
journal = {Indonesia Journal of Biomedical Science},
year = {2026},
volume = {20},
number = {1},
pages = {135--138},
doi = {10.15562/ijbs.v20i1.686}
}
EUVIMED