MIMIC-III与PIC数据库中新生儿胆红素轨迹表型与复合不良结局的潜在类别分析
Neonatal bilirubin trajectory phenotypes and composite adverse outcomes: a two-database latent-class analysis (MIMIC-III and PIC)
一项双数据库回顾性队列研究发现,MIMIC-III和PIC数据库中的高危胆红素轨迹类别与院内死亡、坏死性小肠结肠炎或晚发性脓毒症显著相关。MIMIC-III调整后OR值为5.66和3.16,PIC为2.13。轨迹模型在PIC数据库中未通过外部验证,提示轨迹更多反映基础疾病严重程度而非胆红素特异性效应。
- Article
- Open access
- Published: 25 October 2026
Scientific Reports (2026) Cite this article
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
Abstract
Total serum bilirubin (TSB) is measured routinely in neonates, but whether its trajectory over the first postnatal week adds prognostic value beyond a single measurement is unclear. In this preregistered two-database retrospective cohort study, we fitted latent-class trajectory models separately within each database (primary analysis) and on the pooled population (secondary analysis), including neonates admitted at < 28 days of age with at least three TSB measurements on postnatal days 0–7 (MIMIC-III, n = 2,951; PIC, n = 1,817). Site-specific models identified four classes in MIMIC-III and two in PIC; agreement with pooled classes was low (adjusted Rand index 0.23 and 0.25). High-risk classes were associated with the composite outcome of in-hospital death, necrotizing enterocolitis, or late-onset sepsis in both databases (MIMIC-III: OR 5.66 [95% CI 3.72–8.61] and 3.16 [2.10–4.77]; PIC: OR 2.13 [1.40–3.24]). Outcome drivers differed: MIMIC-III associations reflected necrotizing enterocolitis and late-onset sepsis, whereas the PIC association reflected death. In MIMIC-III, birth-weight adjustment substantially attenuated the associations (Class 4 3.05→1.16; Class 1 5.47→1.85), whereas in PIC, where birth weight was unavailable, admission-weight adjustment strengthened the Class 1 association (OR 2.45→3.81). The trajectory model did not exceed the peak TSB value in MIMIC-III (AUC 0.825 vs. 0.827) and performed poorly in PIC (AUC 0.582; corrected calibration slope 0.719), indicating failed external validation. The hour-specific Bhutani nomogram comparison was applied outside its validated ≥ 35-week population. Bilirubin trajectories largely mark underlying immaturity and illness severity rather than a bilirubin-specific effect; these hypothesis-generating findings require prospective validation, particularly in infants born at ≥ 35 weeks.
Abbreviations
- AUC:
-
Area under the receiver-operating-characteristic curve
- CI:
-
Confidence interval
- DCA:
-
Decision-curve analysis
- EOS:
-
Early-onset sepsis
- FiO2:
-
Fraction of inspired oxygen
- GBTM:
-
Group-based trajectory modeling
- ICD:
-
International Classification of Diseases
- IDI:
-
Integrated discrimination improvement
- IQR:
-
Interquartile range
- LCMM:
-
Latent-class linear mixed model
- LOS:
-
Late-onset sepsis
- MIMIC-III:
-
Medical Information Mart for Intensive Care III
- NEC:
-
Necrotizing enterocolitis
- NRI:
-
Net reclassification improvement
- OR:
-
Odds ratio
- OSF:
-
Open Science Framework
- PIC:
-
Pediatric Intensive Care database
- TSB:
-
Total serum bilirubin
Acknowledgements
The authors thank the MIMIC-III project team and the PIC database team for the publicly available datasets.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Leshan People’s Hospital, Leshan, Sichuan, China
Jingchuan Lu, Meiding Wang, Haiyi Xu & Shuangyan Zhu
Authors
- Jingchuan Lu
- Meiding Wang
- Haiyi Xu
- Shuangyan Zhu
Corresponding author
Correspondence to Jingchuan Lu.
Ethics declarations
Ethics approval and consent to participate
This secondary analysis used only publicly available, de-identified data and required no additional ethics approval. MIMIC-III was used under its credentialed data-use agreement (PhysioNet); PIC was established with IRB approval of the Children’s Hospital, Zhejiang University School of Medicine (approval no. 2019_IRB_052) with waived informed consent. All procedures followed the Declaration of Helsinki.
Consent for publication
Not applicable (no individually identifiable data are presented).
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Lu, J., Wang, M., Xu, H. et al. Neonatal bilirubin trajectory phenotypes and composite adverse outcomes: a two-database latent-class analysis (MIMIC-III and PIC). Sci Rep (2026). https://doi.org/10.1038/s41598-026-76725-0
Received: 01 September 2026
Accepted: 14 October 2026
Published: 25 October 2026
DOI: https://doi.org/10.1038/s41598-026-76725-0
Keywords
来源:Nature 健康护理主题 · nature.com