跳到正文
原文
Nature 健康护理主题·· 9 天前AI 评分35

外部验证与本地校准:急性肾损伤模型在 INSPIRE 和 MIMIC-IV 重症监护队列中的应用

External validation and local recalibration of an acute kidney injury model across INSPIRE and MIMIC-IV critical-care cohorts

AI 导读

研究者在外源性验证集和本地校准模型中评估了急性肾损伤(AKI)模型,发现重新校准后的模型在外部验证集上的AUROC从0.676提高到0.699,校准斜率从0.595调整到0.809,表明模型预测能力有所提升。然而,源校准并不能保证外部校准,临床应用前需评估结果可比性和稳定性。

正文

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

Transport of acute kidney injury (AKI) models between databases depends on patient selection and outcome ascertainment. We developed XGBoost in 8322 INSPIRE patients, calibrated it on an independent validation partition, and evaluated 6,557 MIMIC-IV patients with procedure-date-linked critical-care encounters. Internal test AUROC was 0.814 (0.783, 0.843); parentheses denote 95% patient-bootstrap confidence intervals. Full external-cohort AUROC was 0.669 (0.654, 0.683) and calibration slope 0.541 (0.486, 0.593). On the same fixed external test (1312 patients), a separately refitted, source-recalibrated model using date-aligned source outcomes changed AUROC from 0.676 to 0.699, slope from 0.595 to 0.809, and calibration-in-the-large from 0.148 to 0.836, indicating greater average underprediction. The two outcome frameworks received parallel, post hoc reporting prominence. Across 100 repeated 300-patient samples, logistic updating yielded median slope 1.086 (empirical 95% interval 0.792–1.804) and modest median Brier improvement of 0.0037. Five alternative target partitions gave median slopes 0.90–1.09. Source calibration did not ensure external calibration. Outcome comparability, selection and local updating stability require assessment before clinical use.

Acknowledgements

The author thanks the INSPIRE and MIMIC-IV dataset creators and maintainers.

Author information

Authors and Affiliations

  1. Department of Anesthesiology, The First Affiliated Hospital, Zhejiang University School of Medicine, 1367 Wenyi West Rd, Yuhang District, Hangzhou, 311121, Zhejiang Province, People’s Republic of China

    Ge Gao

Authors

  1. Ge Gao

Corresponding author

Correspondence to Ge Gao.

Ethics declarations

Competing interests

The authors declare no competing interests.

Funding

No funding was received for this study.

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-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gao, G. External validation and local recalibration of an acute kidney injury model across INSPIRE and MIMIC-IV critical-care cohorts. Sci Rep (2026). https://doi.org/10.1038/s41598-026-78904-5

Download citation

  • Received: 03 July 2026

  • Accepted: 28 October 2026

  • Published: 11 November 2026

  • DOI: https://doi.org/10.1038/s41598-026-78904-5

来源:Nature 健康护理主题 · nature.com