Predicting major organ complications in primary Sjögren’s disease using a machine learning ensemble strategy: a dual-center retrospective clinical study - Summary - MDSpire

Predicting major organ complications in primary Sjögren’s disease using a machine learning ensemble strategy: a dual-center retrospective clinical study

  • By

  • Wenqi Xia

  • Jiayun Wu

  • Jingyu Zhang

  • Yuening Kang

  • Yuling Chen

  • Ruyi Liao

  • Xiaomin Li

  • Ya Wen

  • Shenghui Wen

  • Fanxuan Meng

  • Huifen Liu

  • Zhiyang He

  • Jieruo Gu

  • Ou Jin

  • Yong Ren

  • Qing Lv

  • May 29, 2026

  • 0 min

Share

Objective:

To predict major organ complications in primary Sjögren’s disease (SjD) using machine learning algorithms and identify important risk factors.

Key Findings:
  • The ensemble model achieved an AUC of 0.725, accuracy of 71%, and a negative predictive value of 79%.
  • Complement C3 and immunoglobulin G (IgG) were identified as the most important predictors of complication risk.
Interpretation:

Remove unsupported conclusions about the model's performance.

Limitations:
  • The study is retrospective and requires further external validation.
  • Model optimization is necessary before clinical implementation.
Conclusion:

Revise to avoid unsupported claims regarding the model's utility.

Original Source(s)

Predicting major organ complications in primary Sjögren’s disease using a machine learning ensemble strategy: a dual-center retrospective clinical study

  • by Wenqi Xia, Jiayun Wu, Jingyu Zhang, Yuening Kang, Yuling Chen, Ruyi Liao, Xiaomin Li, Ya Wen, Shenghui Wen, Fanxuan Meng, Huifen Liu, Zhiyang He, Jieruo Gu, Ou Jin, Yong Ren, Qing Lv

  • May 29, 2026

Related Content

Beyond cartilage degeneration: osteoarthritis as a systems failure of inflammation regulation

Innate immune regulation of adaptive immunity: mechanisms, implications, and bias

Sleep-related problems among patients with rheumatoid arthritis in the World Health Organization Eastern Mediterranean region: a systematic review and meta-analysis