Scholarlink

Background <p>Current clinical models for pediatric multiple magnet ingestion lack accuracy in asymptomatic patients and in cases with unreliable histories, while radiologic predictors remain unvalidated. This study aimed to develop and validate two nomograms integrating radiologic (± clinical) features for individualized risk stratification.</p> Method <p>This retrospective study analyzed data from 146 pediatric patients treated at Shenzhen Children’s Hospital between 2017 and 2023. Risk factors were identified using the least absolute shrinkage and selection operator (LASSO) and multivariable logistic regression. Model performance was assessed using multiple metrics, with internal validation using 1000 bootstrap resamples.</p> Results <p>Complications occurred in 57 patients (39.0%; 65.1% male; median age, 4.6 years). Nomogram R (area under the receiver operating characteristic curve [AUC]: 0.85, 95% confidence interval [CI] 0.79–0.92; internally validated AUC: 0.86) included magnets size &gt; 25&#xa0;mm, blurred magnet edges, gap between magnets, and lack of progression on serial radiographs. Nomogram C&amp;R (AUC: 0.95, 0.92–0.99; internally validated AUC: 0.96) incorporated multiple ingestions, vomiting, abdominal pain, magnets size &gt; 25&#xa0;mm, and lack of progression on serial radiographs. Both models demonstrated good fit (Hosmer–Lemeshow test), good calibration in calibration plots, and high clinical utility. Nomogram C&amp;R demonstrated superior net reclassification improvement and integrated discrimination improvement. Sensitivity, specificity, and accuracy were 82.5%, 70.8%, and 75.3% for Nomogram R, and 87.7%, 92.1%, and 90.4% for Nomogram C&amp;R. Web-based tools were created for clinical use.</p> Conclusion <p>Nomogram C&amp;R enables accurate complication risk stratification and supports personalised clinical decision-making in pediatric multiple magnet ingestion. Future prospective validation and integration into clinical pathways are warranted to further establish its role in routine practice.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of complications from pediatric multiple magnet ingestion: development and validation of radiologic and clinical-radiologic nomograms in a large retrospective study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月14日

查看全文
  • Xiaohui Lai,
  • Yizhen Luo,
  • Songyu Teng,
  • Yijiang Zhuang,
  • Chenrui Zheng,
  • Jianli Zhou,
  • Xiongjian Cui,
  • Qiru Su,
  • Zhiyong Li,
  • Hongwu Zeng

摘要

Background

Current clinical models for pediatric multiple magnet ingestion lack accuracy in asymptomatic patients and in cases with unreliable histories, while radiologic predictors remain unvalidated. This study aimed to develop and validate two nomograms integrating radiologic (± clinical) features for individualized risk stratification.

Method

This retrospective study analyzed data from 146 pediatric patients treated at Shenzhen Children’s Hospital between 2017 and 2023. Risk factors were identified using the least absolute shrinkage and selection operator (LASSO) and multivariable logistic regression. Model performance was assessed using multiple metrics, with internal validation using 1000 bootstrap resamples.

Results

Complications occurred in 57 patients (39.0%; 65.1% male; median age, 4.6 years). Nomogram R (area under the receiver operating characteristic curve [AUC]: 0.85, 95% confidence interval [CI] 0.79–0.92; internally validated AUC: 0.86) included magnets size > 25 mm, blurred magnet edges, gap between magnets, and lack of progression on serial radiographs. Nomogram C&R (AUC: 0.95, 0.92–0.99; internally validated AUC: 0.96) incorporated multiple ingestions, vomiting, abdominal pain, magnets size > 25 mm, and lack of progression on serial radiographs. Both models demonstrated good fit (Hosmer–Lemeshow test), good calibration in calibration plots, and high clinical utility. Nomogram C&R demonstrated superior net reclassification improvement and integrated discrimination improvement. Sensitivity, specificity, and accuracy were 82.5%, 70.8%, and 75.3% for Nomogram R, and 87.7%, 92.1%, and 90.4% for Nomogram C&R. Web-based tools were created for clinical use.

Conclusion

Nomogram C&R enables accurate complication risk stratification and supports personalised clinical decision-making in pediatric multiple magnet ingestion. Future prospective validation and integration into clinical pathways are warranted to further establish its role in routine practice.