Scholarlink

Background <p>Preterm birth is a leading cause of neonatal mortality worldwide. Socioeconomic status (SES) is associated with preterm birth risk, yet existing studies in low-resource settings have rarely examined whether incorporating SES into clinical models can enhance prediction performance. This study aimed to develop and validate a SES-integrated risk prediction model for preterm birth in Northwest China.</p> Methods <p>A retrospective cohort of 2,781 deliveries from county-level hospitals (January–November 2024) was analyzed. Logistic regression was used to construct a baseline model (clinical variables) and an extended model incorporating SES indicators (education, occupation, living environment). Predictive performance was evaluated using the area under the ROC curve (AUC) and calibration plots.</p> Results <p>Including SES and healthcare accessibility variables markedly improved model discrimination (AUC increased from 0.688 to 0.854) and showed good calibration, indicating strong agreement between predicted and observed outcomes.</p> Conclusion <p>Socioeconomic factors significantly enhance the accuracy and calibration of preterm birth prediction models. Integrating SES indicators into perinatal assessment can improve risk identification and inform equitable maternal health policies.</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

Integrating socioeconomic determinants into preterm birth risk prediction: evidence from Northwest China

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月16日

查看全文
  • Xiaoning Wang,
  • Qin Yuan,
  • Yongmei Yang

摘要

Background

Preterm birth is a leading cause of neonatal mortality worldwide. Socioeconomic status (SES) is associated with preterm birth risk, yet existing studies in low-resource settings have rarely examined whether incorporating SES into clinical models can enhance prediction performance. This study aimed to develop and validate a SES-integrated risk prediction model for preterm birth in Northwest China.

Methods

A retrospective cohort of 2,781 deliveries from county-level hospitals (January–November 2024) was analyzed. Logistic regression was used to construct a baseline model (clinical variables) and an extended model incorporating SES indicators (education, occupation, living environment). Predictive performance was evaluated using the area under the ROC curve (AUC) and calibration plots.

Results

Including SES and healthcare accessibility variables markedly improved model discrimination (AUC increased from 0.688 to 0.854) and showed good calibration, indicating strong agreement between predicted and observed outcomes.

Conclusion

Socioeconomic factors significantly enhance the accuracy and calibration of preterm birth prediction models. Integrating SES indicators into perinatal assessment can improve risk identification and inform equitable maternal health policies.