Scholarlink

Objective <p>Intrahepatic cholestasis of pregnancy (ICP) complicated by twin pregnancy significantly increases the risk of preterm birth, and no tailored predictive tools for gestational age (GA) at delivery have been developed for this specific population to date. This study aimed to develop and validate a twin-specific nomogram with dynamic total bile acid (TBA) monitoring, medication history and curative effect for preterm birth prediction in this population.</p> Methods <p>A retrospective cohort of 258 twin pregnancies complicated by ICP was enrolled (November 2024–November 2025). The data included demographic, clinical, biochemical (dynamic TBA parameters, liver enzymes), and therapeutic variables (ursodeoxycholic acid (UDCA) usage, combination regimens, TBA response posttreatment). LASSO regression was used to select predictors, which were incorporated into a logistic regression-based nomogram. The model was validated using discrimination (the area under the receiver operating characteristic curve (AUC)), classification accuracy (sensitivity, specificity, PPV, NPV), calibration (Hosmer–Lemeshow test, calibration curves), and clinical utility (decision curve analysis (DCA)).</p> Results <p>In this cohort, the incidence of preterm birth was 83.3%. The independent predictors of preterm birth included GA at ICP diagnosis, UDCA usage, GA at TBA peak, TBA severity group at peak, predelivery TBA (TBA end), aspartate aminotransferase (AST), and treatment curative effect (all <i>P</i> &lt; 0.05). The discriminatory performance of the nomogram, as measured by the area under the curve (AUC), was 0.812 (95% CI: 0.721–0.903) in the training set and 0.740 (95% CI: 0.590–0.889) in the test set. Calibration curves and Hosmer–Lemeshow tests (training set <i>P</i> = 0.1527; test set <i>P</i> = 0.6991) confirmed good agreement between the predicted and actual outcomes. DCA demonstrated significant net benefits across a clinically relevant risk threshold (0–0.833). The model exhibited high specificity (93.8%) and negative predictive value (85.7%) in the test set.</p> Conclusion <p>To our knowledge, this is among the first nomograms for preterm birth prediction in twin pregnancies with ICP that integrate dynamic TBA monitoring and therapeutic variables. This model is intended primarily as a low-risk exclusion tool to support clinical monitoring strategies, rather than to guide high-risk prediction or delivery decisions. Notably, the model predicts a composite preterm birth outcome modified by both biological risk and clinical intervention rather than purely spontaneous preterm birth, and its low sensitivity further restricts its utility for high-risk preterm birth prediction. Its clinical utility for this purpose requires further rigorous prospective and future external validation studies.</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

A nomogram integrating dynamic total bile acid monitoring, medication history, and curative effect for preterm birth prediction in twin pregnancies with intrahepatic cholestasis of pregnancy

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月9日

查看全文
  • Fei Ding,
  • Jia Li,
  • Minjie Zhang,
  • Yanhua Zhao,
  • Shanshan Liang,
  • Qianwen Zhang

摘要

Objective

Intrahepatic cholestasis of pregnancy (ICP) complicated by twin pregnancy significantly increases the risk of preterm birth, and no tailored predictive tools for gestational age (GA) at delivery have been developed for this specific population to date. This study aimed to develop and validate a twin-specific nomogram with dynamic total bile acid (TBA) monitoring, medication history and curative effect for preterm birth prediction in this population.

Methods

A retrospective cohort of 258 twin pregnancies complicated by ICP was enrolled (November 2024–November 2025). The data included demographic, clinical, biochemical (dynamic TBA parameters, liver enzymes), and therapeutic variables (ursodeoxycholic acid (UDCA) usage, combination regimens, TBA response posttreatment). LASSO regression was used to select predictors, which were incorporated into a logistic regression-based nomogram. The model was validated using discrimination (the area under the receiver operating characteristic curve (AUC)), classification accuracy (sensitivity, specificity, PPV, NPV), calibration (Hosmer–Lemeshow test, calibration curves), and clinical utility (decision curve analysis (DCA)).

Results

In this cohort, the incidence of preterm birth was 83.3%. The independent predictors of preterm birth included GA at ICP diagnosis, UDCA usage, GA at TBA peak, TBA severity group at peak, predelivery TBA (TBA end), aspartate aminotransferase (AST), and treatment curative effect (all P < 0.05). The discriminatory performance of the nomogram, as measured by the area under the curve (AUC), was 0.812 (95% CI: 0.721–0.903) in the training set and 0.740 (95% CI: 0.590–0.889) in the test set. Calibration curves and Hosmer–Lemeshow tests (training set P = 0.1527; test set P = 0.6991) confirmed good agreement between the predicted and actual outcomes. DCA demonstrated significant net benefits across a clinically relevant risk threshold (0–0.833). The model exhibited high specificity (93.8%) and negative predictive value (85.7%) in the test set.

Conclusion

To our knowledge, this is among the first nomograms for preterm birth prediction in twin pregnancies with ICP that integrate dynamic TBA monitoring and therapeutic variables. This model is intended primarily as a low-risk exclusion tool to support clinical monitoring strategies, rather than to guide high-risk prediction or delivery decisions. Notably, the model predicts a composite preterm birth outcome modified by both biological risk and clinical intervention rather than purely spontaneous preterm birth, and its low sensitivity further restricts its utility for high-risk preterm birth prediction. Its clinical utility for this purpose requires further rigorous prospective and future external validation studies.