An interpretable machine-learning model for predicting the efficacy of nonsteroidal anti-inflammatory drugs for closing hemodynamically signiﬁcant patent ductus arteriosus in preterm infants

Q: What was the goal of this study regarding NSAID treatment in preterm infants?

To develop and validate an interpretable machine-learning model that predicts the efficacy of NSAIDs for closing hemodynamically significant patent ductus arteriosus in preterm infants.

Q: Which model and evaluation approach were used?

A random-forest model was built, evaluated using the AUC of the receiver operating characteristic curve, and interpreted with variable-importance and marginal-effect plots.

Q: What variables were identified as top contributors to the model’s predictions?

The most important features were plasma albumin level and platelet count measured 72 hours before treatment, and 24-hour urine volume measured before treatment.

Aug 6, 2026

Upload by Jiven

Interpretable machine-learning is developed to predict whether nonsteroidal anti-inflammatory drugs (NSAIDs) will successfully close hemodynamically significant patent ductus arteriosus in preterm infants. A cohort of 182 infants (≤30 weeks) treated with NSAIDs forms “success” and “failure” groups using demographic, clinical, laboratory, and echocardiographic variables collected within 72 hours prior to medication. A random-forest model is trained and assessed by AUC, with variable-importance and marginal-effect plots explaining key drivers, then externally validated using an ibuprofen cohort. The model shows AUC 0.792 and identifies plasma albumin level and urine volume as top contributors.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Introduction

Patent ductus arteriosus and hemodynamically significant hsPDA

NSAIDs as first-line therapy and treatment limitations

Methods

Study population and outcome definition

Feature collection and model construction

Performance assessment and interpretability

External validation cohort

Results

Success vs failure group associations

Independent risk factors

Model performance and key contributing features

External cohort prediction

Conclusion

TYPE Original Research PUBLISHED 04 April 2023 DOI 10.3389/fped.2023.1097950 EDITED BY Giovanni Vento, Catholic University of the Sacred Heart, Italy REVIEWED BY Tuuli Metsvaht, University of Tartu, Estonia Konrad Heimann, University Hospital RWTH Aachen, Germany *CORRESPONDENCE Li-Ping Shi [slping2022@163.com](slping2022@163.com) SPECIALTY SECTION This article was submitted to Neonatology, a section of the journal Frontiers in Pediatrics RECEIVED 14 November 2022 ACCEPTED 22 March 2023 PUBLISHED 04 April 2023 CITATION Liu T-X, Zheng J-X, Chen Z, Zhang Z-C, Li D and Shi L-P (2023) An interpretable machinelearning model for predicting the efﬁcacy of nonsteroidal anti-inﬂammatory drugs for closing hemodynamically signiﬁcant patent ductus arteriosus in preterm infants. Front. Pediatr. 11:1097950 . doi: 10.3389/fped.2023.1097950 COPYRIGHT © 2023 Liu, Zheng, Chen, Zhang, Li and Shi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. An interpretable machine-learning model for predicting the efﬁcacy of nonsteroidal anti-inﬂammatory drugs for closing hemodynamically signiﬁcant patent ductus arteriosus in preterm infants Tai-Xiang Liu1, Jin-Xin Zheng2,3, Zheng Chen1,4, Zi-Chen Zhang1, Dan Li1 and Li-Ping Shi1* 1Department of NICU, The Children’s Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Child Health, Hangzhou, China, 2Department of Nephrology, Ruijin Hospital, Institute of Nephrology, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 3School of Global Health, Chinese Center for Tropical Diseases Research, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 4Yiwu Branch, Children’s Hospital Zhejiang University School of Medicine, Yiwu, China Background: Nonsteroidal anti-inﬂammatory drugs (NSAIDs) have been widely used in the closure of ductus arteriosus in premature infants. We aimed to develop and validate an interpretable machine-learning model for predicting the efﬁcacy of NSAIDs for closing hemodynamically signiﬁcant patent ductus arteriosus (hsPDA) in preterm infants. Methods: We assessed 182 preterm infants ≤ 30 weeks of gestational age ﬁrst treated with NSAIDs to close hsPDA. According to the treatment outcome, patients were divided into a “success” group and “failure” group. Variables for analysis were demographic features, clinical features, as well as laboratory and echocardiographic parameters within 72 h before medication use. We developed the machine-learning model using random forests. Model performance was assessed by the area under the receiver operating characteristic curve (AUC) . Variable-importance and marginal-effect plots were constructed to explain the predictive model. The model was validated using an external cohort of two preterm infants who received ibuprofen (p.o.) to treat hsPDA. Results: Eighty-three cases (45 . 6%) were in the success group and 99 (54 .4%) in the failure group. Infants in the success group were associated with maternal chorioamnionitis (p = 0 . 002), multiple births (p = 0 . 007), gestational age at birth (p = 0.020), use of indometacin (p = 0.007), use of inotropic agents (p < 0.001), noninvasive ventilation (p = 0 . 001), plasma albumin level (p < 0 . 001), PDA size (p = 0 . 038) and Vmax (p = 0 . 013) . Multivariable binary logistic regression analysis showed that maternal chorioamnionitis, multiple births, use of indomethacin, use of inotropic agents, plasma albumin level, and PDA size were independent risk factors inﬂuencing the efﬁcacy of NSAIDs (p < 0 . 05)

FAQs

What was the goal of this study regarding NSAID treatment in preterm infants?

To develop and validate an interpretable machine-learning model that predicts the efficacy of NSAIDs for closing hemodynamically significant patent ductus arteriosus in preterm infants.

Which model and evaluation approach were used?

A random-forest model was built, evaluated using the AUC of the receiver operating characteristic curve, and interpreted with variable-importance and marginal-effect plots.

What variables were identified as top contributors to the model’s predictions?

The most important features were plasma albumin level and platelet count measured 72 hours before treatment, and 24-hour urine volume measured before treatment.

Aug 6, 2026

Upload by Jiven