About

Andrea F. Duncan, M.D., M.S., is a Professor of Pediatrics specializing in Neonatology and Newborn Services at the Children's Hospital of Philadelphia. She holds leadership roles including Medical Director of the Neonatal Follow-up Program, Associate Chief of Diversity, Equity, and Inclusion in the Division of Neonatology, and Associate Chair of Diversity, Equity, and Inclusion in the Department of Pediatrics at the Children's Hospital of Philadelphia. Dr. Duncan is involved in faculty search advisory roles and serves on the Committee on Appointments and Promotions for the Department of Pediatrics at the University of Pennsylvania Perelman School of Medicine. Her educational background includes a B.A. from the University of Texas at Austin, an M.D. from the University of Texas Southwestern at Dallas, and an M.S. in Clinical Research from the University of New Mexico. Her research contributions encompass a focus on health equity, anti-racist practices in research, neurodevelopmental outcomes of high-risk infants, and interventions for preterm infants. She has authored numerous publications in these areas, emphasizing her commitment to advancing neonatal care and addressing disparities in health outcomes.

Research topics

• Medicine
• Pediatrics
• Internal medicine
• Psychiatry
• Radiology
• Anesthesia
• Intensive care medicine
• Physical therapy
• Emergency medicine

Selected publications

• Enhancing Neurodevelopmental Outcomes of High-Risk Infants

  JAMA Pediatrics · 2026-03-02

  articleSenior author
  PublisherDOI

• A preregistered, open pipeline for early cerebral palsy risk assessment from infant videos

  GigaScience · 2026-01-01

  articleOpen access

  Cerebral palsy (CP), affecting approximately 1 in 500 children due to abnormal brain development, impacts movement control. Early risk assessment via the general movements assessment (GMA) at 3-4 months is highly predictive for CP but relies on trained clinicians. Machine-learning-based approaches for predicting GMA score from video have shown considerable promise, but typically rely on dataset-specific preprocessing, custom feature sets, and manually designed model pipelines, which make external benchmarking more difficult. This, combined with strict privacy constraints on sharing data, makes it challenging to train and evaluate models across datasets, which is important for assessing clinical utility. There is therefore a need to develop approaches that will work across different datasets to enable multi-site dataset aggregation and model training. To address this gap, we developed an end-to-end pipeline that uses off-the-shelf pose estimation, general-purpose feature extraction, and automated machine learning-none of which are tuned to a specific dataset. We applied this approach to a newly generated large dataset of 1053 infants (with approximately 10-12% positive class for adverse GMA outcome, drawn from a high-risk clinical cohort) within a preregistered study design. Model performance was evaluated on a strict "lock-box" test set, which remained untouched during any phase of model development or preprocessing optimization, and only used for evaluation once the final model and pipeline had been preregistered. The developed model achieved moderate predictive accuracy for clinician-assessed GMA scores (area under the receiver operating characteristic curve, ROC-AUC = 0.77; area under the precision-recall curve, PR-AUC = 0.41). The moderate accuracy is noteworthy given the 10-12% positive class prevalence, and power-law scaling of ROC-AUC as a function of increasing dataset size. By releasing de-identified feature data and open-source code, and simplifying the training pipeline using AutoML, our work establishes essential groundwork for future robust, globally relevant CP screening tools suitable for low-resource settings.

  PublisherDOI

• Unconditional cash transfers to low-income preterm infants and their families: a pilot randomized controlled trial

  Journal of Perinatology · 2025-04-11 · 2 citations

  articleOpen accessSenior author

  OBJECTIVE: Unconditional cash transfers (UCTs)-no strings attached monthly payments-to low-income families may reduce financial stress and improve health outcomes. We sought to determine the feasibility and acceptability of randomizing low-income caregivers of preterm infants to a high- or low-value UCT for 4 months. STUDY DESIGN: Parallel, pilot randomized controlled trial that was preregistered (ClinicalTrials.gov NCT05930327). We enrolled 24 birthing parent-infant dyads. The intervention was a $325 monthly UCT and the active control was a $25 monthly UCT. RESULT: The intervention was feasible and universally acceptable among families in the high-value cash transfer arm. Exploratory outcomes revealed a high degree of financial strain, stress, and depressive symptoms. CONCLUSION: This study provides feasibility, acceptability, and preliminary efficacy data to inform a future, larger trial to examine the impacts of UCTs to low-income birthing parents of preterm infants. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov ID NCT05930327.

  PublisherOA PDFDOI

• Data-Driven Early Prediction of Cerebral Palsy Using AutoML and interpretable kinematic features

  medRxiv · 2025-02-12

  preprintOpen access

  Early identification of cerebral palsy (CP) remains a major challenge due to the reliance on expert assessments that are time-intensive and not scalable. Consequently, a range of studies have aimed at using machine learning to predict CP scores based on motion tracking, e.g. from video data. These studies generally predict clinical scores which are a proxy for CP risk. However, clinicians do not REALLY want to estimate scores, they want to estimate the patients' risk of developing clinical symptoms. Here we present a data-driven machine-learning (ML) pipeline that extracts movement features from infant video based motion tracking and estimates CP risk using AutoML. Using AutoSklearn, our framework minimizes risk of overfitting by abstracting away researcher-driver hyperparameter optimization. Trained on movement data from 3- to 4-month-old infants, our classifier predicts a highly indicative clinical score (General Movements Assessment [GMA]) with an ROC-AUC of 0.78 on a held-out test set, indicating that kinematic movement features capture clinically relevant variability. Without retraining, the same model predicts the risk of cerebral palsy outcomes at later clinical follow-ups with an ROC-AUC of 0.74, demonstrating that early motor representations generalize to long-term neurodevelopmental risk. We employ pre-registered lock-box validation to ensure rigorous performance evaluation. This study highlights the potential of AutoML-powered movement analytics for neurodevelopmental screening, demonstrating that data-driven feature extraction from movement trajectories can provide an interpretable and scalable approach to early risk assessment. By integrating pre-trained vision transformers, AutoML-driven model selection, and rigorous validation protocols, this work advances the use of video-derived movement features for scalable, data-driven clinical assessment, demonstrating how computational methods based on readily available data like infant videos can enhance early risk detection in neurodevelopmental disorders.

  PublisherOA PDFDOI

• Consensus Recommendations for Antiracist Child Health Research: A Modified Delphi Study

  Pediatrics Open Science · 2025-11-26

  articleOpen access

  OBJECTIVE This study aimed to develop consensus on a comprehensive and abridged list of recommendations for conducting antiracist research across all stages of a pediatric research project. Antiracist research aims to ensure that the construct of race is correctly interpreted and that racially and ethnically minoritized communities are fairly engaged throughout a research project. METHODS Using a modified Delphi approach, experts in equitable pediatric research methods completed 3 rounds of surveys and virtual focus groups between April and December 2023. Round 1 asked experts to add to or revise themes and subthemes gathered from a systematic review of published antiracist practices. Round 2 included ranking items’ importance; items voted in the top 50% by 60% of experts were included in an abridged list of essential practices. In Round 3, experts reviewed the final guidance and had the option to rescue items for inclusion in the abridged guidance. RESULTS Fourteen experts with diverse personal and professional backgrounds participated. Experts added new themes and edited existing ones, creating a comprehensive list of 68 recommendations by consensus and identifying 36 to include in the abridged list of essential practices. They also discussed complex topics such as who these recommendations apply to, the nuances of equitable community engagement, and the dangers of health equity tourism. CONCLUSIONS An expert panel achieved consensus on a comprehensive and abridged list of recommendations for how to conduct research in an antiracist, equitable way. These can inform research teams, regulatory agencies, and funders to improve pediatric research.

  PublisherOA PDFDOI

• Hydrocortisone in Preterm Infants and School-Age Functional Outcomes

  JAMA Pediatrics · 2025-12-08

  articleOpen access

  Importance: Bronchopulmonary dysplasia (BPD) is the most common in-hospital morbidity of prematurity, associated with significant long-term medical and neurodevelopmental sequelae and health resource utilization. The Neonatal Research Network (NRN) Hydrocortisone for BPD Trial evaluated the efficacy and safety of hydrocortisone to prevent BPD in high-risk very preterm infants; the impact of hydrocortisone on school-age outcomes in this trial cohort is previously unreported. Objective: To evaluate the impact of neonatal hydrocortisone treatment on early school-age functional motor, cognitive, academic, and pulmonary outcomes among children who participated in the Hydrocortisone for BPD Trial. Design, Setting, and Participants: This prospective long-term cohort study is a follow-up of a randomized clinical trial, the Hydrocortisone for BPD Trial, conducted at 19 centers of the Eunice Kennedy Shriver National Institute of Child Health and Human Development NRN. Participants, enrolled from August 2011 to February 2018, included intubated infants who had been born before 30 weeks' gestational age and had been mechanically ventilated for at least 7 days by postnatal day 14 to 28. They were eligible for a single, in-person, early school-age visit between corrected age 5 years 0 months and 7 years 11 months, conducted from September 2017 to July 2024. Data analysis was performed from July 2024 to September 2025. Intervention: Participants were randomized to a 10-day tapering course of hydrocortisone or placebo beginning at 14 to 28 postnatal days. Main Outcomes and Measures: Early school-age study visits were performed by certified, masked assessors. The primary outcome of functional impairment was defined as any of the following: cognitive delay, motor delay, academic delay, or poor functional exercise capacity. Results: The primary outcome was available for 545 of 674 eligible children (80.9%), including 272 children in the hydrocortisone group (152 [55.9%] female; mean [SD] gestational age, 24.9 [1.5] weeks; mean [SD] age at visit, 5.3 [0.6] years) and 273 in the placebo group (108 [39.6%] female; mean [SD] gestational age, 24.8 [1.5] weeks; mean [SD] age at visit, 5.4 [0.6] years). There was no difference in the rate of functional impairment between the hydrocortisone group (194 of 272 children [71.3%]) and the placebo group (200 of 273 children [73.3%]) (adjusted relative risk, 0.99; 95% CI, 0.89-1.10), nor were there differences in the rates of the individual components. Motor delay was the most common impairment (308 of 510 children [60.4%]), followed by poor functional exercise capacity (175 of 484 children [36.2%]). Conclusions and Relevance: In this study, neonatal hydrocortisone treatment of preterm infants with high risk for BPD did not impact functional impairment or its components; nearly three-quarters of the children demonstrated functional impairment at school age. Trial Registration: ClinicalTrials.gov Identifier: NCT01353313.

  PublisherOA PDFDOI

• Call for Applications for the <i>JAMA Pediatrics</i> Editorial Fellowship

  JAMA Pediatrics · 2025-06-30

  articleSenior author
  PublisherDOI

• Evidence for missed cases of postpartum depression based on paediatric clinical care screenings

  The British Journal of Psychiatry · 2025-06-01 · 2 citations

  articleOpen access
  PublisherOA PDFDOI

• Consensus Recommendations for Anti-Racist Child Health Research: A Modified Delphi Study

  UNC Libraries · 2025-12-11

  articleOpen access

  OBJECTIVES This study aimed to develop consensus on a comprehensive and abridged list of recommendations for conducting anti-racist research across all stages of a pediatric research project. Anti-racist research aims to ensure that the construct of race is correctly interpreted, and that racially and ethnically minoritized communities are fairly engaged throughout a research project. METHODS Using a modified Delphi approach, experts in equitable pediatric research methods completed three rounds of surveys and virtual focus groups between April and December 2023. Round 1 asked experts to add to or revise themes and subthemes gathered from a systematic review of published anti-racist practices. Round 2 included ranking items&rsquo; importance; items voted in the top 50% by 60% of experts were included in an abridged list of essential practices. In Round 3, experts reviewed the final guidance and had the option to rescue items for inclusion in the abridged guidance. RESULTS Fourteen experts with diverse personal and professional backgrounds participated. Experts added new themes and edited existing ones, creating a comprehensive list of 68 recommendations by consensus and identifying 36 to include in the abridged list of essential practices. They also discussed complex topics such as who these recommendations apply to, the nuances of equitable community engagement, and the dangers of health equity tourism. CONCLUSIONS An expert panel achieved consensus on a comprehensive and abridged list of recommendations for how to conduct research in an anti-racist, equitable way. These can inform research teams, regulatory agencies, and funders to improve pediatric research.

  PublisherDOI

• Whole-Body Hypothermia for Neonatal Encephalopathy in Preterm Infants 33 to 35 Weeks’ Gestation

  JAMA Pediatrics · 2025-02-24 · 30 citations

  articleOpen access

  Importance: Hypothermia begun less than 6 hours after birth reduces death or disability in infants with encephalopathy due to hypoxia-ischemia at 36 or more weeks' gestation. Trials of hypothermia for infants younger than 36 weeks' gestation are lacking. Objective: To assess the probability that hypothermia at less than 6 hours after birth decreases death or disability in infants 33 to 35 weeks' gestation with moderate or severe hypoxic-ischemic encephalopathy. Design, Setting, and Participants: This randomized clinical trial was conducted between July 2015 and December 2022 for infants 33 to 35 weeks' gestation with moderate or severe hypoxic-ischemic encephalopathy at less than 6 hours after birth. Bayesian and intention-to-treat analyses were prespecified. The setting included 19 US Neonatal Research Network centers. Data were analyzed from March 2023 to November 2024. Interventions: Infants received unblinded targeted esophageal temperature management. Infants with hypothermia were maintained at 33.5 °C (acceptable 33-34 °C) for 72 hours and then rewarmed. Infants with normothermia were to be maintained at 37 °C (acceptable 36.5-37.3 °C). Main Outcomes and Measures: Composite of death or disability (moderate or severe) at 18 to 22 months' corrected age adjusted for level of encephalopathy and center. Results: A total of 168 infants with hypothermia and normothermia were preterm (mean [SD] age, 34.0 [0.8] weeks' gestation and 34.1 [0.8] weeks' gestation, respectively), while 46 of 88 (52%) and 45 of 80 (56%) were male, respectively. Randomization occurred at mean (SD) 4.5 (1.2) hours and 4.5 (1.3) hours for the groups with hypothermia and normothermia, respectively. The primary outcome occurred in 29 of 83 infants (35%) with hypothermia and 20 of 69 infants (29%) with normothermia (adjusted relative risk [hypothermic/normothermic], 1.11; 95% credibility interval, 0.74-2.00), and death occurred in 18 of 88 infants (20%) with hypothermia and 9 of 78 infants (12%) with normothermia (adjusted relative risk, 1.38; 95% credibility interval, 0.79-2.85). Bayesian analysis with neutral prior indicated 74% probability of increased death or disability and 87% probability of increased death with hypothermia. Conclusions and Relevance: Among infants 33 to 35 weeks' gestation with hypoxic-ischemic encephalopathy, hypothermia at less than 6 hours' age did not reduce death or disability at 18 to 22 months' corrected age. Trial Registration: ClinicalTrials.gov Identifier: NCT01793129.

  PublisherOA PDFDOI

Frequent coauthors

• Betty R. Vohr

  Brown University

  112 shared

• Elisabeth C. McGowan

  111 shared

• Susan R. Hintz

  Stanford University

  101 shared

• Brenda B. Poindexter

  99 shared

• Rosemary D. Higgins

  Eunice Kennedy Shriver National Institute of Child Health and Human Development

  95 shared

• Ira Adams‐Chapman

  Research Network (United States)

  86 shared

• Barbara J. Stoll

  Children's Nutrition Research Center at Baylor College of Medicine

  83 shared

• Abbot R. Laptook

  Women & Infants Hospital of Rhode Island

  82 shared