EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Predicting Prenatal Depression and Assessing Model Bias Using Machine Learning Models

Yongchao Huang, Suzanne Alvernaz, Sage J Kim, Pauline Maki, Yang Dai, Beatriz Peñalver Bernabé

Biological Psychiatry Global Open Science · 2024-11 · 4 (6) · 100376

DOI: 10.1016/j.bpsgos.2024.100376 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biological Psychiatry Global Open Science, 2024-11. Verfasst von Yongchao Huang und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Engineering, Colleges of Engineering and Medicine, University of Illinois, Chicago, Illinois. Die Arbeit wurde bisher 19-mal zitiert (Crossref) und nennt 97 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: NICHD NIH HHS, NIMHD NIH HHS, NCATS NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Perinatal depression is one of the most common medical complications during pregnancy and postpartum period, affecting 10% to 20% of pregnant individuals, with higher rates among Black and Latina women who are also less likely to be diagnosed and treated. Machine learning (ML) models based on electronic medical records (EMRs) have effectively predicted postpartum depression in middle-class White women but have rarely included sufficient proportions of racial/ethnic minorities, which has contributed to biases in ML models. Our goal is to determine whether ML models could predict depression in early pregnancy in racial/ethnic minority women by leveraging EMR data.

Methods

We extracted EMRs from a large U.S. urban hospital serving mostly low-income Black and Hispanic women (n = 5875). Depressive symptom severity was assessed using the Patient Health Questionnaire-9 self-report questionnaire. We investigated multiple ML classifiers using Shapley additive explanations for model interpretation and determined prediction bias with 4 metrics: disparate impact, equal opportunity difference, and equalized odds (standard deviations of true positives and false positives).

Results

Although the best-performing ML model's (elastic net) performance was low (area under the receiver operating characteristic curve = 0.61), we identified known perinatal depression risk factors such as unplanned pregnancy and being single and underexplored factors such as self-reported pain, lower prenatal vitamin intake, asthma, carrying a male fetus, and lower platelet levels. Despite the sample comprising mostly low-income minority women (54% Black, 27% Latina), the model performed worse for these communities (area under the receiver operating characteristic curve: 57% Black, 59% Latina women vs. 64% White women).

Conclusions

EMR-based ML models could moderately predict early pregnancy depression but exhibited biased performance against low-income minority women. Perinatal depression affects 10% to 20% of pregnant individuals, with higher rates among racial/ethnic minorities who are underdiagnosed and undertreated. This study used machine learning models on electronic medical record data from a hospital serving mostly low-income Black and Hispanic women to predict early pregnancy depression. While the best model performed moderately well, it exhibited bias, predicting depression less accurately for Black and Latina women compared with White women. The study identified some known risk factors such as unplanned pregnancy and underexplored factors such as self-reported pain, lower prenatal vitamin intake, and carrying a male fetus that may contribute to perinatal depression.

Schlagwörter

Stichwörter der Autor:innen

Electronic medical recordsHealth disparitiesMachine learningModel performance biasPerinatal depression

Autor:innen und Institutionen

  • Department of Biomedical Engineering, Colleges of Engineering and Medicine, University of Illinois, Chicago, Illinois.
  • Division of Health Policy and Administration, School of Public Health, University of Illinois, Chicago, Illinois.
  • Department of Psychiatry, College of Medicine, University of Illinois, Chicago, Illinois.
  • Department of Psychology, College of Medicine, University of Illinois, Chicago, Illinois.
  • Department of Obstetrics and Gynecology, College of Medicine, University of Illinois, Chicago, Illinois.
  • Center of Bioinformatics and Quantitative Biology, University of Illinois, Chicago, Illinois.

Förderung

  • NICHD NIH HHS
  • NIMHD NIH HHS
  • NCATS NIH HHS
  • National Institutes of Health
  • National Institute on Minority Health and Health Disparities
  • National Center for Advancing Translational Sciences
  • National Institute of Mental Health
  • Eunice Kennedy Shriver National Institute of Child Health and Human Development

Zitieren

APA

Huang, Y., Alvernaz, S., Kim, S. J., Maki, P., Dai, Y., & Peñalver Bernabé, B. (2024). Predicting Prenatal Depression and Assessing Model Bias Using Machine Learning Models. Biological Psychiatry Global Open Science, 4(6), 100376. https://doi.org/10.1016/j.bpsgos.2024.100376

Vancouver

Huang Y, Alvernaz S, Kim SJ, Maki P, Dai Y, Peñalver Bernabé B. Predicting Prenatal Depression and Assessing Model Bias Using Machine Learning Models. Biological Psychiatry Global Open Science. 2024;4(6):100376. doi:10.1016/j.bpsgos.2024.100376

BibTeX
@article{Huang2024,
  author = {Huang, Yongchao and Alvernaz, Suzanne and Kim, Sage J and Maki, Pauline and Dai, Yang and Peñalver Bernabé, Beatriz},
  title = {Predicting Prenatal Depression and Assessing Model Bias Using Machine Learning Models},
  journal = {Biological Psychiatry Global Open Science},
  year = {2024},
  volume = {4},
  number = {6},
  pages = {100376},
  doi = {10.1016/j.bpsgos.2024.100376}
}
Kostenlos starten →