Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Prediction of extubation failure among low birthweight neonates using machine learning

Journal of Perinatology volume 43, pages 209–214 (2023)Cite this article

Subjects

Abstract

Objective

To develop machine learning models predicting extubation failure in low birthweight neonates using large amounts of clinical data.

Study design

Retrospective cohort study using MIMIC-III, a large single-center, open-source clinical dataset. Logistic regression and boosted-tree (XGBoost) models using demographics, medications, and vital sign and ventilatory data were developed to predict extubation failure, defined as reintubation within 7 days.

Results

1348 low birthweight (≤2500 g) neonates who received mechanical ventilation within the first 7 days were included, of which 350 (26%) failed a trial of extubation. The best-performing model was a boosted-tree model incorporating demographics, vital signs, ventilator parameters, and medications (AUROC 0.82). The most important features were birthweight, last FiO2, average mean airway pressure, caffeine use, and gestational age.

Conclusions

Machine learning models identified low birthweight ventilated neonates at risk for extubation failure. These models will need to be validated across multiple centers to determine generalizability of this tool.

Access through your institution
Buy or subscribe

This is a preview of subscription content, access via your institution

Access options

Access through your institution

Buy this article

  • Purchase on SpringerLink
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Study Cohort.
Fig. 2: Receiver operator curves for each prediction model.

Similar content being viewed by others

Data availability

The datasets described in the manuscript are publicly accessible (https://physionet.org/content/mimiciii/1.4/). The models described are available upon request to the corresponding author.

References

  1. Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptook AR, Walsh MC, et al. Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics. 2010;126:443–56.

    Article  Google Scholar 

  2. Consortium on Safe Labor, Hibbard JU, Wilkins I, Sun L, Gregory K, Haberman S, et al. Respiratory morbidity in late preterm births. JAMA. 2010;304:419–25.

    Article  Google Scholar 

  3. Chavez TA, Lakshmanan A, Figueroa L, Iyer N, Stavroudis TA, Garingo A, et al. Resource utilization patterns using non-invasive ventilation in neonates with respiratory distress syndrome. J Perinatol. 2018;38:850–6.

    Article  Google Scholar 

  4. Berger J, Mehta P, Bucholz E, Dziura J, Bhandari V. Impact of early extubation and reintubation on the incidence of bronchopulmonary dysplasia in neonates. Am J Perinatol. 2014;31:1063–72.

    Article  Google Scholar 

  5. Williams E, Dassios T, Arnold K, Hickey A, Greenough A. Prolonged ventilation and postnatal growth of preterm infants. J Perinat Med. 2019;48:82–6.

    Article  Google Scholar 

  6. Robbins M, Trittmann J, Martin E, Reber KM, Nelin L, Shepherd E. Early extubation attempts reduce length of stay in extremely preterm infants even if re-intubation is necessary. J Neonatal Perinat Med. 2015;8:91–7.

    Article  CAS  Google Scholar 

  7. Vliegenthart RJS, van Kaam AH, Aarnoudse-Moens CSH, van Wassenaer AG, Onland W. Duration of mechanical ventilation and neurodevelopment in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2019;104:F631–5.

    Article  Google Scholar 

  8. Jensen EA, DeMauro SB, Kornhauser M, Aghai ZH, Greenspan JS, Dysart KC. Effects of multiple ventilation courses and duration of mechanical ventilation on respiratory outcomes in extremely low-birth-weight infants. JAMA Pediatr. 2015;169:1011–7.

    Article  Google Scholar 

  9. Chawla S, Natarajan G, Shankaran S, Carper B, Brion LP, Keszler M, et al. Markers of successful extubation in extremely preterm infants, and morbidity after failed extubation. J Pediatr. 2017;189:113–9.e2.

    Article  Google Scholar 

  10. Shalish W, Kanbar L, Kovacs L, Chawla S, Keszler M, Rao S, et al. The impact of time interval between extubation and reintubation on death or bronchopulmonary dysplasia in extremely preterm infants. J Pediatr. 2019;205:70–76.e2.

    Article  Google Scholar 

  11. Shalish W, Latremouille S, Papenburg J, Sant’Anna GM. Predictors of extubation readiness in preterm infants: a systematic review and meta-analysis. Arch Dis Child Fetal Neonatal Ed. 2019;104:F89–97.

    Article  Google Scholar 

  12. Dimitriou G, Greenough A, Endo A, Cherian S, Rafferty G. Prediction of extubation failure in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2002;86:F32–5.

    Article  CAS  Google Scholar 

  13. Szymankiewicz M, Vidyasagar D, Gadzinowski J. Predictors of successful extubation of preterm low-birth-weight infants with respiratory distress syndrome. Pediatr Crit Care Med. 2005;6:44–9.

    Article  Google Scholar 

  14. Gupta D, Greenberg RG, Sharma A, Natarajan G, Cotten M, Thomas R, et al. A predictive model for extubation readiness in extremely preterm infants. J Perinatol. 2019;39:1663–9.

    Article  Google Scholar 

  15. Silva MGF, Gregório ML, de Godoy MF. Does heart rate variability improve prediction of failed extubation in preterm infants? J Perinat Med. 2019;47:252–7.

    Article  Google Scholar 

  16. Precup D, Robles-Rubio CA, Brown KA, Kanbar L, Kaczmarek J, Chawla S, et al. Prediction of extubation readiness in extreme preterm infants based on measures of cardiorespiratory variability. Conf Proc IEEE Eng Med Biol Soc. 2012;2012:5630–3.

    Google Scholar 

  17. Kaczmarek J, Chawla S, Marchica C, Dwaihy M, Grundy L, Sant’Anna GM. Heart rate variability and extubation readiness in extremely preterm infants. Neonatology. 2013;104:42–8.

    Article  CAS  Google Scholar 

  18. Beam AL, Kohane IS. Big data and machine learning in health care. JAMA. 2018;319:1317–8.

    Article  Google Scholar 

  19. Shalish W, Kanbar LJ, Rao S, Robles-Rubio CA, Kovacs L, Chawla S, et al. Prediction of Extubation readiness in extremely preterm infants by the automated analysis of cardiorespiratory behavior: study protocol. BMC Pediatr. 2017;17:167.

    Article  Google Scholar 

  20. Mueller M, Almeida JS, Stanislaus R, Wagner CL. Can machine learning methods predict extubation outcome in premature infants as well as clinicians? J Neonatal Biol. 2013;2. https://doi.org/10.4172/2167-0897.1000118.

  21. Daunhawer I, Kasser S, Koch G, Sieber L, Cakal H, Tütsch J, et al. Enhanced early prediction of clinically relevant neonatal hyperbilirubinemia with machine learning. Pediatr Res. 2019;86:122–7.

    Article  Google Scholar 

  22. Turova V, Sidorenko I, Eckardt L, Rieger-Fackeldey E, Felderhoff-Müser U, Alves-Pinto A, et al. Machine learning models for identifying preterm infants at risk of cerebral hemorrhage. PLoS ONE. 2020;15:e0227419.

    Article  CAS  Google Scholar 

  23. Podda M, Bacciu D, Micheli A, Bellù R, Placidi G, Gagliardi L. A machine learning approach to estimating preterm infants survival: development of the Preterm Infants Survival Assessment (PISA) predictor. Sci Rep. 2018;8:13743.

    Article  Google Scholar 

  24. Mikhno A, Ennett CM. Prediction of extubation failure for neonates with respiratory distress syndrome using the MIMIC-II clinical database. Conf Proc IEEE Eng Med Biol Soc. 2012;2012:5094–7.

    Google Scholar 

  25. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD). Circulation. 2015;131:211–9.

    Article  Google Scholar 

  26. Johnson A, Pollard T, Mark R. MIMIC-III Clinical Database (version 1.4). PhysioNet. 2016. Available from: https://doi.org/10.13026/C2XW26.

  27. Johnson AEW, Pollard TJ, Shen L, Lehman LH, Feng M, Ghassemi M, et al. MIMIC-III, a freely accessible critical care database. Sci Data. 2016;3:160035.

    Article  CAS  Google Scholar 

  28. Goldberger AL, Amaral LA, Glass L, Hausdorff JM, Ivanov PC, Mark RG, et al. PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. Circulation. 2000;101:E215–20.

    Article  CAS  Google Scholar 

  29. Giaccone A, Jensen E, Davis P, Schmidt B. Definitions of extubation success in very premature infants: a systematic review. Arch Dis Child Fetal Neonatal Ed. 2014;99:F124–7.

    Article  Google Scholar 

  30. Chen T, Guestrin C. XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, NY, USA: Association for Computing Machinery; 2016. p. 785–94.

  31. Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B, et al. From local explanations to global understanding with explainable AI for trees. Nat Mach Intell. 2020;2:56–67.

    Article  Google Scholar 

  32. Keller RL, Feng R, DeMauro SB, Ferkol T, Hardie W, Rogers EE, et al. Bronchopulmonary dysplasia and perinatal characteristics predict 1-year respiratory outcomes in newborns born at extremely low gestational age: a prospective cohort study. J Pediatr. 2017;187:89–97.e3.

    Article  Google Scholar 

  33. Lodha A, Seshia M, McMillan DD, Barrington K, Yang J, Lee SK, et al. Association of early caffeine administration and neonatal outcomes in very preterm neonates. JAMA Pediatr. 2015;169:33–8.

    Article  Google Scholar 

Download references

Acknowledgements

The authors would like to thank the 2019 course directors of HST 953 Collaborative Data Science in Medicine at the Massachusetts Institute of Technology for their guidance and support in working with the MIMIC-III database.

Funding

ALB is funded by the NIH NHBLI (K01HL141771).

Author information

Authors and Affiliations

  1. Philips Research North America, Cambridge, MA, USA

    Annamalai Natarajan

  2. Department of Computer Science, Stanford University, Palo Alto, CA, USA

    Grace Lam

  3. Department of Biomedical Informatics, Harvard Medical School, Boston, MA, USA

    Jingyi Liu

  4. Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA

    Andrew L. Beam

  5. CAUSALab, Harvard T.H. Chan School of Public Health, Boston, MA, USA

    Andrew L. Beam

  6. Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, MA, USA

    Kristyn S. Beam & Jonathan C. Levin

  7. Division of Newborn Medicine, Boston Children’s Hospital, Boston, MA, USA

    Kristyn S. Beam & Jonathan C. Levin

Authors
  1. Annamalai Natarajan
    View author publications

    Search author on:PubMed Google Scholar

  2. Grace Lam
    View author publications

    Search author on:PubMed Google Scholar

  3. Jingyi Liu
    View author publications

    Search author on:PubMed Google Scholar

  4. Andrew L. Beam
    View author publications

    Search author on:PubMed Google Scholar

  5. Kristyn S. Beam
    View author publications

    Search author on:PubMed Google Scholar

  6. Jonathan C. Levin
    View author publications

    Search author on:PubMed Google Scholar

Contributions

AN designed the study, performed initial and subsequent analyses, and drafted the initial manuscript. JCL conceptualized and designed the study, provided supervision, participated in the initial and subsequent analyses, interpreted the data, and drafted the initial manuscript. KSB conceptualized and designed the study, participated in the initial and subsequent analyses, interpreted the data, and revised the manuscript. ALB conceptualized and designed the study, provided supervision, participated in the initial and subsequent analyses, and revised the manuscript. GL and JL participated in the initial analyses and revised the manuscript. All authors gave final approval to be published and agree to be accountable for all aspects of the work.

Corresponding author

Correspondence to Jonathan C. Levin.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Natarajan, A., Lam, G., Liu, J. et al. Prediction of extubation failure among low birthweight neonates using machine learning. J Perinatol 43, 209–214 (2023). https://doi.org/10.1038/s41372-022-01591-3

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue date:

  • DOI: https://doi.org/10.1038/s41372-022-01591-3

This article is cited by

Search

Advanced search

Quick links