An Empirical Assessment of Machine Learning Models for Delirium Prediction Using the eICU-CRD v2.0 Database

• Márcio Wellington de Souza Hospital Israelita Albert Einstein / Hospital Beneficência Portuguesa
• Diego Minatel Universidade de São Paulo (USP)

DOI: https://doi.org/10.5753/sbbd_estendido.2026.249653

Resumo

Population aging intensifies the challenges faced by healthcare systems, particularly in managing conditions such as hospital-acquired delirium, a neuropsychiatric disorder characterized by acute onset and a fluctuating course, and associated with increased mortality and functional decline. Despite its clinical relevance, delirium frequently remains underdiagnosed during hospitalization. This paper investigates the use of machine learning to predict delirium early in intensive care units, using data from the eICU-CRD v2.0 database. A domain expert clinician selected the analyzed attributes, ensuring clinical coherence in the modeling process. We evaluated five classification algorithms and four data-balancing techniques using stratified cross-validation and metrics appropriate for imbalanced scenarios. Random Forest with Random Undersampling achieved the best overall performance. Our findings suggest that machine learning-based approaches can support earlier recognition of delirium and enhance clinical decision-making.

Palavras-chave: Delirium, Early Prediction, Healthcare Analytics, Imbalanced Classification, Clinical Decision Support

Referências

Bishara, A., Chiu, C., Whitlock, E. L., Douglas, V. C., Lee, S., Butte, A. J., Leung, J. M., and Donovan, A. L. (2022). Postoperative delirium prediction using machine learning models and preoperative electronic health record data. BMC anesthesiology, 22(1):8.

Contreras, M., Kapoor, S., Zhang, J., Davidson, A., Ren, Y., Guan, Z., Ozrazgat-Baslanti, T., Sena, J., Nerella, S., Bihorac, A., et al. (2025). A large language model for delirium prediction in the intensive care unit using structured electronic health records. Scientific Reports, 15(1):38890.

Davoudi, A., Ozrazgat-Baslanti, T., Ebadi, A., Bursian, A. C., Bihorac, A., and Rashidi, P. (2017). Delirium prediction using machine learning models on predictive electronic health records data. In 2017 IEEE 17th International Conference on Bioinformatics and Bioengineering (BIBE), pages 568–573. IEEE.

Ely, E. W., Shintani, A., Truman, B., Speroff, T., Gordon, S. M., Harrell Jr, F. E., Inouye, S. K., Bernard, G. R., and Dittus, R. S. (2004). Delirium as a predictor of mortality in mechanically ventilated patients in the intensive care unit. Jama, 291(14):1753–1762.

Faceli, K., Lorena, A. C., Gama, J., and Carvalho, A. C. P. d. L. F. d. (2021). Inteligência artificial: uma abordagem de aprendizado de máquina. LTC, Rio de Janeiro, 2nd edition.

Inouye, S. K., Westendorp, R. G., and Saczynski, J. S. (2014). Delirium in elderly people. The lancet, 383(9920):911–922.

Milbrandt, E. B., Deppen, S., Harrison, P. L., Shintani, A. K., Speroff, T., Stiles, R. A., Truman, B., Bernard, G. R., Dittus, R. S., and Ely, E. W. (2004). Costs associated with delirium in mechanically ventilated patients. Critical care medicine, 32(4):955–962.

Pollard, T. J., Johnson, A. E., Raffa, J. D., Celi, L. A., Mark, R. G., and Badawi, O. (2018). The eicu collaborative research database, a freely available multi-center database for critical care research. Scientific data, 5(1):1–13.

Sarker, I. H. (2021). Machine learning: Algorithms, real-world applications and research directions. SN computer science, 2(3):160.

United Nations (2019). World population ageing 2019: Highlights.

Viegas, A., Von Rekowski, C. P., Araújo, R., Viana-Baptista, M., Macedo, M. P., and Bento, L. (2025). Predicting icu delirium in critically ill covid-19 patients using demographic, clinical, and laboratory admission data: A machine learning approach. Life, 15(7):1045.

Witlox, J., Eurelings, L. S., de Jonghe, J. F., Kalisvaart, K. J., Eikelenboom, P., and Van Gool, W. A. (2010). Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. Jama, 304(4):443–451.

Yang, T., Yang, H., Liu, Y., Liu, X., Ding, Y.-J., Li, R., Mao, A.-Q., Huang, Y., Li, X.-L., Zhang, Y., et al. (2024). Postoperative delirium prediction after cardiac surgery using machine learning models. Computers in Biology and Medicine, 169:107818.

Zhang, Y., Hu, J., Hua, T., Zhang, J., Zhang, Z., and Yang, M. (2023). Development of a machine learning-based prediction model for sepsis-associated delirium in the intensive care unit. Scientific reports, 13(1):12697.