Predicting Mortality and the Need for Early Intubation in Intensive Care Patients Using Machine Learning: A Pilot Study

Abstract

Aim: This pilot study aimed to assess the feasibility and preliminary performance of machine learning models for predicting ICU mortality and the need for intubation within 48 hours using routinely collected admission data.

Method: Ten adult intensive care patients were included in this single-center, retrospective observational pilot study. For ANN development, 13 prespecified clinical and laboratory features per patient were restructured in long format, yielding 130 feature-level records. These records represented feature entries derived from 10 patients rather than independent patient-level observations. Demographic characteristics, primary diagnosis, APACHE II and SOFA scores, arterial blood gas parameters, oxygenation indicators (FiO₂, PaO₂/FiO₂), and routine metabolic–biochemical variables were analyzed. ML-based classification models predicted ICU mortality and the need for intubation within the first 48 hours.

Results: In the exploratory feature-level internal assessment, both the ICU mortality and early-intubation classifications yielded an AUC of 1.00, sensitivity of 100%, specificity of 100%, and an F1-score of 1.00. These estimates were derived from records originating from 10 patients and should be interpreted cautiously.

Conclusion: This pilot study demonstrates the feasibility of applying machine learning methods to routinely collected ICU admission data. The findings are preliminary and hypothesis-generating and should not be interpreted as established clinical predictive performance. Validation in larger, independent, prospective, and multicenter cohorts is required.

Keywords

• Intensive care
• machine learning
• mortality
• intubation
• pilot study
• decision support

Supporting Institution

The authors declare that no financial support was received during the conduct of this study.

Ethical Statement

Ethical approval for this study was obtained from the Non-Interventional Clinical Research Ethics Committee of Niğde Ömer Halisdemir University Faculty of Medicine (Decision No: 2026/3). The study was conducted in accordance with the principles of the Declaration of Helsinki.

Yoğun Bakım Hastalarında Mortalite ve Erken Entübasyon Gereksiniminin Makine Öğrenmesi ile Öngörülmesi: Pilot Çalışma

Abstract

Amaç: Bu pilot çalışmada, rutin yatış verilerine dayalı makine öğrenmesi modellerinin yoğun bakım mortalitesi ile ilk 48 saat içinde entübasyon gereksinimini öngörmedeki uygulanabilirliğinin ve ön performansının değerlendirilmesi amaçlandı.

Yöntem: Tek merkezli, retrospektif, gözlemsel pilot çalışmaya 10 erişkin yoğun bakım hastası dahil edildi. Yapay sinir ağı geliştirilmesi için her hastaya ait önceden belirlenmiş 13 klinik ve laboratuvar özelliği uzun veri formatında yeniden düzenlenerek toplam 130 özellik düzeyinde kayıt oluşturuldu. Bu kayıtlar, bağımsız hasta gözlemlerini değil, 10 hastadan türetilen özellik kayıtlarını temsil etmekteydi. Demografik özellikler, primer tanı, APACHE II ve SOFA skorları, arter kan gazı parametreleri, oksijenasyon göstergeleri (FiO₂, PaO₂/FiO₂) ve rutin metabolik–biyokimyasal değişkenler analiz edildi. Yoğun bakım mortalitesi ve ilk 48 saat içinde entübasyon gereksiniminin öngörülmesi amacıyla ML tabanlı sınıflandırma modelleri kullanıldı.
Bulgular: Özellik düzeyindeki keşifsel iç değerlendirmede, hem yoğun bakım mortalitesi hem de erken entübasyon sınıflandırması için AUC 1,00, duyarlılık %100, özgüllük %100 ve F1-skoru 1,00 olarak bulundu. Bu tahminler 10 hastadan türetilen kayıtlara dayandığından dikkatle yorumlanmalıdır.

Sonuç: Bu pilot çalışma, yoğun bakım yatışında rutin olarak elde edilen verilere makine öğrenmesi yöntemlerinin uygulanabilir olduğunu göstermektedir. Bulgular ön ve hipotez oluşturucu nitelikte olup, yerleşik bir klinik öngörü performansının kanıtı olarak değerlendirilmemelidir. Modellerin daha geniş, bağımsız, prospektif ve çok merkezli hasta gruplarında doğrulanması gereklidir.

Keywords

• Yoğun bakım
• makine öğrenmesi
• mortalite
• entübasyon
• pilot çalışma
• karar destek sistemi

Supporting Institution

Yazarlar, bu çalışmanın yürütülmesi sırasında herhangi bir finansal destek alınmadığını beyan eder.

Ethical Statement

Bu çalışma için etik onay, Niğde Ömer Halisdemir Üniversitesi Tıp Fakültesi Girişimsel Olmayan Klinik Araştırmalar Etik Kurulu’ndan alınmıştır (Karar No: 2026/3). Çalışma, Helsinki Bildirgesi ilkelerine uygun olarak yürütülmüştür.

References

1. 1. Sun H, Kang M, Zhang H, Jia J, Wang Q. Machine learning for predicting mortality in intensive care unit patients: a prognostic performance systematic review and meta-analysis. Nursing in Critical Care. 2025;30(6):e70206.
2. 2. Iwase S, Nakada T, Shimada T, et al. Prediction algorithm for ICU mortality and length of stay using machine learning. Scientific Reports. 2022;12:12912.
3. 3. Pang K, Li L, Ouyang W, Liu X, Tang Y. Establishment of ICU mortality risk prediction models with machine learning algorithm using MIMIC-IV database. Diagnostics. 2022;12(5):1068.
4. 4. Lim L, Gim U, Cho K, et al. Real-time machine learning model to predict short-term mortality in critically ill patients: development and international validation. Critical Care. 2024;28(1):76.
5. 5. Thadani S, Wu TC, Wu DTY, et al. Machine learning-based prediction model for ICU mortality after continuous renal replacement therapy initiation in children. Critical Care Explorations. 2024;6(12):e1188.
6. 6. Yang S, Sun Y, Wang T, et al. Machine learning-based prediction of mortality and multidrug-resistant infection risks in ICU patients with suspected infection: a prospective national multicenter cohort study. BMC Infectious Diseases. 2026;26:139.
7. 7. Liu J, Duan X, Duan M, et al. Development and external validation of an interpretable machine learning model for the prediction of intubation in the intensive care unit. Scientific Reports. 2024;14:27174.
8. 8. Li R, Xu Z, Xu J, et al. Predicting intubation for intensive care units patients: a deep learning approach to improve patient management. International Journal of Medical Informatics. 2024;186:105425.

9. 9. Venturini M, Van Keilegom I, De Corte W, Vens C. Predicting time-to-intubation after critical care admission using machine learning and cured fraction information. Artificial Intelligence in Medicine. 2024;150:102817.
10. 10. Amory IA, Rashidi Khazaee P, Yousefi S. Improving sepsis mortality prediction with machine learning using full region synthetic sampling approach. Health Science Reports. 2025;8(11):e71556.
11. 11. Gao J, Lu Y, Ashrafi N, et al. Prediction of sepsis mortality in ICU patients using machine learning methods. BMC Medical Informatics and Decision Making. 2024;24:228.
12. 12. An S, Ye Z, Che W, et al. Development and validation of machine learning models to predict in-hospital mortality in ICU patients with sepsis and chronic kidney disease. BMC Infectious Diseases. 2025;25(1):1504.
13. 13. Silveira EC, Pretti SM, Santos BA, et al. Prediction of hospital mortality in intensive care unit patients from clinical and laboratory data: a machine learning approach. World Journal of Critical Care Medicine. 2022;11(5):317–329.
14. 14. Li H, Ashrafi N, Kang C, Zhao G, Chen Y, Pishgar M. A machine learning-based prediction of hospital mortality in mechanically ventilated ICU patients. PLoS ONE. 2024;19(9):e0309383.
15. 15. Weaver M, Goodin DA, Miller HA, et al. Prediction of prolonged mechanical ventilation in the intensive care unit via machine learning: a COVID-19 perspective. Scientific Reports. 2024;14:30173.
16. 16. Zhang G, Xie Q, Wang C, Xu J, Liu G, Su C. Intelligent alert system for predicting invasive mechanical ventilation needs via noninvasive parameters: employing an integrated machine learning method with integration of multicenter databases. Medical & Biological Engineering & Computing. 2024;62(11):3445–3458.
17. 17. Hannon DM, Syed JDA, McNicholas B, Madden M, Laffey JG. The development of a C5.0 machine learning model in a limited data set to predict early mortality in patients with ARDS undergoing an initial session of prone positioning. Intensive Care Medicine Experimental. 2024;12(1):103.
18. 18. Misseri G, Piattoli M, Cuttone G, Gregoretti C, Bignami EG. Artificial intelligence for mechanical ventilation: a transformative shift in critical care. Therapeutic Advances in Pulmonary and Critical Care Medicine. 2024;19:29768675241298918.
19. 19. El-Rashidy N, Tarek Z, Elshewey AM, et al. Multitask multilayer-prediction model for predicting mechanical ventilation and the associated mortality rate. Neural Computing and Applications. 2025;37:1321–1343.