Prediction of Clinical Deterioration in Hospitalized Adult Patients with Hematologic Malignancies Using a Neural Network Model

摘要: Introduction Clinical deterioration (ICU transfer and cardiac arrest) occurs during approximately 5–10% of hospital admissions. Existing prediction models have a high false positive rate, leading to multiple alarms alarm fatigue. We used routine vital signs laboratory values obtained from the electronic medical record (EMR) along with machine learning algorithm called neural network develop model that would increase predictive accuracy decrease rates. Design Retrospective cohort study. Setting The hematologic malignancy unit in an academic center United States. Patient Population Adult patients admitted 2009 2010. Intervention None. Measurements Main Results Vital were system then as predictors (features). A was build predict clinical events arrest). The performance compared VitalPac Early Warning Score (ViEWS). Five hundred sixty five consecutive total admissions available 43 resulting deterioration. Using simulation, outperformed ViEWS value 82% 24%, respectively. Conclusion We developed tested network-based for hospitalized unit. Our existing model, substantially increasing value, allowing clinician be confident raised. This can readily implemented real-time fashion EMR systems.