Machine Learning-Based Prediction of COVID-19 Prognosis Using Clinical and Hematologic Data

Area of Science:

• Medical Informatics
• Hematology
• Infectious Diseases

Background:

• The COVID-19 pandemic presents significant global healthcare challenges.
• Accurate prognosis prediction is crucial for managing COVID-19 patient care and resource allocation.

Purpose of the Study:

• To evaluate the efficacy of machine learning models in predicting COVID-19 patient outcomes and disease severity.
• To identify key clinical and hematological parameters that serve as predictors for COVID-19 prognosis.

Main Methods:

• A multicenter retrospective study involving 485 COVID-19 patients.
• Analysis of demographic data, symptoms, hematological variables, treatments, and clinical outcomes.
• Application and comparison of machine learning algorithms: random forest, multilayer perceptron, and support vector machine.

Main Results:

• Machine learning models demonstrated high performance in predicting disease severity and clinical outcomes, achieving an Area Under the Curve (AUC) of 0.96.
• Hematological parameters, specifically neutrophils, lymphocytes, D-dimer, and monocytes, were identified as the most significant predictors.
• All evaluated machine learning approaches exhibited comparable predictive capabilities.

Conclusions:

• Machine learning techniques are feasible and effective for predicting COVID-19 patient outcomes and severity.
• Hematological markers are critical indicators for assessing COVID-19 prognosis and patient outcomes.
• This study highlights the potential of leveraging routinely collected data for improved COVID-19 patient management.