Identification of high-risk COVID-19 patients using machine learning

Area of Science:

• Computational biology
• Epidemiology
• Medical informatics

Background:

• The COVID-19 pandemic caused by SARS-CoV-2 has led to significant global mortality and economic impact.
• Accurate prediction of patient outcomes is crucial for effective resource allocation and treatment strategies.

Purpose of the Study:

• To develop and validate a machine-learning algorithm for predicting COVID-19 patient survival.
• To assist healthcare professionals in identifying high-risk patients for timely intervention and improved hospital capacity planning.

Main Methods:

• A machine-learning algorithm was trained using historical data from confirmed and suspected COVID-19 cases in Mexico.
• The dataset included medical history, demographic information, and COVID-19 specific data.
• The algorithm was evaluated for its accuracy in predicting survival versus mortality across different clinical stages.

Main Results:

• The machine-learning algorithm demonstrated high accuracy in identifying high-risk COVID-19 patients.
• The method proved effective across four distinct clinical stages of the disease.
• The algorithm's predictive capabilities can enhance hospital capacity planning and facilitate prompt medical treatment.

Conclusions:

• The developed machine-learning tool offers a valuable resource for medical professionals in assessing COVID-19 patient prognosis.
• The algorithm can support real-time decision-making for prioritizing healthcare needs during the pandemic.
• The methodology shows potential for application in statistical hypothesis testing within biological and medical research.