Reuse of Clinical COVID-19 Patient Data: Pre-Processing for Future Classification

Area of Science:

• Medical Informatics
• Public Health
• Computational Biology

Background:

• The COVID-19 pandemic necessitates tools to predict patient outcomes.
• Identifying patients at high risk for severe disease is critical for resource allocation and treatment.
• Existing methods require enhancement for accurate clinical evolution prediction.

Purpose of the Study:

• To develop a decision support system for identifying COVID-19 patients likely to experience worse clinical evolution.
• To leverage advanced statistical methods for feature extraction from patient data.
• To pinpoint key indicators associated with severe COVID-19 outcomes.

Main Methods:

• Data collection and pre-processing to ensure dataset consistency.
• Application of advanced statistical techniques, including Principal Component Analysis (PCA).
• Feature extraction to identify significant predictors of clinical outcomes.

Main Results:

• Preliminary analysis identified influential features for predicting severe COVID-19.
• The presence of cardiac and liver illnesses emerged as significant factors.
• Elevated inflammatory parameters at the time of diagnosis were highly influential.

Conclusions:

• Decision support systems can be effectively developed using patient data and statistical analysis.
• Cardiac and liver comorbidities, along with inflammatory markers, are crucial for predicting COVID-19 severity.
• These findings can aid medical staff in proactive patient management and intervention.