Managing COVID-19 With a Clinical Decision Support Tool in a Community Health Network: Algorithm Development and...

Area of Science:

• Infectious Diseases
• Clinical Informatics
• Health Services Research

Background:

• The COVID-19 pandemic has caused significant global morbidity and mortality.
• Healthcare systems face strain due to high intensive care needs.
• Effective COVID-19 severity assessment is crucial for triage and resource allocation.

Purpose of the Study:

• To develop, validate, and scale a clinical decision support system (CDSS) and mobile application.
• To aid in COVID-19 severity assessment, management, and patient care.
• To provide an evidence-based tool for healthcare providers.

Main Methods:

• A two-tiered model was developed using data from 701 COVID-19 patients.
• Tier 1 uses non-laboratory data for initial assessment (hospitalization/testing needs).
• Tier 2 predicts mortality risk using biomarkers (CRP, procalcitonin, D-dimer) and age.
• Models were validated on external datasets from Wuhan, China.

Main Results:

• Biomarker levels were significantly higher in non-survivors (P<.001).
• Internal validation showed AUCs of 0.79 for Tier 1 and 0.95 for Tier 2.
• External validation demonstrated AUCs of 0.79 for Tier 1 and 0.97 for Tier 2.
• Both models showed strong predictive performance.

Conclusions:

• The CDSS and mobile app are validated and ready for clinical use.
• These tools can assist healthcare providers in evidence-based COVID-19 management.
• Deployment can improve patient outcomes and support cost containment in community settings.