Development of a novel risk score to predict mortality in patients admitted to hospital with COVID-19

Area of Science:

• Medical Science
• Infectious Diseases
• Critical Care Medicine

Background:

• Hospitalised patients with Coronavirus Disease 2019 (COVID-19) face significant mortality risks.
• Early identification of high-risk individuals is crucial for timely intervention and improved patient outcomes.
• Objective prognostic guidance is needed for families and healthcare providers managing COVID-19 patients.

Purpose of the Study:

• To develop and validate a predictive score for 30-day mortality in hospitalised COVID-19 patients.
• To identify key clinical and biochemical factors associated with adverse outcomes.
• To create a simple, accessible tool for risk stratification.

Main Methods:

• A retrospective cohort study was conducted on 316 COVID-19 patients admitted to a UK hospital.
• Multivariable logistic regression analysis was used to develop a mortality prediction score.
• The score incorporated patient demographics, laboratory results (platelet count, international normalised ratio), and vital signs (Glasgow Coma Scale, respiratory rate, blood pressure).

Main Results:

• The developed mortality score demonstrated strong predictive power for 30-day mortality (Area Under Curve: 0.7933).
• An optimal cut-point (score ≥ 4) identified patients with a 7.6-fold increased odds of 30-day mortality.
• The score achieved a sensitivity of 78.36% and specificity of 67.59% at the optimal cut-point.

Conclusions:

• A simple, easy-to-use risk score effectively predicts 30-day mortality in hospitalised COVID-19 patients.
• The score can aid in identifying patients requiring closer monitoring and intensive early intervention.
• Further external validation is recommended, but the score shows potential for guiding prognosis and clinical management.