The Predictive Score for Patients Hospitalized With COVID-19 in Resource-Limited Settings

Area of Science:

• Infectious Diseases
• Critical Care Medicine
• Biostatistics

Background:

• The second wave of COVID-19 severely impacted India and developing nations, overwhelming healthcare systems.
• Despite improved understanding and treatments, mortality rates increased, highlighting the need for better patient triage.
• Existing predictive models often require online resources, limiting accessibility in resource-constrained settings.

Purpose of the Study:

• To identify simple, routinely available laboratory tests as predictors of intensive care unit (ICU) admission and mortality in hospitalized COVID-19 patients.
• To develop an easy-to-use scoring system for risk stratification.
• To aid in the efficient allocation of limited healthcare resources during COVID-19 waves.

Main Methods:

• Retrospective review of an institutional database of patients hospitalized during the second COVID-19 wave.
• Analysis of routine laboratory tests to identify predictors of ICU admission and mortality.
• Calculation of odds ratios (OR) and confidence intervals (CI) to assess the predictive value of identified variables.

Main Results:

• Absolute neutrophil count >4,200 predicted ICU admission (OR: 3.1), and >7,200 predicted mortality (aOR: 4.2).
• Blood urea >45 (aOR: 8.0), serum ferritin >500 (aOR: 2.7), and LDH >675 (OR: 9.2) predicted ICU admission.
• A right shift in partial pressure of oxygen (p50c) >26.5 (OR: 2.6) and serum protein <7 g/dL (OR: 2.8) also predicted ICU admission.

Conclusions:

• Several easily accessible laboratory parameters can effectively predict severe COVID-19 outcomes.
• A novel scoring system based on these predictors can help identify high-risk patients needing ICU care.
• This approach supports optimized resource management in healthcare settings facing surges in COVID-19 cases.