The Strong Predictive Role and Their Day-Dependent Behavior of Blood Urea Nitrogen and Complete Blood Count in COVID-19 Inpatients Prognosis

Published in Medical Laboratory Journal, 2024

Title

The Strong Predictive Role and Their Day-Dependent Behavior of Blood Urea Nitrogen and Complete Blood Count in COVID-19 Inpatients Prognosis

Authors

Mohammad Fayaz, Vahid Tajari, Mohammad Hosein Taziki Balajelini, Abdolhalim Rajabi, and Seyed Mehran Hosseini


Journal Information

Journal: Medical Laboratory Journal (MLJ)
Year: 2024
Volume: 18
Issue: 1
Pages: 16–22


Keywords

  • COVID-19
  • Survival Analysis
  • Cox Regression
  • Generalized Additive Models
  • Blood Urea Nitrogen (BUN)
  • Complete Blood Count (CBC)
  • Prognosis
  • Hospitalized Patients
  • Mortality Prediction
  • Clinical Biomarkers

Abstract

Background

The prognosis of hospitalized COVID-19 patients can be predicted using demographic, clinical, laboratory, and imaging indicators. Identifying the strongest predictors and understanding their dynamic behavior during hospitalization can improve clinical decision-making and patient management.

Methods

This retrospective cohort study included moderate-to-severe COVID-19 patients hospitalized between 26 January 2020 and 13 January 2021. The hospitalization period was defined as the interval between admission and discharge, with the final outcome categorized as either survival or death.

Demographic, clinical, and laboratory information were extracted from the hospital information system. A total of 131 laboratory variables were evaluated. The analysis employed:

  • Generalized Additive Models (GAM)
  • Cox Proportional Hazards Regression Models

to investigate both static and time-varying prognostic effects.

Results

Among 1,520 hospitalized COVID-19 patients:

  • 232 patients (15.26%) died
  • 1,288 patients survived or reached the end of the 30-day follow-up period

Key findings included:

  • Blood Urea Nitrogen (BUN) levels were approximately twice as high in deceased patients:
    • Deceased: 44.60 ± 25.41 mg/dL
    • Survived: 21.30 ± 13.32 mg/dL
  • Lymphocyte counts demonstrated the opposite pattern and were lower among patients with poor outcomes.

  • Both BUN and lymphocyte count showed statistically significant hazard ratios in survival models.

  • Dynamic trend analyses demonstrated a rapid increase in hazard functions after approximately 15 days of hospitalization, indicating a critical period for disease progression.

Conclusion

Blood urea nitrogen and complete blood count biomarkers provide strong prognostic information for hospitalized COVID-19 patients. Their dynamic changes, particularly during the second week of hospitalization, can serve as early warning indicators of unfavorable outcomes and may support clinical monitoring and risk stratification strategies.


Full Text


Research Significance

This study highlights the prognostic value of routinely collected laboratory biomarkers in COVID-19 management. By combining survival analysis with flexible generalized additive models, the research demonstrates that both baseline values and temporal changes in laboratory markers contain clinically meaningful information regarding patient outcomes.

The findings emphasize the importance of:

  • Dynamic monitoring of laboratory indicators
  • Early identification of high-risk patients
  • Time-dependent risk assessment during hospitalization
  • Integration of laboratory trajectories into prognostic models

Citation

Fayaz M, Tajari V, Taziki Balajelini MH, Rajabi A, Hosseini SM. The strong predictive role and their day-dependent behavior of blood urea nitrogen and complete blood count in COVID-19 inpatients prognosis. Medical Laboratory Journal. 2024;18(1):16–22.

Recommended citation: Fayaz M, Tajari V, Taziki Balajelini MH, Rajabi A, Hosseini SM. The strong predictive role and their day-dependent behavior of blood urea nitrogen and complete blood count in COVID-19 inpatients prognosis. Medical Laboratory Journal. 2024;18(1):16–22.
Download Paper