Serial measurements of lymphocytes, D-dimer, LDH, and CRP do not improve the ability to predict COVID-19 adverse outcomes

Abstract

We evaluated the discriminatory ability of variations in lymphocyte, D-dimer, C-reactive protein (CRP), and lactate dehydrogenase (LDH) serum levels at 48 to 72 hours of hospitalization compared with baseline measurements to predict unfavorable clinical outcomes in patients with COVID-19. We analyzed diagnostic test results based on a retrospective cohort to determine the ability of variations (gradients or ratios) in patients' lymphocyte, D-dimer, CRP, and LDH serum levels taken 48 to 72 hours after hospital admission to predict adverse outcomes such as death, mechanical ventilation, or intensive care unit (ICU) admission developing. Among 810 patients (56.1% men, age 61.6 ± 16.2 years), 37.5% had at least one adverse outcome; 28.2% required ICU admission, 26.5% required mechanical ventilation, and 19.4% died during hospitalization. In comparing baseline measurements with measurements at 48 to 72 hours, D-dimer, lymphocyte delta, LDH, and CRP had similar discriminatory ability (area under the receiver operating characteristic curve [AUC] 0.57 vs. 0.56, 0.53 vs. 0.57, 0.64 vs. 0.66, and 0.62 vs. 0.65, respectively). Measuring serum risk markers upon hospital admission can be used to evaluate risk of adverse outcomes in hospitalized patients with COVID-19. Repeating these measurements at 48 to 72 hours does not improve discriminatory ability.