Neutrophil-to-Lymphocyte Ratio in COVID-19 as a Predictor for Hospital Admission, Need for ICU Admission and 30-Day Mortality

Abstract

Rationale: The neutrophil-to-lymphocyte ratio (NLR) as a predictor of outcomes has been studied in different disease states such as sepsis, ARDS and pancreatitis. In COVID-19 patients, NLR has been shown to be an independent risk factor associated with disease severity with a few studies suggesting an association between NLR and outcomes. Using data from a large multi-hospital healthcare system over a 6-month period, we sought to independently evaluate the NLR's ability to predict need for hospitalization, ICU admission, need for mechanical ventilation, 30-day mortality, and its association with hypoxemia. Methods: We collected demographic, historic and clinical data from patients with a naso-and/or oropharyngeal swab positive for SARS-CoV-2 PCR visiting the ED for COVID-19-related complaints between March-September 2020. Fraction of inspired oxygen (FiO2) was calculated using 1 liter per minute (lpm) supplemental O2 to 3% FiO2 conversion and capped at 15 lpm O2. We examined the association of NLR with hypoxemia, need for hospitalization, ICU admission, need for mechanical ventilation within 7 days of ED visit and 30-day mortality after adjustment for history of hypertension, diabetes, severity of hypoxemia and age. Results: From 510 patients analyzed, 357 had a complete blood count (CBC) drawn (median age 57, 51% female). Median NLR was 2.09 with IQR of 3.69-5.88. NLR negatively correlated with degree of hypoxemia as quantified by SpO2-FiO2 (SF) ratio (Rho =-0.33, p < 0.001). Patients requiring hospital admission had higher NLR compared to those who did not (median 2.59 IQR [4.45-7.6] vs 1.57 [2.75-4.36], p < 0.001). In patients admitted to the hospital, NLR was higher in those requiring ICU admission compared to those who did not [2.99 [6.21-10.8] vs 2.37 [4.05-6.05], p < 0.01]. In those with NLR above the median there was a trend towards higher need for intubation within 7 days, although this did not achieve statistical significance (OR 2.28, 95% CI 0.94-5.53, p = 0.07). 30-day mortality was higher in those with NLR above the median compared to those below the median (OR 2.84, 95% CI 1.29-6.28, p < 0.01). Conclusions: In patients presenting to the ED with COVID-19, an increased NLR is associated with need for hospitalization, ICU admission and worse 30-day mortality while being inversely correlated with degree of hypoxemia. The cost-effectiveness and wide-spread availability of CBC testing makes NLR an easily implementable prognostication tool in COVID-19 patients.