Usefulness of the neutrophil-to-lymphocyte ratio in predicting the severity of COVID-19 patients: a retrospective cohort study

ABSTRACT BACKGROUND: Quick and accurate identification of critically ill patients ensures appropriate and correct use of medical resources. In situations that threaten public health, like pandemics, rapid and effective methods are needed for early disease detection among critically ill patients. OBJECTIVE: To determine the relationship between the neutrophil-to-lymphocyte ratio (NLR) of coronavirus disease-19 (COVID-19) patients upon admission to the emergency department (ED) and these patients’ prognosis. DESIGN AND SETTING: Retrospective cohort study among COVID-19 patients in the ED of a tertiary-level hospital. METHODS: Data on patients’ age, gender, vital signs, chronic diseases, laboratory tests and clinical outcomes were collected from electronic medical records. Receiver operating characteristic (ROC) curve analysis was performed. The area under the curve (AUC) was used to assess the accuracy of NLR for predicting in-hospital mortality risk and intensive care unit (ICU) requirement. The Youden J index (YJI) was used to determine optimal threshold values. RESULTS: 1,175 patients were included. Their median age was 63 years (IQR, 48-75). With an NLR cutoff value of 5.14, the sensitivity, specificity, PPV, AUC and YJI for ICU requirement were calculated as 77.87%, 74.08%, 92.4%, 0.811 and 0.5194, respectively. With the same cutoff value, the sensitivity, specificity, AUC and YJI for in-hospital mortality were 77.27%, 75.82%, 0.815 and 0.5309, respectively. In addition, advanced age, leukocytosis, anemia and lymphopenia were found to be associated with poor prognosis. CONCLUSION: The NLR, which is a widely available simple parameter, can provide rapid insights regarding early recognition of critical illness and prognosis among COVID-19 patients.


METHODS
This retrospective cohort study was carried out in the emergency department (ED) of Kartal Dr. Lütfi Kırdar City Hospital between October 1, 2020, and March 1, 2021. The hospital's institutional review board approved the analysis and issued a waiver of consent (Ethics Committee Ruling number: 2021/514/198/30; date: March 29, 2021).
All COVID-19 patients over the age of 18 who were hospitalized between October 1, 2020, and March 1, 2021, were included in this study. The diagnosis of COVID-19 was determined based on the WHO guidelines. This study includes only patients who had positive results from a real-time reverse transcriptase-polymerase chain reaction (RT-PCR) test on nasal and pharyngeal swab samples. 12 A data form was used to collect patients' age, gender, vital signs, chronic diseases, laboratory tests and clinical outcome data from the electronic medical records.

Measurements
Chronic diseases presented by these patients, such as chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), coronary artery disease (CAD), atrial fibrillation (AF), chronic renal failure (CRF), chronic neurological disease (CND), diabetes mellitus (DM) and hypertension (HT) were recorded by scanning digital file records stored in the hospital information management system (HIMS).
The NLR was calculated using the following simple formula: Absolute number of neutrophils/Absolute number of lymphocytes.

Outcomes
ED outcomes were determined as inpatient unit (IU) or ICU; and hospitalization outcomes as survivor or non-survivor.

Data analysis
Statistical analyses were performed using the Statistical Package Continuous data did not meet the assumption of normality.
The Mann-Whitney U test was used for analyses on continuous data and the chi-square test was used for analyses on categorical data. Continuous data were reported as medians and interquartile ranges (IQR; 25 th to 75 th percentile), while categorical data were reported as frequencies and percentages (Tables 1 and 2).
A P-value of less than 0.05 was considered statistically significant.

RESULTS
This study was conducted using data from 1,175 patients, among whom 439 were women and 483 were men. There were 889 patients in the survivor group and 286 patients in the non-survivor group; and there were 922 patients in the IU group and 253 patients in the ICU group ( Table 1).
The median age of the population included in the study was 63 years (IQR: 48-75), with a minimum age of 19 Table 2).
Analysis on the effects of chronic diseases on the prognosis for COVID-19 showed that there were significant differences between the survivor and non-survivor groups and between the IU and ICU groups regarding COPD, CHF, CAD, AF, CRF and CND.
While there was a significant difference between the survivor and non-survivor groups regarding hypertension (HT), there was no significant difference between the IU and ICU groups ( Table 1).
While there were statistically significant differences in both pairs of groups regarding HR and SpO 2 , which are among the vital signs, there was no statistically significant difference regarding SBP ( Table 2). While significant differences (P < 0.001) were detected in both pairs of groups regarding WBC, Neu, Lym, Hgb, Plt and NLR among the laboratory parameters, there was no significant difference in either pair of groups regarding monocytes ( Table 2).
The data on the study population are presented in Tables 1 and 2. Predictive values for NLR, in relation to in-hospital mortality and ICU requirement, were analyzed by means of ROC analysis.

DISCUSSION
COVID-19 has a significant impact on the hematopoietic system. Dysregulation of the hematological and immunological systems plays a key role in the pathological process of this infection. 15 In this study, we concluded that the NLR values of COVID-19 patients at the time of admission to the ED can be used as a predictor for ICU requirement and mortality risk. In addition, advanced age, leukocytosis, anemia and lymphopenia were found to be associated with poor prognosis. Over recent years, the diagnostic and prognostic accuracy of various ratios such as neutrophil-to-lymphocyte, thrombocyte-to-lymphocyte and monocyte-to-lymphocyte have been studied in relation to many inflammatory conditions. Similar studies are ongoing in the COVID-19 pandemic.
In a study based on retrospective analysis of clinical data from 443 patients with COVID-19, Shang et al. reported that NLR, IU = inpatient unit; ICU = intensive care unit; IQR = interquartile range (25 th to 75 th percentile); SBP = systolic blood pressure; HR = heart rate; SpO 2 = blood oxygen saturation; WBC = white blood cells; Lym = lymphocytes; Neu = neutrophils; Mon = monocytes; Hgb; hemoglobin; Plt = platelets, NLR = neutrophil-tolymphocyte ratio.  In a recent study conducted in Turkey, the relationship between ICU requirement for COVID-19 patients and their hemogram parameters at the time of initial admission was investigated.
It was highlighted that high NLR and monocyte-to-lymphocyte and low platelet-to-lymphocyte ratios can be predictors for ICU requirement. 17 In a prospective study conducted in Pakistan, it was reported that use of the NLR successfully enabled early recognition of severe conditions among patients with COVID-19 pneumonia.
In a study assessing the prognostic accuracy of the NLR, the AUC was calculated as 0.831 and the YJI was 0.589. In addition, it was reported that, for the optimum NLR threshold value of 4.795, the sensitivity was 83.9% and the specificity was 75%. 18 In our study, the best threshold value was found to be 5.14. In terms of ICU requirement, we calculated the sensitivity as 77.87%, specificity 74.08%, AUC 0.811 and YJI 0.519 for this NLR cutoff value of 5.14. Evaluation of the same cutoff value in terms of mortality prediction gave rise to sensitivity calculated as 72.27%, specificity 50.7%, AUC 0.815 and YJI 0.530.
The immune system is the system that is most affected by COVID-19 infection, after the respiratory system. 19 Therefore, it is not surprising that the NLR has high predictive accuracy. In assessing the pathogenesis of the disease, it is seen that necrosis, bleeding and atrophy occur in the spleen, and also that there are signif-    and bacteremia. 21 In a meta-analysis examining 15 studies, it was reported that the neutrophil count and NLR were higher, but that the lymphocyte count was lower in severe COVID-19 cases, compared with non-severe cases. 22 Additionally, recent studies have reported that NLR can be a reliable predictor, not only for inflammatory diseases and infections, but also for other acute medical conditions, including cerebral hemorrhage, acute coronary syndrome and ischemic stroke. [23][24][25] In general, elderly patients have been shown to be the "most vulnerable" group with regard to COVID-19 mortality. 26,27 In one study, COVID-19 patients aged 60 years and over were shown to have greater severity of clinical outcomes and higher mortality rates, compared with those who were under 60. 28 Similarly, in our study, we concluded that greater age was associated with increased risk of ICU requirement and mortality.
This study had certain limitations, such as having a relatively small sample size and being a single-center study. For more accurate and precise results, wider generalizability of the findings and validation of our results, multicenter clinical studies with larger sample sizes are required.

CONCLUSION
The NLR, which is a widely available simple parameter, can provide rapid insights regarding early recognition of critical illness and prognosis among COVID-19 patients.