Abstract
We investigated the association between complete blood count, including neutrophil-to-lymphocyte ratio (NLR) in combination with patient characteristics, and coronavirus disease (COVID-19) outcomes to identify the best prognostic indicator. We analyzed data of patients with confirmed COVID-19 from the nationwide database of the Japan COVID-19 Task Force between February 2020 and November 2021. A composite outcome was defined as the most severe condition, including noninvasive positive-pressure ventilation, high-flow nasal cannula, invasive mechanical ventilation, extracorporeal membrane oxygenation, or death. Of 2425 patients in the analysis, 472 (19.5%) experienced a composite outcome. NLR was the best predictor of composite outcomes, with an area under the curve (AUC) of 0.81, and a sensitivity and specificity of 72.3% and 75.7%, respectively, using a cut-off value of 5.04. The combination of NLR and an oxygen requirement on admission had the highest AUC (0.88). This simple combination may help identify patients at risk of progression to severe disease.
Keywords: COVID-19, SARS-CoV-2, Prognostic indicator, Neutrophil-to-lymphocyte ratio, Area under the curve
1. Introduction
A simple method of predicting severe outcomes is required in institutions with several patients with coronavirus disease (COVID-19). Changes in the blood count are outcome predictors in these patients. Generally, higher neutrophil and lower lymphocyte counts are associated with severe disease [1]. The neutrophil-to-lymphocyte ratio (NLR) is a better predictor of mortality than the complete blood count (CBC) [2]. Additionally, a higher heart rate or respiratory rate, and lower oxygen saturation, are associated with higher mortality [3]. We investigated the association between CBC, including the NLR, and severe COVID-19 outcomes. We also determined the best combination of prognostic indicators.
2. Patients and methods
This retrospective cohort study analyzed data from the nationwide database of the Japan COVID-19 Task Force from February 2020 to November 2021 [4]. The study was approved by the Ethics Committees of the Keio University School of Medicine (IRB#: 20200061) and related research institutions. Written or oral informed consent was obtained from all patients. A composite outcome was defined as the most severe condition, including the need for respiratory support by noninvasive positive pressure ventilation (NPPV), high-flow nasal cannula (HFNC), invasive mechanical ventilation (IMV), extracorporeal membrane oxygenation (ECMO), or death. There were three groups: severe (patients requiring support using low-flow oxygen devices), mild (symptomatic patients not requiring oxygen support), and asymptomatic (asymptomatic patients not requiring oxygen support) [5]. Patients who were non-Japanese, treated with chemotherapy, treated with steroids before admission, pregnant, had concomitant or prior hematologic malignancies, had incomplete CBC results, or had composite outcomes were excluded (Supplementary Fig. 1).
To identify predictors, we determined which components of the CBC were the best predictors of composite outcomes. We used the area under the receiver-operating characteristic (ROC) curve (AUC) to combine patient characteristics and symptoms/signs with AUCs of ≥0.6 with the best predictor from the CBC. The cut-off was determined by the Youden index. The DeLong test was used to compare the AUCs.
To investigate the independent association between the NLR and composite outcomes, we performed a multivariable logistic regression analysis after adjustments for known severity factors (age [≥65 years], sex, BMI [≥25 kg/m2], smoking history, hypertension, cardiovascular disease, and chronic kidney disease) [6]. Subgroup (severity subgroups) multivariable logistic regression analyses using an NLR cut-off were also performed. Since the database did not distinguish between asymptomatic and mild disease on admission, we compared the asymptomatic/mild and the severe groups. The most severe group on admission was excluded from analysis since they had composite outcomes. Statistical analyses were conducted using R (The R Foundation for Statistical Computing, Vienna, Austria) and JMP version 16 (SAS Institute Japan Ltd, Tokyo, Japan). The figures were plotted using R and GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA).
3. Results
The baseline characteristics of 2425 patients included in the study are summarized in Table 1 . The median age was 57 years, and 1678 (69.2%) patients were male. Of the 2425 patients, 330 patients experienced IMV (13.6%), 118 (4.9%) HFNC, 88 (3.6%) death, 48 (2.0%) ECMO, and 4 (0.2%) NPPV. Table 2 shows the results of the ROC analysis of predictors of the composite outcomes. NLR was higher in patients with worse severity (Fig. 1A). NLR was the best predictor of composite outcomes, with an AUC of 0.81, a sensitivity of 72.3%, and a specificity of 75.5%, using a cut-off of 5.04. We combined age, respiratory rate, dyspnea, and an oxygen requirement on admission with the NLR (Fig. 1B ). The highest AUC was 0.88 for NLR combined with an oxygen requirement on admission. Using each cut-off, a composite outcome was observed in 44 patients (3.3%) with an NLR below the cut-off and no oxygen requirement; 114 patients (19.1%) with an NLR above the cut-off without an oxygen requirement or below the cut-off and with an oxygen requirement; and 314 patients (63.2%) with an NLR above the cut-off and an oxygen requirement (Fig. 1C). Of the patients with composite outcomes, 37.1% of those without an oxygen requirement had an NLR above the cut-off, and 66.7% of those with an NLR below the cut-off required oxygenation.
Table 1.
Clinical characteristics of patients with COVID-19.
| Clinical characteristics | Total (n = 2425) |
|---|---|
| Age, years | 57 (45–71) |
| Sex, male | 1678 (69.2) |
| BMI, kg/m2 | 24.4 (21.9–27.4) |
| Smoking history | |
| Non-smoker | 1179 (52.4) |
| Past smoker | 718 (31.9) |
| Current smoker | 352 (15.6) |
| Days from onset | 6 (3–8) |
| Oxygen requirement on admission | 779 (32.1) |
| Vital signs | |
| Body temperature, °C | 37.1 (36.6–38.0) |
| Systolic blood pressure, mmHg | 127 (114–140) |
| Diastolic blood pressure, mmHg | 80 (71–89) |
| Heart rate,/min | 86 (76–98) |
| Respiratory rate,/min | 19 (16–22) |
| Symptoms and signs | |
| Coma | 92 (3.9) |
| Fever | 2968 (82.1) |
| Cough | 1506 (63.7) |
| Sputum | 685 (29.2) |
| Sore throat | 593 (25.4) |
| Nasal discharge | 360 (15.4) |
| Taste disorder | 444 (19.0) |
| Smell disorder | 364 (15.6) |
| Dyspnea | 919 (39.3) |
| Stomach ache | 86 (3.7) |
| Abdominal distension | 19 (0.8) |
| Hematochezia | 13 (0.6) |
| Diarrhea | 456 (19.5) |
| Nausea or vomiting | 226 (9.7) |
| Fatigue | 1302 (55.2) |
| WBC,/μL | 5100 (4000–6800) |
| Neutrophils,/μL | 3577 (2581–5188) |
| Neutrophil, % | 71.7 (62.7–80.5) |
| Lymphocytes,/μL | 969 (695–1320) |
| Lymphocyte, % | 19.8 (12.6–27.3) |
| Eosinophils,/μL | 9 (0–40) |
| Eosinophil, % | 0.2 (0.0–0.8) |
| Hb, g/dL | 14.4 (13.1–15.5) |
| Plt ( × 104),/μL | 18.7 (14.9–23.4) |
| NLR | 3.63 (2.30–6.40) |
| ELR | 0.03 (0.01–0.07) |
| CRP | 3.27 (0.82–8.03) |
| Treatment | |
| Steroids | 1286 (53.5) |
| Remdesivir | 923 (38.5) |
| Anti-IL-6 antibody | 273 (11.4) |
| Severity | |
| Asymptomatic | 60 (2.5) |
| Mild | 1099 (45.3) |
| Severe | 794 (32.7) |
| Critical | 472 (19.5) |
| Composite outcome | |
| Death | 88 (3.6) |
| NPPV | 4 (0.2) |
| NHFC | 118 (4.9) |
| IMV | 330 (13.6) |
| ECMO | 48 (2.0) |
Data are shown as number of patients (%) or median (interquartile range).
Abbreviations: BMI, body mass index; COVID-19, coronavirus disease; CRP, C-reactive protein; ECMO, extracorporeal membrane oxygenation; ELR, eosinophil-to-lymphocyte ratio; Hb, hemoglobin; IL-6, interleukin-6; IMV, invasive mechanical ventilation; NHFC, high-flow nasal cannula; NLR, neutrophil-to-lymphocyte ratio; NPPV, non-invasive positive pressure ventilation; Plt, platelet; WBC, white blood cell.
Table 2.
Area under the receiver operating curve, 95% confidence interval, sensitivity and specificity of CBC, age, sex, BMI, vital signs, oxygen requirement on admission, and symptoms and signs.
| AUC | 95% CI | Sensitivity | Specificity | Cut-off | |
|---|---|---|---|---|---|
| WBC | 0.68 | 0.65–0.71 | 51.7 | 80.9 | 6740 |
| Neutrophil | 0.73 | 0.71–0.76 | 55.1 | 83.8 | 5326 |
| Lymphocyte | 0.74 | 0.71–0.76 | 69.3 | 67.1 | 864 |
| Eosinophil | 0.72 | 0.69–0.74 | 79.2 | 62.2 | 5 |
| Hb | 0.59 | 0.56–0.62 | 63.4 | 50.7 | 14.4 |
| Plt | 0.52 | 0.49–0.55 | 32.6 | 73.5 | 15.2 × 104 |
| NLR | 0.81 | 0.78–0.83 | 72.3 | 75.7 | 5.04 |
| ELR | 0.50 | 0.44–0.56 | 19.8 | 86.7 | 0.11 |
| CRP | 0.76 | 0.74–0.79 | 79.0 | 60.6 | 3.68 |
| Age, years | 0.64 | 0.61–0.66 | 72.3 | 53.1 | 57 |
| Sex, male | 0.55 | 0.52–0.57 | 76.7 | 32.6 | |
| BMI, kg/m2 | 0.58 | 0.55–0.61 | 62.1 | 51.9 | 24.3 |
| Vital signs | |||||
| Body temperature, °C | 0.51 | 0.48–0.54 | 29.8 | 75.2 | 38.0 |
| Systolic blood pressure, mmHg | 0.51 | 0.48–0.54 | 36.4 | 70.4 | 137 |
| Diastolic blood pressure, mmHg | 0.56 | 0.53–0.59 | 32.9 | 76.7 | 71 |
| Heart rate,/min | 0.50 | 0.47–0.53 | 30.1 | 73.8 | 97 |
| Respiratory rate,/min | 0.67 | 0.64–0.70 | 52.5 | 74.3 | 21 |
| Oxygen requirement on admission | 0.83 | 0.81–0.85 | 85.2 | 80.7 | |
| Symptoms and signs | |||||
| Coma | 0.55 | 0.81–0.85 | 12.3 | 98.0 | |
| Fever | 0.50 | 0.48–0.52 | 82.4 | 17.9 | |
| Cough | 0.52 | 0.50–0.52 | 67.7 | 37.2 | |
| Sputum | 0.54 | 0.51–0.54 | 35.0 | 72.1 | |
| Sore throat | 0.45 | 0.43–0.47 | 83.0 | 27.2 | |
| Nasal discharge | 0.46 | 0.45–0.48 | 91.0 | 16.8 | |
| Taste disorder | 0.45 | 0.44–0.47 | 88.8 | 20.3 | |
| Smell disorder | 0.44 | 0.42–0.45 | 94.7 | 17.8 | |
| Dyspnea | 0.71 | 0.69–0.71 | 74.1 | 68.7 | |
| Stomachache | 0.51 | 0.50–0.52 | 4.8 | 96.6 | |
| Abdominal distension | 0.50 | 0.50–0.51 | 1.2 | 99.3 | |
| Hematochezia | 0.51 | 0.50–0.51 | 1.7 | 99.7 | |
| Diarrhea | 0.47 | 0.46–0.49 | 84.7 | 20.4 | |
| Nausea or vomiting | 0.50 | 0.48–0.51 | 91.1 | 10.0 | |
| Fatigue | 0.59 | 0.56–0.61 | 69.2 | 48.0 | |
Abbreviations: AUC, area under the receiver operating curve; BMI, body mass index; CI, confidence interval; CBC, complete blood count; CRP, C-reactive protein; ELR, eosinophil-to-lymphocyte ratio; Hb, hemoglobin; NLR, neutrophil-to-lymphocyte ratio; Plt, platelet; WBC, white blood cell.
Fig. 1.
NLR by severity, receiver operating characteristic curve of NLR + age/dyspnea/RR/oxygen, and the proportion of composite outcomes. (A) Boxplot of Neutrophil-to-lymphocyte ratio (NLR) compared by severity. (B) Receiver operating characteristic curves (ROCs) and area under the receiver operating curve (AUC) of NLR + age, dyspnea, respiratory rate (RR), and oxygen requirement at referral (Oxygen). (C) Proportion of composite outcomes. Number of composite outcomes obtained using a combination of NLR >5.04 and oxygen requirement at referral. Composite outcomes were observed in 44 (3.3%) patients in the negative group, 114 (19.1%) in either positive group, and 314 (63.2%) in both positive groups.
The adjusted multivariable analysis revealed that NLR was a significant predictor of composite outcomes (adjusted odds ratio [aOR], 1.07; 95% confidence interval [CI], 1.06–1.09) (Supplementary Fig. 2). Subgroup multivariable analyses using severity on admission also showed that NLR was significantly associated with composite outcomes in both the asymptomatic/mild (aOR, 2.31; 95% CI, 1.39–3.83) and severe groups (aOR, 3.37; 95% CI, 2.27–5.03) (Supplementary Table 1).
4. Discussion
We analyzed a large number of Japanese patients and confirmed previously reported factors. Further, NLR using a cut-off of 5.04, was useful in predicting composite outcomes based on the severity of COVID-19 on admission. A previous meta-analysis showed that patients with severe COVID-19 on admission and non-survivors had higher NLR levels than those of patients with less severe disease and survivors, respectively [7]. Another study showed that corticosteroids reduced mortality in patients with an NLR >6.11, but not in those with an NLR ≤6.11 [8]. Therefore, NLR may be a predictor of treatment effectiveness.
In this study, an oxygen requirement on admission combined with NLR had the highest AUC. In another study, the mortality rate of patients with COVID-19 who required mechanical ventilation or had a fraction of inspired oxygen (FiO2) of ≥60% was 61.5% [9], suggesting that the need for oxygenation may be a predictor for severe disease. Among patients with COVID-19, those who require oxygenation and tracheal intubation have higher mortality rates than those who do not [10]. Among patients who receive IMV, a lower FiO2 ratio is associated with a better prognosis [11]. The use of NLR and oxygen demand as prognostic indicators is useful in primary care settings because they are cheap and easily assessable [12]. Our results showed that the combination of a high NLR with an oxygen requirement has a high predictive value. Therefore, these patients should ideally be monitored in hospitals and not care facilities. Notably, the predictive value of NLR alone was low in patients without an oxygen requirement, suggesting that we should carefully evaluate other risk factors in these patients.
There are several hypotheses regarding the mechanisms behind high NLR in patients with severe COVID-19: hemolysis of infected lymphocytes through angiotensin converting enzyme 2 receptors expressed on their surface [13], and lymphocyte apoptosis in the presence of interleukin-6 and tumor necrosis factor-alpha [14]; however, the actual mechanism remains unclear. The exact reason for the increase in diagnostic accuracy when combining NLR and an oxygen requirement is unclear. Conditions requiring oxygen are caused by inflammation and other pathophysiological states such as underlying diseases. Therefore, combination of NLR with an oxygen requirement can work complementarily.
This study had several limitations. First, this was a retrospective analysis with no prospective validation. Second, since we registered only hospitalized patients aged ≥18 years, this study's population was different from that of the hospital's outpatients department, and the percentages of patients who needed oxygen and had composite outcomes were higher than those in the Japanese surveillance. Further studies are required to confirm these results.
5. Conclusions
The combination of NLR and an oxygen requirement can predict severe outcomes in patients with COVID-19 very accurately. This simple, cost-effective combination may help to predict which patients are most likely to progress to severe disease, thus enabling early intervention.
Funding source
This work was supported by AMED [grant numbers: JP20nk0101612, JP20fk0108415, JP21jk0210034, JP21km0405211, JP21km0405217, JP21wm0325031, and JP21fk0108573], JST CREST [grant number: JPMJCR20H2], JST PRESTO [grant number: JPMJPR21R7], and MHLW [grant number: 20CA2054].
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Conflict of Interest
The authors have no conflicts of interest.
Acknowledgments
We would like to thank all participants who were involved in this study and all members of the Japan COVID-19 Task Force engaged in clinical work and research on COVID-19.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.resinv.2023.03.007.
Appendix A. Supplementary data
The following is the Supplementary data to this article.
References
- 1.Terpos E., Ntanasis-Stathopoulos I., Elalamy I., Kastritis E., Sergentanis T.N., Politou M., et al. Hematological findings and complications of COVID-19. Am J Hematol. 2020;95:834–847. doi: 10.1002/ajh.25829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Liu Y., Du X., Chen J., Jin Y., Peng L., Wang H.H.X., et al. Neutrophil-to-lymphocyte ratio as an independent risk factor for mortality in hospitalized patients with COVID-19. J Infect. 2020;81:e6–e12. doi: 10.1016/j.jinf.2020.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rechtman E., Curtin P., Navarro E., Nirenberg S., Horton M.K. Vital signs assessed in initial clinical encounters predict COVID-19 mortality in an NYC hospital system. Sci Rep. 2020;10 doi: 10.1038/s41598-020-78392-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Namkoong H., Edahiro R., Fukunaga K., Shirai Y., Sonehara Y., Tanaka H., et al. 2021. Japan COVID-19 Task Force: a nation-wide consortium to elucidate host genetics of COVID-19 pandemic in Japan. MedRxiv [Preprint] [DOI] [Google Scholar]
- 5.Tanaka H., Lee H., Morita A., Namkoong H., Chubachi S., Kabata H., et al. Int J Infect Dis. 2021;113:74–81. doi: 10.1016/j.ijid.2021.09.070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fukushima T., Chubachi S., Namkoong H., Asakura T., Tanaka H., Lee H., et al. Clinical significance of prediabetes, undiagnosed diabetes and diagnosed diabetes on critical outcomes in COVID-19: integrative analysis from the Japan COVID-19 task force. Diabetes Obes Metabol. 2023;25:144–155. doi: 10.1111/dom.14857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Simadibrata D.M., Calvin J., Wijaya A.D., Ibrahim N.A.A. Neutrophil-to-lymphocyte ratio on admission to predict the severity and mortality of COVID-19 patients: a meta-analysis. Am J Emerg Med. 2021;42:60–69. doi: 10.1101/2020.09.14.20191098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cai J., Li H., Zhang C., Chen Z., Liu H., Lei F., et al. The neutrophil-to-lymphocyte ratio determines clinical efficacy of corticosteroid therapy in patients with COVID-19. Cell Metabol. 2021;33:258–269. doi: 10.1016/j.cmet.2021.01.002. e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yang X., Yu Y., Xu J., Shu H., Xia J., Liu H., et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med. 2020;8:475–481. doi: 10.1016/s2213-2600(20)30079-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Rojas-Marte G., Hashmi A.T., Khalid M., Chukwuka N., Fogel J., Munoz-Martinez A., et al. Outcomes in patients with COVID-19 disease and high oxygen requirements. J Clin Med Res. 2021;13:26–37. doi: 10.14740/jocmr4405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Grasselli G., Greco M., Zanella A., Albano G., Antonelli M., Bellani G., et al. Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy. JAMA Intern Med. 2020;180:1345–1355. doi: 10.1001/jamainternmed.2020.3539. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kerboua K.E. NLR: a cost-effective nomogram to guide therapeutic interventions in COVID-19. Immunol Invest. 2021;50:92–100. doi: 10.1080/08820139.2020.1773850. [DOI] [PubMed] [Google Scholar]
- 13.Xu H., Zhong L., Deng J., Peng J., Dan H., Zeng X., et al. High expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa. Int J Oral Sci. 2020;12:8. doi: 10.1038/s41368-020-0074-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liao Y.C., Liang W.G., Chen F.W., Hsu J.H., Yang J.J., Chang M.S. IL-19 induces production of IL-6 and TNF-alpha and results in cell apoptosis through TNF-alpha. J Immunol. 2002;169:4288–4297. doi: 10.4049/jimmunol.169.8.4288. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.