key: cord-0752569-926yxpbf
authors: Shang, Weifeng; Dong, Junwu; Ren, Yali; Tian, Ming; Li, Wei; Hu, Jianwu; Li, Yuanyuan
title: The value of clinical parameters in predicting the severity of COVID‐19
date: 2020-06-02
journal: J Med Virol
DOI: 10.1002/jmv.26031
sha: 89bfdc43925348b07760a0894db129e72fc98bf4
doc_id: 752569
cord_uid: 926yxpbf

To study the relationship between clinical indexes and the severity of coronavirus disease 2019 (COVID‐19), and to explore its role in predicting the severity of COVID‐19. Clinical data of 443 patients with COVID‐19 admitted to our hospital were retrospectively analyzed, which were divided into nonsevere group (n = 304) and severe group (n = 139) according to their condition. Clinical indicators were compared between different groups. The differences in sex, age, the proportion of patients with combined heart disease, leukocyte, neutrophil‐to‐lymphocyte ratio (NLR), neutrophil, lymphocyte, platelet, D‐dimer, C‐reactive protein (CRP), procalcitonin, lactate dehydrogenase, and albumin on admission between the two groups were statistically significant (P < .05). Multivariate logistic regression analysis showed NLR and CRP were independent risk factors for severe COVID‐19. Platelets were independent protective factors for severe COVID‐19. The receiver operating characteristic (ROC) curve analysis demonstrated area under the curve of NLR, platelet, CRP, and combination was 0.737, 0.634, 0.734, and 0.774, respectively. NLR, CRP, and platelets can effectively assess the severity of COVID‐19, among which NLR is the best predictor of severe COVID‐19, and the combination of three clinical indicators can further predict severe COVID‐19.

The general condition of the patients and the results of leukocyte, neutrophils, lymphocytes, neutrophil-to-lymphocyte ratio (NLR), hemoglobin, platelets, D-dimer, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), procalcitonin, lactate dehydrogenase (LDH), uric acid, creatinine, albumin, chest computed tomographic (CT) scans on the first test during the period of hospitalization, and real-time reverse transcriptase polymerase chain reaction results of respiratory viruses were collected from electronic medical records. The diagnosis of COVID-19 was confirmed by RNA detection of the SARS-CoV-2 in the Wuhan Institute of Virology, Chinese Academy of Sciences.

All data were entered twice into a computer using Epidata 3.1 for the double check to ensure accuracy. The counting data were expressed as a proportion, and χ 2 statistic was used to compare between severe group and nonsevere group. For the data of continuous variables, the normality test was first conducted. After the normal distribution test, the only albumin conformed to the normal distribution, the variable was expressed by mean (SD), and t test was used to compare between the two groups. The rest of the continuous variables that did not conform to the normal distribution were represented by the median (interquartile range, IQR), and the rank-sum test was used for clinical parameters comparison between the two groups. Binary logistic regression (step forward likelihood ratio approach into analysis) was used to perform multifactor analysis and calculate OR values and 95% confidence intervals of the risk factors that were chosen on the basis of likely and relevant confounders after univariate analysis. We depicted the receiver operating characteristic 

A total of 443 cases were included in this study, including 220 males and 223 females. The median age was 56.00 years (IQR, 43.25-66.75).

There were 139 cases in the severe group and 304 cases in the nonsevere group. There were statistically significant differences in the distribution of sex, whether patients had heart disease or not between the two groups (P < .05). The differences in age, leukocyte, NLR, neutrophil, lymphocyte, platelet, D-dimer, CRP, procalcitonin, LDH, creatinine, and albumin on admission between the two groups were statistically significant (P < .05). The differences between the other parameters were not statistically significant (P > .05) ( Table 1) . Table 2 , parameters with P < .1 in Table 1 were included in the logistic regression model, that is, sex, age, whether patients had heart disease or not, leukocyte, NLR, neutrophil, lymphocyte, platelet, D-dimer, CRP, procalcitonin, LDH, creatinine, and albumin. And a forward step (likelihood ratio) method was used for binary logistic regression analysis. Results showed that NLR and CRP were independent risk factors for severe COVID-19, and their OR values were 1.222 and 1.017, respectively. In addition, platelet was an independent protective factor for severe COVID-19 with an OR value of 0.995.

The ROC curves were drawn for severe COVID-19 based on the logistic regression analysis model in Table 2 . The curves were shown in Figure 1 . Since the platelet was a protective factor, all patients' platelet values were multiplied by -1 to make its ROC curve above the reference line. The above parameters were all valuable for predicting the severity of COVID-19 (P < .05). The AUC of the logistic regression model was 0.774 (95%CI: 0.722-0.827). The AUC and optimal thresholds of each independent risk or protection factors can be found in Table 3 .

There are six previously identified coronaviruses that infect humans, four of which are more common and less pathogenic in humans. The other two are known as SARS-Cov-1 and MERS-Cov, which cause severe respiratory diseases. The mortality of SARS has been reported as more than 10% and MERS at more than 35%. 8, 9 The SARS-CoV-2, known as the seventh human coronavirus, belongs to β coronavirus.

The case fatality rate of hospitalized COVID-19 patients has been reported to be 4.3% to 15%. 6,10-12 Therefore, early identification and timely treatment of severe cases are vitally important.

In terms of laboratory tests, the main characteristics of COVID-19

were normal or decreased total number of white blood cells, decreased lymphocyte count, increased CRP, increased ESR, and increased liver enzymes and myoenzymes in some patients. As a new inflammatory indicator, NLR changes can not only reflect the role of neutrophils in infection but also reflect the changes of lymphocytes in vivo. It has been reported that NLR has some predictive value in the diagnosis of disease and severity in patients with influenza virus and other inflammation-related diseases. [13] [14] [15] [16] Our results showed that the NLR in the severe group was significantly higher than that in the nonsevere group. The area under the ROC curve of the NLR predicting the severity of COVID-19 was the largest. The optimal working point was 4.283, and the sensitivity and specificity to predict the severity of COVID-19 were 56.3% and 83.7%, respectively. CRP is a useful inflammatory marker and indicator that plays an important role in host resistance to invading pathogens and inflammation. 18 CRP was highly correlated to the acute lung injury in 2019-nCoV-infected patients. 19 Our study showed that CRP in the severe group was significantly higher than that in the nonsevere group, and CRP was an independent risk factor for severe COVID-19.

The optimal working point was 38.55 mg/L. This is consistent with previous research showing that hypoalbuminemia, lymphopenia, and CRP more than equal to 40 mg/L were the predictive factors for pneumonia progression to respiratory failure in MERS-CoV-infected patients. 20 Besides, higher CRP has been linked to unfavorable aspects of COVID-19 diseases, such as cardiac injury, and ARDS development, and death. 21 Therefore, the detection of CRP levels in COVID-19

patients is of great value in assessing the severity of their condition.

Our study also found that platelet count was lower in the severe group than in the nonsevere group. Binary logistic regression analysis showed platelet count was an independent protective factor for severe COVID-19. The optimal working point was 177.5 (10 9 /L). This partially matches the results of Georges et al 22 who conducted a multicenter retrospective study showing that severe thrombocytopenia was an independent predictor of mortality for severe community-acquired pneumonia. But Elmaraghy's study showed that both thrombocytopenia and thrombocytopenia were significantly associated with mortality in patients with community-acquired pneumonia. 23 Therefore, whether platelets are protective factors for severe COVID-19 needs to be further verified.

This study has several limitations. First, our study was a singlecenter retrospective study, which may affect the generalization of the results due to the limitation of enrolled patients. Second, some patients are still hospitalized among the 443 cases. Therefore, it is difficult to assess risk factors for poor outcomes. Third, the sample size was relatively small, which may have some impact on the statistical results.

In conclusion, NLR, CRP, and platelets can effectively assess the severity of COVID-19, among which NLR is the best predictor of severe COVID-19, and the combination of three clinical indicators can further predict severe COVID-19. To date, there is no specific antiviral drug for COVID-19, and the case fatality rate is still high.

Therefore, clinicians need to carry out a timely and accurate F I G U R E 1 ROC curves of NLR, platelet, CRP, and logistic regression model in patients with severe COVID-19. CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio; ROC, reactive oxygen curve T A B L E 3 The AUC and optimal thresholds of each independent risk or protection factors

A novel coronavirus from patients with pneumonia in China

A new coronavirus associated with human respiratory disease in China

Understanding of COVID-19 based on current evidence

Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72314 cases from the Chinese Center for Disease Control and Prevention

Update on novel coronavirus infected pneumonia situation as of 24:00 on

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

New coronavirus pneumonia prevention and control program

From SARS to MERS, thrusting coronaviruses into the spotlight

SARS and other coronaviruses as causes of pneumonia

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China

Clinical characteristics of novel coronavirus cases in tertiary hospitals in Hubei Province

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Role of hematological parameters in the diagnosis of influenza virus infection in patients with respiratory tract infection symptoms

Neutrophil-to-lymphocyte ratio and platelet-to-lymphocyte ratio: novel markers for the diagnosis and prognosis in patients with restenosis following CAS

Neutrophil-lymphocyte ratio as an early new marker in AIV-H7N9-infected patients: a retrospective study

Comparison of diagnostic values of procalcitonin, C-reactive protein and blood neutrophil/lymphocyte ratio levels in predicting bacterial infection in hospitalized patients with acute exacerbations of COPD

Neutrophil-to-lymphocyte ratio and lymphocyteto-C-reactive protein ratio in patients with severe coronavirus disease 2019 (COVID-19): a meta-analysis

C-reactive protein and inflammation: conformational changes affect function

Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury

Predictive factors for pneumonia development and progression to respiratory failure in MERS-CoV infected patients

Hematological findings and complications of COVID-19

Thrombocytosis in patients with severe community-acquired pneumonia

Platelet count: is it a possible marker for severity and outcome of community acquired pneumonia?

The value of clinical parameters in predicting the severity of COVID-19