Clinical Characteristics of COVID-19 and the Value of Mulbsta Scoring System in Prognosis Evaluation

Background COVID-19, a worldwideinfectious disease, has killed more than 420000 people, whichis extremely harmful. Methods In this single-center retrospective study, we included the novel coronavirus pneumonia conrmed in our hospital. This study collected the basic information andclinical examination features. Results The enrolled 704 patients were armed infected with 2019-nCoV by the test of throat swabs. There are 334 men and 369 women, and gender, age, combined with basic diseasesare distinct in diverse disease classication (p<0.05). From the symptom analysis, the proportion of fever over 38 degrees, dyspnea, fatigue, poor appetite and other symptoms is diverse in different types of diseases (p<0.05). As the severity of the disease increases, the median lymphocyte count decreases, C-reactive protein increase, erythrocyte sedimentation rateincrease, albumin decrease, pleural effusion increase, D-Dimer and NT-proBNP increase signicantly (p<0.05).As the disease severity increases, the average value of MuLBSTA score obviously ascend (p<0.05), MuLBSTA scoring system predicts novel coronavirus pneumonia patients' prognosis is still insucient, and may require additional indicators including anorexia, fatigue, C reactive protein, etc (p<0.05). The MuLBSTA evaluation system has certain value for the evaluation of the disease, but it needs to be improved.


Introduction
The novel coronavirus 2 of severe acute respiratory syndrome (SARS-CoV-2), a novel beta-coronavirus, has caused COVID-19 [1,2], which has become a worldwide pandemic disease, with a total number of more than 4.5 million people infected. On March 11, who o cially de ned the pneumonia epidemic as "pandemic" in Geneva, and so far, novel coronavirus pneumonia has resulted in over 420 thousand deaths. The most common clinical symptoms at onset are fever, dry cough, chest distress, dyspnea, fatigue, poor appetite, and loss of smell [3]. Clinically, we found that patients with the same clinical manifestations have great differences in disease changes and prognosis. Owing to the complexity of COVID-19, it is signi cantly important to analyze the clinical characteristics of patients and judge the prognosis of patients in the early stage.

Page 3/16
The MuLBSTA score, an early prediction model of mortality in viral pneumonia [4,5]. Lingxi Guo found that the MuLBSTA score, based on 6 parameters (multilobular in ltration, lymphopenia, bacteria coinfection, smoking history, hypertension, and age), has a powerful prediction 90-day mortality [5]. Nanshan Chen mentioned that further examination is required to evaluate the usability of MuLBSTA score on mortality of COVID-19 [4]. Our study not only axamined the predictive value of the MuLBSTA score for the risk of mortality in novel coronary pneumonia, but also investigated the relationship between the MuLBSTA score and the disease level of the novel coronary pneumonia. What's more, novel coronavirus pneumonia prognosis is evaluated by MuLBSTA score in this article, and we also assessed the value of the MuLBSTA score and the parameters that may be needed to further re ne it.

Research design
For this single-center retrospective study, we included the novel coronavirus pneumonia con rmed in our hospital for the rst time (excluding patients hospitalized outside the hospital) from January to March. MuLBSTA score The MuLBSTA score consists of multilobular in ltration (5 points), lymphopenia ≤ 0.8 (4 points), bacteria co-infection (4 points), smoking history (Acute-smoker 3 points, Quit-smoker 2 points), hypertension (2 points), and age ≥ 60 (2 points). If the patient has the above situation, calculate the score of each item and add it up [5].

Statistical analysis
We expressed continuous variables as the medians and interquartile ranges (IQR) or average. Categorical variables were presented as the counts. N represents the total number of people in each group.

History characteristics
The enrolled 704 patients were a rmed infected with 2019-nCoV by the test of throat swabs. There are 334 men and 369 women, 48% and 52% respectively. More concretely, the number of female patients with light, normal, heavy and critical weight was 8 (2%), 291 (79%), 52 (14%), 18 (5%) respectively. Correspondingly, the number of male patients with light, normal, heavy and critical was 16 (5%), 234 (70%), 60 (18%) and 24 (7%) respectively. Gender is different in disease classi cation (p < 0.05). The research presented the age as median (IQR), and we know that as disease grades get more severe, the median age gets larger from 55.5(43.8-72.2) to 73.5(64.8-80.0). Comorbidities occurred in 60% of the patients, and the most common comorbidity was hypertension, followed by diabetes, coronary heart disease, COPD and cerebrovascular disease. Through analysis, we know that hypertension, chronic renal injury and cerebrovascular disease are different in the proportion of patients with different clinical types (p < 0.05), what's more, patients without comorbidities are in a lighter condition (p < 0.05). In terms of clinical symptoms, the frequency of occurrence was fever, dry cough, fatigue, chest tightness, dyspnea, poor appetite, and diarrhea. In addition, there were two cases of olfactory sense. In clinical symptoms, fever, dry cough, fatigue, chest distress, dyspnea, poor appetite and diarrhea were the most frequent. In addition, there were two cases of olfactory loss. Among them, dry cough, chest distress, dyspnea, asthenia, poor appetite and fever of different degrees are diverse in the frequency of disease classi cation (p < 0.05) ( Table 1).

Clinical examination features
On admission, the groups of leukocyte count below the normal range, above the normal range and in the normal range have distinct frequency in diverse disease levels (p < 0.05) ( Table 2). Neutrophil count below the normal value in 81 patients (12%) and neutrophil count above the normal value in 622 patients (88%), but there is no obvious difference in the frequency of the two in different levels of disease (p > 0.05) ( Table 2). As the severity of the disease increases, the median lymphocyte count decreases signi cantly (p < 0.05), and the lymphocyte count below the normal value to a certain extent indicates a serious condition ( Table 2). Platelet count, hemoglobin, C-reactive protein, and D-Dimer median (IQR) also have signi cant difference in diverse disease levels (p < 0.05) ( Table 2). The groups of blood routine index (ESR, Procalcitonin, AST, Creatinine, NT-proBNP, Troponin I) below the normal range and above the normal range also have signi cant difference in diverse disease levels (p < 0.05) ( Table 2). In addition, similar to lymphocyte results, as the severity of the disease increases, the median albumin decreases signi cantly (p < 0.05). Through more detailed analysis, we know that different albumin content groups (< 25, 25-30, 30-35, ≥ 35 g/L) have distinct frequency in diverse disease levels (p < 0.05) ( Table 2). In CT imaging, nearly 97% of patients have ground glass opacity, and In addition, this study found that the frequency of pleural effusion in diverse disease levels was also different (p < 0.05).   Relationship between MuLBSTA and disease grade As Lingxi Guo previously reported that MuLBSTA < 12 points represents low-risk and MuLBSTA ≥ 12 points means high-risk [5]. We further subdivided the 12 to 22 segments (12 ≤ Mu ≤ 15, 15 < Mu ≤ 18, 18 < Mu ≤ 22), and as the disease severity increases, the average value of MuLBSTA score obviously ascend (p < 0.05). Meanwhile, the research found that the groups of distinct MuLBSTA (< 12, 12-15, 15-18, 18-22) have remarkable difference in diverse disease levels (p < 0.05) ( Table 3).  (Table 4). In biochemical examination, the research noticed that a few of tests median value (C-reactive protein, Albumin, D-Dimer, NT-proBNP) becomes lower and lower as the MuLBSTA score gets higher and higher, which also indicates that the tests (C-reactive protein, Albumin, D-Dimer, NT-proBNP) may be required to improve the MuLBSTA scoring system (Table 4).

Discussion
This is a retrospective study on the clinical features of COVID-19 and the MuLBSTA scoring system, including 703 patients of COVID-19 from Wuhan forth hospital.
Clinically, we found that some patients with COVID-19 have mild or no symptoms, but they have the same infectivity [6]. At the same time, they also have the possibility of transforming into severe cases, which is the di culty of our epidemic prevention and treatment, so it is particularly important to preliminarily determine the patient's condition change through the disease state at the time of admission. Therefore, we studied the most serious level of the disease course and the patient's basic situation, clinical symptoms, biochemical examination, etc., and also evaluate the MuLBSTA evaluation system prediction for the most severe state of the patient's disease course, and the indexes that may need to be improved are also discussed. Studies have shown that COVID-19 has more male patients than female patients [1], and a research reported that gender related ratios of COVID-19 patients were similar [7]. Among the patients with the novel coronary pneumonia we included, men accounted for 48% and women accounted for 52%, and the difference between the two genders was small, but the proportions of the two in diverse disease levels were distinct, speci cally, there are more male critically ill patients. This may be related to the X chromosome and sex hormones, which exert a vital role in innate and adaptive immunity [8], or it may be related to smoking. Our data show that novel coronavirus pneumonia is usually more severe in older patients, which is the same as others [9,10]. Novel coronavirus pneumonia patients with other comorbidity are relatively severe, and thirty-two percent of them with hypertension, what's more, we also found that patients with hypertension are more likely to develop severe and critical diseases [5]. Chronic kidney injury is also the same, and many dead patients have renal failure, meanwhile, acute kidney injury during novel coronavirus pneumonia course indicates poor prognosis and increases mortality [11].
Correspondingly, there was no obviously difference in the proportion of coronary heart disease, diabetes mellitus, COPD and cancer in diverse disease levels. Thus it may be known, we need to be particularly vigilant for patients of novel coronary pneumonia complicated with hypertension and kidney injury. From the analysis of clinical symptoms, fever is commonly found in the patients with severe and critical illness. The proportion of fever in the patients with severe illness is as high as 80%. Meanwhile, dry cough, chest distress, dyspnea, fatigue and poor appetite are also more common in the patients with severe and critical illness.
According to the analysis of the laboratory examination results of patients at admission, there are signi cant differences in many blood examination indexes among patients with diverse grades of diseases (Leukocyte count, Neutrophil count, Lymphocyte count, Platelet count, Hemoglobin, C-reactive protein, ESR, Procalcitonin, Creatinine, Albumin, D-Dimer, D-Dimer, Troponin I). Although there are dramatically differences in some of these indicators in diverse disease grades, the magnitude of the differences is hard to distinguish clinically, and the exceptions are lymphocyte count, C-reactive protein, ESR, Albumin, D-Dimer, NT-proBNP, which results are similar to other research results [12][13][14][15].
Contradictory to the previous research showing that pleural effusion has no speci c type in patients with novel coronary pneumonia [16], the study found that severe patients are more likely to develop pleural effusion. The main reason is that the sample size of the previous study is too small.
The MuLBSTA score consists of multilobular in ltration, lymphopenia ≤ 0.8, bacteria co-infection, smoking history, hypertension, and age ≥ 60, and these data are very easy to obtain clinically. Intuitively, the MuLBSTA score is getting higher and higher with the severity of the disease, in addition, we found that the MuLBSTA score of moderate and severe diseases was more than 12 points, which was similar to the initial study of MuLBSTA [5], We also found that the difference of MuLBSTA score in distinct disease grades was signi cant. The higher the MuLBSTA score, the more serious the disease was, but there were 58 serious patients with MuLBSTA score less than 12 points in the data, which also deserves our attention. According to the analysis, the cause of this phenomenon is the inaccurate record of smoking history in case data, or the congenital de ciency of MuLBSTA scoring system. For the rst reason, we can strengthen the inquiry and record of smoking history, and for the second reason, we need to consider the evaluation indicators that MuLBSTA may need to improve. Next, we make more efforts to this goal in depth analysis.
Considering that the index of MuLBSTA scoring system does not involve clinical symptoms, we try to analyze the possible symptom indexes. We included six symptoms, namely fever, poor appetite, fatigue, chest tightness, dyspnea and dry cough, and we discovered that only poor appetite and fatigue were diverse in MuLBSTA score segment, which represents that poor appetite and fatigue are likely to be selected as symptom indicators. In addition, there may be some indexes that need to be improved in terms of laboratory examination indexes. We know that the MuLBSTA scoring system includes a laboratory examination index-lymphocyte count. Are there any other examination indexes that can be included? We selected some indexes with signi cant distinct in the values of various disease grade (CRP, albumin, D-dimer, pro BNP). Our analysis showed that the indicates-CRP, albumin, D-dimer, NT-proBNP, which are MuLBSTA may need to include. Because of the limited number of cases in our single center study, the above conclusions need to be con rmed by more extensive studies.

Conclusion
For this single-center retrospective study, the analysis displayed that to some extent, we can predict the development trend of patients' diseases from the state of admission. From the perspective of patient classi cation, patients over 60 years old with basic diseases, especially hypertension, chronic renal injury and cerebrovascular disease, are the high-level group of patients with high risk of serious illness. From the symptom analysis-fever over 38 degrees, dyspnea, fatigue, poor appetite-may indicate that the disease will develop into severe or critical illness. According to the results of the examination, lymphopenia, high CRP, high ESR, bacterial infection, low albumin, high D-dimer, high pro BNP and pleural effusion indicate the severity of the disease. From the analysis of MuLBSTA, the higher the MuLBSTA score, the more serious the disease was, and we need to be alert to patients with high MuLBSTA scores. Of course, the evaluation of MuLBSTA for novel coronavirus pneumonia patients are not irreproachable, further, poor appetite, fatigue and some abnormal laboratory test results may be new indicators to be included.