3.5 Longitudinal variations of laboratory findings
As disease progressed, differences in longitudinal trends of the laboratory findings were observed between the non-survived patients and survived severe patients. As shown in Fig. 5, leucocyte and neutrophil counts increased in the early stage of hospitalization (3-7 days) and gradually decreased in the late stage of hospitalization (8-14 days) in severe survived patients, but continuously increased in non-survived patients. Even though lymphopenia was observed in both groups during hospitalization, the lymphocyte count was significantly lower in the non-survived group compared to survived severe group. Also, it is noticeable that sustained eosinopenia and progressing thrombocytopenia were seen in non-survived patients, but both blood cell numbers partially relieved in survived severe patients. Sustained high levels of NLR, CRP, PCT, AST, BUN and serum creatinine were associated with the fatal clinical outcome of severe patients.
Discussion
This retrospective cohort study reported the clinical, radiological and laboratory characteristics of 289 hospitalized COVID-19 patients in Wuhan, China, and identified several risk factors for mortality of COVID-19. Of the 289 laboratory-confirmed COVID-19 cases in this study, most of the patients [183 (63.3%)] were more than 50 years old. 53.3% of the patients were male. 58.5% of the patients were associated with underlying comorbidities, including hypertension (28%), diabetes mellitus (9.3%) and coronary heart disease (6.2%), which were similarly demonstrated in previous studies.16-19
The mortality of the 289 hospitalized cases in the present study was 17.0% (49/289), which was lower than that reported by Zhou et al. [54/191 (28.3%)],18 but much higher than that reported by Guan et al. (1.4%)19 and that in a large-scale analysis reported by the Chinese Center for Disease Control and Prevention [1023/44672 (2.3%)].20 The difference may due to the distinct sample sizes and case inclusion criteria of studies. 56.1% (162/289) patients were non-severe and ultimately discharged from the hospitals.
In this study, elder age, more underlying comorbidities (including hypertension), surgery history, higher prevalence of abnormal CT images as well as the increased leucocyte and neutrophil counts, elevated levels of serum CRP, PCT, D-dimer, ALT, AST and BUN on admission were found in survived severe cases, compared to non-severe cases. Significant differences were identified between non-survived patients and survived severe patients, regarding age, clinical symptoms of chest tightness/dyspnea and loss of appetite, neutrophil counts, NLR, and levels of serum CRP, PCT and D-dimer on admission.
It is interesting to find that affected lobe numbers in CT-scans were associated with the severity of the disease. This was not reported previously. In addition, affected lobe numbers have correlated with age, CRP, D-dimer and BUN, which correlated with each other. Continuous surveillance of chest CT is apparently useful for the evaluation of the disease course of COVID-19. However, in critically ill patients, it is not easy to obtain the second chest CT scan during hospitalization, especially in intubated and ventilated patients. In this situation, bedside X-ray will be an alternative. Most of the non-survived patients of this study lacked the second CT scan.
The risk factors of mortality identified in this study include elder age, higher level of CRP, more number of affected lobe(s), chest tightness/dyspnea and smoking history, according to the logistic regression model, which indicate that these parameters can be used for the prediction of the risk of death in severe patients. Although neutrophil counts and biochemical parameters (including D-dimer) were significantly different between the survived severe and non-survived patients in this study, these variables were not found to be an independent risk factor for the mortality of COVID-19 patients.
A previous study has demonstrated that elder age and high level of D-dimer were associated with poor prognosis of hospitalized COVID-19 patients.18 The present study confirmed that increased age was associated with higher risk of death in COVID-19 patients, this may be caused by the dampened immune function and more underlying comorbidities, which potentially lead to the poor outcome of elder patients.21
Several studies have reported that hypertension, hypoxia, leukocytosis, lymphopenia and high serum LDH level were independent predictors for in-hospital death.7-10 However, in the present study, only dyspnea and leukocytosis, but not other variables, were found to be independent risk predictors of death in critically ill COVID-19 patients, which is different to previous reports. In addition, previous reports identified that lower baseline levels and/or progressively decreasing platelet counts were associated with higher mortality of COVID-19 patients.11 Although no significant difference in baseline platelet levels was identified between the severe survived and non-survived patients, longitudinal in-hospital follow up data showed progressing thrombocytopenia during hospitalization was more significant in the non-survived patients in the current study.
CRP is a clinically wide used inflammatory marker. High levels of CRP indicate inflammation caused by various conditions including infections. Elevated IL-6, which is the trigger of CRP synthesis in the liver, was also observed in COVID-19 patients. 22 Cytokine storm has been suggested to be an important cause for poor prognosis of critically ill COVID-19 patients. In the current study, higher levels of CRP were found to be associated with the poor clinical outcome of severe patients. Wang et al. identified higher levels of CRP in non-survivors compared with the survivors within 15 days of COVID-19 hospitalization.10 Additionally, Wu et al. found that elevated high-sensitivity C-reactive protein (hs-CRP) was significantly associated with higher risks of acute respiratory distress syndrome (ARDS) in COVID-19 patients.7 These data suggested that CRP was one of the indicators of cytokine storm in COVID-19 and associated with the mortality of this disease.
Higher NLR has been suggested to be an independent risk factor of mortality in hospitalized COVID-19 patients in a recent study. In the same study, the levels of serum procalcitonin were positively correlated with the rate of in-hospital death. Although the univariate analysis found that the odd of in-hospital death was higher in patients with higher NLR, it was not an independent risk factor of death for severe patients in the current study.23
Interestingly, the prevalence of patients with smoking history is 9.7% in the current study which is lower than that reported by Guan. et al. (current smoker, 12.6%) ,19 but significantly higher than that reported in a previous study by our team (smokers, 6.4%).17 We think this difference may reflect the difference in the population of patients and may also be caused by the insufficient data collection of the smoking history, when in consideration of the infectious nature of COVID-19. However, the presence of smoking history was found to be a risk factor of death in critically ill patients, indicating that smoking may be associated with mortality in patients with severe COVID-19. This finding is consistent with previous study by Mehra et al., which concluded that current smokers had higher in-hospital death in COVID-19 patients.24 However, the limit of number of smokers in the present study may limit the statistical power of the conclusion. Angiotensin‐converting enzyme 2 (ACE2) was demonstrated as a receptor for SARS-CoV-2, which was highly expressed in airway epithelial cells and plays an important role in SARS‐CoV‐2 infection.25The higher levels of expression of ACE2 in the lower respiratory tract of current smokers may contribute to the risk of developing severe COVID-19 infection.26-27
In concert with previous reports,16,28 bilateral lung involvement was predominant in patients with abnormal chest CT images, mainly manifested as multiple ground glass opacities and subpleural lesions. In addition, five pulmonary lobes were affected in more than half of the patients with abnormal chest CT images. The strong positive correlations identified between the affected lobe numbers and patients’ age, neutrophil counts and lymphocyte counts indicate that more severe pneumonia was associated with the elderly age and higher degree of lymphopenia, which supports that the affected lobe number(s) could be a possible risk factor for severe cases and the in-hospital mortality of severe COVID-19 patients. Leukocytosis, eosinopenia and lymphopenia may be associated with the progression of inflammatory status, and more severe illness was associated with older patients, given the increased level of CRP, SAA, PCT and D-dimer.
Comparison of the dynamic profile of laboratory findings in both groups revealed that sustained increase in leucocyte count, neutrophil count, biological markers, as well as progressing decrease in platelet count, lymphopenia and eosinopenia may be the predictor of death during hospitalization. A previous study has reported an inverse correlation between the serum D-dimer level and the duration of antiviral treatment, which indicates the decreased level of serum D-dimer may represent the status of virus clearance.29 However, the similar dynamic trend in levels of D-dimer was not observed in the present series of patients during hospitalization, this may due to the limited number of patients with complete laboratory result of biochemical parameters throughout the period of hospitalization.
This study was limited to the relatively small number of patients which may limit the statistic power and the inclusion of hospitalized patients exclusively (non-hospitalized patients were not included in the analysis), these may cause statistical bias and hence the significant difference identified for demographic and symptomatic characteristics, as well as the laboratory findings between the groups. Also, missing data on some variables, such as information of CT images and biochemical parameters, may cause bias in the identification of risk factors for mortality in severe patients.
In summary, this retrospective, bi-center study revealed that elder age, level of CRP, the number of affected pulmonary lobes, the clinical symptoms manifested with chest tightness/dyspnea and the presence of a smoking history were the independent risk factors of mortality for the non-survived patients, in comparison of severe and survived patients. Assessment of these parameters may help to identify severe COVID-19 patients in high risk of death. Earlier medical intervention and support on these patients with high risk may reduce the fatality of this disease.