Clinical Features and Outcome Analysis of Patients Infected with Severe and Critical COVID-19 Associated Pneumonia

Until now, information on the clinical characteristics of severe and critical patients with COVID-19 is extremely limited.The aim of the present study was to analyse the clinical features of these patients and influencing factors of clinical outcome, and explore treatment effects of prone position on COVID-19 patients with severe ARDS. A retrospective analysis was performed on 55 COVID-19 patients in the ICU of Zhongnan Hospital of Wuhan University from January 6 to February 15, 2020 in Wuhan, China. Case data from each patient were collected and related clinical outcomes on day 14 of ICU admission were recorded. The follow-up deadline was February 29, 2020.


3
The majority of patients with severe and critical COVID-19 in the ICU were over 50 years old and male. 52.7% need invasive mechanical ventilation. On the 14th day of admission, 56.4% of the patients improved, 34.5% of the patients deteriorated. The rate of deaths during hospitalization was 21.8%. The worsening of COVID-19 patients might be related to excessive inflammatory and immune responses. In addition, prone ventilation may improve oxygenation in some COVID-19 patients with severe ARDS, but a significant mortality benefit with proning was uncertain.

Background
In December 2019, multiple unexplained pneumonia cases occurred in Wuhan, China, and laboratory tests later confirmed that the pathogen was a new coronavirus (SARS-CoV-2) with significantly different genetic characteristics than SARS-CoV and MERS-CoV. On February 11, 2020, the World Health Organization officially named SARS-CoV-2 infected pneumonia as COVID-19 (the 2019 novel coronavirus disease). COVID-19 is mainly manifested by fever and respiratory symptoms, and is transmitted through droplets and contact. COVID-19 is extremely contagious. It has quickly spreaded to the entire Hubei province and affected major provinces and cities in China and even whole world since the first patient appeared in December 2019. The latest epidemiological data analysis of 8866 COVID-19 patients in China [1] showed that the basic reproduction number(R0) of COVID-19 was 3.77, which was equivalent to severe acute respiratory syndrome (SARS, R0 was 2.9 to 3.3), and much higher than Middle East Respiratory Syndrome (MERS). As of 24:00 on March 4, 2020, there were already 80422 confirmed case and 2984 death in China.
According to the clinical classification criteria in the fifth edition of the new Coronavirus diagnosis and treatment program [2] issued by the National Health Commission of China, most patients with COVID-19 are light or common, and could recover after bed rest and 4 enhanced supportive treatment. Some severe or critical patients might develop organ dysfunction, such as ARDS, septic shock, acute kidney injury, acute myocardial injury and so on, or even die. A study by Wang D et al. [3] showed that about 26% of hospitalized COVID-19 patients need to enter the ICU for respiratory and organ function supportive care. Therefore, active and effective treatment of severe and critical COVID-19 patients is the key to reduce the mortality of COVID-19. However, patients in most current clinical studies on COVID-19 [3][4][5][6]

Research design and patients inclusion
This study was a retrospective single-center case study, and enrolled all patients diagnosed with COVID-19 who were admitted to the ICU isolation ward in Zhongnan Hospital of Wuhan University from January 6 to February 15, 2020. The diagnostic criteria of COVID-19 referred to the fifth edition of the new coronavirus diagnosis and treatment program issued by the Chinese National Health Commission [2] . SARS-CoV-2 nucleic acid detection in throat swab specimens of enrolled patients was positive by RT-PCR. This study was approved by the Ethics Committee of Zhongnan Hospital of Wuhan University (No.2020045). All patients had signed informed consent when they entered the ICU. Each patient was monitored for clinical outcomes (improvement, stalemate, or deterioration) on the 14th day of ICU admission. The follow-up deadline was February 29, 2020.

Data collection
Through the electronic case system, epidemiological, demographic, clinical and radiological characteristics, laboratory tests, treatments, and clinical outcomes of the enrolled patients were reviewed and recorded. If the first symptoms and contact history could not be obtained from the case system, communications with the patient and family members in person or on the phone were needed. The onset date was the date on which the first symptoms appeared. The time from onset to tracheal intubation, onset to ICU, tracheal intubation, and ICU hospital stay were recorded. The peak body temperature was the highest body temperature throughout the course of the disease. While SARS-CoV-2 nucleic acid in throat swabs was tested in every patient, other respiratory viruses, such as respiratory syncytial virus, influenza A virus, influenza B virus, H7 avian influenza virus and so on, were also analyzed. Save sputum samples for bacterial culture or fungal testing. The recorded laboratory tests included blood routine tests, procalcitonin, liver function, kidney function, electrolytes, coagulation function, and respiratory pathogenic tests, all of which were the results of the first blood draw after ICU admission. Blood samples should be taken before glucocorticoids were used. Prone position therapy was performed for all patients whose PaO 2 / FiO 2 was still less than 100mmHg after invasive mechanical ventilation. The prone position duration and blood gas analysis results within 4-6 hours before and after each prone position session were recorded. All data was reviewed by two physicians (Ying Feng and Mengwei Li).
The clinical classification criteria for the severity of COVID-19 patients referred to the fifth edition of the new coronavirus diagnosis and treatment program issued by the Chinese National Health Commission [2] . The clinical outcome on the 14th day after entering the 6 ICU referred to improvement, stalemate, ordeterioration. The improvement of condition was defined as the clinical classification changed from severe type when entering the ICU to ordinary type, or from critical to severe or common type; stalemate was defined as no change in clinical classification when compared with which when entering ICU; deterioration was defined as clinical classification from critical type to death, or severe type to critical type or death. Organ damage of every patient was recorded, including ARDS, septic shock, acute kidney injury, and acute myocardial injury. ARDS diagnosis and classification referred to the Berlin definition [7] . The diagnosis of septic shock was based on the "2016 Sepsis Rescue Campaign" guidelines [8]. The diagnosis of acute kidney injury was based on the Guidelines for Kidney Disease Improving Global Outcomes (KDIGO) in 2012 [9] . Acute myocardial injury was defined as the blood concentrations of myocardial injury markers, such as high-sensitivity troponin I, exceeding the 99th percentile reference upper limit or new abnormalities of ECG and echocardiography were displayed [4] .Referring to the study of Sahoo JN et al. [10] , the improvement of PaO 2 / FiO 2 after prone postion treatment was defined as at least 20% increase of PaO 2 / FiO 2 compared with which before prone position; PaCO 2 improvement was defined as at least 6mmHg decrease of PaCO 2 in arterial blood after prone postion compared with which before prone position.

The role of funding sources
The fund sponsor did not participate in the project design, data collection, data analysis, and thesis writing of this research. To guarantee the accuracy and integrity of the research, corresponding author was responsible for quality control and review. The final version of the submission was determined by all authors.

Data analysis
Data analysis was performed using SPSS 20.0 statistical software. Measurement data that was normally distributed was expressed as mean(standard deviation) [X (SD)],], and comparison between groups was performed using independent sample t test.
Measurement data that was not normally distributed was expressed as median (interquartile range) [M(IQR)],], and group comparisons were made using Mann-Whitney U test. Count data was expressed in N ( %),, and comparisons between groups were performed using the X 2 test or Fisher's exact probability method. The difference was statistically significant at P <0.05.

Demographic and epidemiological characteristics of patients
Among the 55 patients enrolled, 35 (63.6%)were male and 20 (36.4%) were female. The 8 age ranged from 29 to 92 years old, with an average age of 63.0 (15.2

Comparison of clinical characteristics of patients with improvement and deterioration on the 14th day after ICU admission
On the 14th day after ICU admission, the clinical characteristics of patients who improved and deteriorated indicated that the average age of deteriorated patients was older (P <0.05),and SOFA(Sequential organ failure assessment) and APACHEII(Acute physiology and chronic health evaluation II) scores of deteriorated patients were higher (P <0.05).
Analysis of Lymphocyte subsets showed that CD4, CD8, and NK cell counts in peripheral blood were significantly lower in deteriorated patients (P <0.05). Meanwhile, the concentrations of IFN-γ, IL-10, IL-4, IL-6 and TNF-α in peripheral blood of deteriorated patients were higher than those of patients with improved disease (P <0.05). In addition, patients with worsening condition were more likely to develop acute kidney injury, septic shock, acute myocardial injury, and secondary infections during the disease course (P <0.05). Show in Table 4 indicated that number of male patients were approximately twice than female patients [5] .
Among severe and critical patients in ICU, it was similar that men were far more than women. This gender-related difference in susceptibility may result from the secretion of female estrogen and the expression of immune response genes on the X chromosome [11] .
Studies have confirmed that estrogen has important immunomodulatory effects, which plays an immunosuppressive role at high concentrations and an immune activation role at low concentrations [12,13] . However, it is worth noting that there was no significant  showed that plasma   concentrations of IL2, IL7, IL10, GSCF, IP10, MCP1, MIP1A and TNF-α were significantly higher in ICU patients than that in non-ICU cases [4] . On January 27, 2020, Wang Fusheng, a academician in the Fifth Medical Center of the General Hospital of the Chinese People's Liberation Army, performed pathological examination of a dead patient with COVID-19 through postmortem biopsies. He found lymphocyte-dominated mononuclear cells inflammatory infiltration in the interstitial tissue of both lungs of the patient. Meanwhile, flow cytometry analysis of peripheral blood indicated that number of CD4 + and CD8 + cells was reduced remarkably. However, both were in over-activated statement. And among CD4 + cells, CCR4 + , CCR6 + , and Th17 subsets with high proinflammatory effects increased [14] . Similar results were obtained in our study. A comparative analysis of lymphocyte subsets in patients on the 14th day of ICU showed that the absolute counts of CD4 + , CD8 + , and NK cells in peripheral blood of improved patient were significantly higher than that in worsend patients. At the same time, pro-inflammatory cytokines, such as IFNγ, IL-6, TNF-α, and anti-inflammatory cytokines, such as IL-10, IL-4, in deteriorated patients were higher than those in improved patients. This result suggests that inflammatory storm may play an important role in deterioration of COVID-19 patients. It is envisaged that glucocorticoids, a commonly used immunosuppressive drugs in the clinic, may have certain effects on COVID-19. However, some experts proposed that while inhibiting lung inflammation, glucocorticoids could also suppress immune response and pathogen clearance [15] . At present, the treatment of glucocorticoids on COVID-19 is still controversial. It still need large-scale randomized clinical trials to confirm whether glucocorticoids can accurately improve the clinical outcome.

13
There is no specific treatment strategy for COVID-19. The main treatment is respiratory support. As a treatment strategy recommended in the guidelines for mechanical ventilation in patients with ARDS [16] , prone position ventilation is also considered in COVID-19 patients with severe ARDS. Theoretically, prone position ventilation could improve oxygenation of patients by reducing gravity-dependent atelectasis, lowering pleural pressure gradients, and improving ventilation-perfusion ratios. However, its effect on mortality in ARDS patients is uncertain. The latest meta-analysis on prone ventilation, publishing in the annual report of the American Thoracic Society in 2017, enrolled 8 randomized controlled clinical trials (2129 patients). It was indicated that no significant difference of 28-day mortality was observed between ARDS patients with prone ventilation and supine ventilation. And subgroup analysis suggested that prolonged prone ventilation duration greater than 12 h/d might be beneficial in reduced mortality in patients with severe ARDS [17] . To date, no related researches on the effectiveness of prone ventilation in patients with COVID-19 and ARDS were reported. In this study, 13 patients underwent prone ventilation during treatment. Data analysis showed that PaO 2 /FiO 2 improved(PaO 2 /FiO 2 increased at least 20% after prone position) in 33.3% of all prone position sessions.
Arterial blood gas analysis showed that PaCO 2 improved (the arterial blood PaCO2 decreased at least 6mmHg after prone position) in 18.5% of prone sessions. Of these 13 patients, 8 patients died, 4 patients were still hospitalized in the ICU, and only 1 patient discharged from ICU after his condition improved. Therefore, we speculated that prone ventilation might improve the oxygenation of some COVID-19 patients with severe ARDS, but a significant mortality benefit with proning was uncertain. However, as the number of patients in this study is relatively small, our speculation need to be confirmed through further studies with large samples.
There are some limitations in this study. Firstly, the number of cases is relatively small, so it is impossible to use multivariate analysis of variance to analyze the risk factors for the worsening of the disease on the 14th day of ICU admission. It will be more representative to use large sample data to analyze the clinical characteristics of severe and critical COVID-19 patients. Secondly, as COVID-19 patients need to return to the general isolation ward for further treatment after being discharged from ICU, the follow-up period in this study is relatively short, and the final clinical outcome (death or cure) of the patient cannot be fully calculated. Two patients in this study returned to the general isolation ward for continue treatment after their condition improved, but suddenly respiration worsening happened and eventually patients died. Therefore, the analysis of the final clinical outcome may be more instructive for the treatment of COVID-19 patient. Finally, recent research [18] found that SARS-CoV-2 load in respiratory specimens of COVID-19 patients was closely related to the severity of the disease and the degree of lung injury. It was believed that early and continuous monitoring of virus load may have certain predictive effects on disease development. However, due to lack of data, this study failed to analyze the correlation between SARS-CoV-2 load in respiratory specimens from COVID-19 patients and clinical outcome.

Conclusion
The majority of patients with severe and critical COVID-19 who need to stay in the ICU were over 50 years old and male. 52.7% need invasive mechanical ventilation. On the 14th day of admission, 56.4% of the patients improved, 34.5% of the patients deteriorated. The rate of deaths during hospitalization was 21.8%. The worsening of patients with COVID-19 might be related to excessive inflammatory and immune responses. In addition, prone ventilation may improve oxygenation in some COVID-19 15 patients with severe ARDS, but a significant mortality benefit with proning was uncertain.  Tables   Due to technical limitations, Tables 1 -5 are only available for download from the Supplementary Files section. Figure 1 Flow chart of the study.