4 DISCUSSION
Our study reports the analysis of clinical characteristics and outcomes of 20 confirmed COVID-19 patients characterized by severe disease. Nine (45%) of these patients died within 35 days of admission to the hospital.
Our critically ill patients with COVID-19 were older which is in line with other studies.8,18 In previous studies involving severe COVID-19 patients, most common symptoms were fever, cough, fatigue, and dyspnea.6-8,9,11,18 Similarly, the incidence of cough, fatigue or myalgia, fever, and dyspnea on admission in our study were 95% (n=19), 90% (n=18), 85% (n=17) and 80% (n=16) respectively. In these patients mean duration of symptoms was 11±9 days from onset to ICU admission, and the previous studies reported similar (from 7-12 days) duration of symptoms prior to ICU admission.5,7,9,11
More than half of the patients (n=11, 55%) in our center had underlying diseases, similar to the data reported by the study from Wuhan (n=64, 46%).8 However, a much higher percentage of patients has been reported in other studies.7,18 Similar to the previous studies,8,18 hypertension was the most common comorbidity, followed by cardiovascular disease and diabetes in our study. This requires more follow-up data for the observation of hypertension, the hypertension treatment received by the patient, and the assessment of high-risk factors.
With regards to laboratory indicators, 11 (55%) patients in our center developed lymphocytopenia on admission, and it mainly occurred in the death group (n=8). Decreased lymphocyte count was significantly associated with the death of COVID-19 patients in ICU. This is consistent with the results of a single factor analysis of the retrospective study from Wuhan.8 Previous studies of Severe Acute Respiratory Syndrome [SARS] and Middle East Respiratory Syndrome [MERS] have also found lymphopenia in their studies. Studies have confirmed that lymphopenia is one of the earliest changes in SARS and a reliable prognostic indicator of SARS.19 In addition, studies have also shown that MERS coronavirus can induce T cell apoptosis through the activation of apoptotic pathways.20 In several previous studies on moderate disease patients infected with SARS-CoV-2, only 35%6 and 40%5 of patients had mild lymphopenia, whereas in other studies severe SARS-CoV-2 associated lymphopenia occurred in more than 70%12 and 80%11 of the infected patients, suggesting that lymphopenia may reflect the severity of SARS-CoV-2 infection.
Most critically ill patients in our center had an increased LDH (n=17, 85%) when they were admitted, slightly higher than the 76%6 and 67%5 observed in patients from the other studies. All patients in our study had elevated D-dimers on admission, and 15 (75%) patients had D-dimers > 1 mg/L. However, in the study of moderate disease COVID-19 patients, only 36% had an increase in D-dimer.6 Another early study from Wuhan also showed that ICU patients had a higher level of D-dimer (median D-dimer level of 2.4 mg/L) than non-ICU patients (median D-dimer level of 0.5 mg/L). Studies have also found that D-dimer > 1 mg/L has been associated with increased mortality of patients.5 Our study did not observe differences between survival and death groups in terms of D-dimer, which may be related to small sample size. The CRP of all patients in our study was increased to 101.46±65.60 mg/L, which was much higher than the average CRP in moderate disease COVID-19 patients, 51.4±41.8 mg/L.6 The probability of CRP increase (100%) was also higher in our study than that of moderate disease COVID-19 patients, 60.7%18 and 86%.6 Similarly, the probability of elevated ProCT (85%, n=17) was much higher in our study than the 6% to 30% probability found in moderate disease COVID-19 patient studies.6,5,18 It is, therefore, suggested that secondary bacterial infection may be a complication in severe patients and cannot be disregarded.
Mechanical ventilation is the main supportive treatment for COVID-19 critically ill patients. In our study, only six patients (30%) received further IMV treatment, which is much lower than reported in other ICU patients: 88% (Lombardy, Italy),10 47% (Wuhan),8 42% (Wuhan),11 and 30% (Wuhan).7 Non-invasive ventilation was used more frequently, with 70% of patients receiving NIV in our study, as compared to the rate of NIV use in other ICU studies: 42% (Wuhan),8 56% (Wuhan),11 and 62% (Wuhan).7
Acute hypoxemic respiratory failure caused by ARDS has been found to be the most common complication (60% to 70% of patients admitted to the ICU), followed by shock (30%), myocardial dysfunction (20% to 30%) and acute kidney disease injury (10%~30%) [5,6 9-11].7-11 In our study, acute heart injury was the most common (19 cases, 95%), followed by ARDS (12 cases, 60%), shock (12 cases, 60%), and pneumothorax (2 cases, 10%). A study of mild disease COVID-19 patients in Wuhan reported that 53% of patients in their cohort died of respiratory failure, 7% died of shock (probably caused by fulminant myocarditis), and 33% of patients died of both conditions.21 In our study, the 10 patients in the death group were all complicated with shock and ARDS. The probability of death and ARDS in the death group was much higher than that in the survival group. Due to the limited sample size, this study failed to show the association of concurrent ARDS or shock with the death of patients. We speculate that concurrent ARDS and shock may be related to the death of severe COVID-19 patients. Due to the limitation of sample size and lack of laboratory data, our study found that only increased BMI and decreased lymphocyte count were significantly associated with an increased chance of death in patients with COVID-19 in the ICU.
This study had several limitations. Firstly, the study was conducted in a single center and was limited by the time of follow-up. Only 20 patients with serious and critical COVID-19 disease patients were included. Smaller sample size may reduce the reliability of the statistical analysis. Secondly, there was a lack of detailed medical and treatment information during the hospitalization of patients, such as mechanical ventilation parameters, blood gases, patient medication, imaging examinations, and other supportive treatments. Finally, this is a retrospective study that failed to include adequate laboratory tests on all patients, including interleukin 6, serum ferritin, etc., resulting in an inability to assess their role in predicting hospital deaths. The findings of our study can be utilized by future studies to further investigate the clinical course and outcome of COVID-19 patients hospitalized in ICU. However, a prospective multicenter study with a larger sample size is necessitated to further explore the factors associated with nosocomial death in severe COVID-19 patients.