Demographic and Clinical Characteristics
Details of demographic and clinical characteristics are shown in table 1. A total of 65 patients were included in our study with 10 (15.38%) mild, 32 (49.23%) general, 8 (12.31%) severe, and 15 (23.08%) critical type. The mean (SD) age was 48.4 (18.46) years and 55.38% of them were males. Three patients (4.62%) were under 15 years old (the youngest one was 3 years), and 15 patients (23.08%) were over 65 years old. All 3 younger patients (all with mild type) had a quick recovery to discharge with median virus RNA clearance time of 4 days. Patients with critical type tended to be older than those with mild type (P<0.0001). There was no significant difference by gender among different clinical types (P = 1.000).
A history of recent travel to or living in Wuhan, contact with people confirmed with COVID19, and contact with people from Wuhan was documented in 55.38%, 26.15% and 7.69% of patients, respectively. Not all patients could provide their exact exposure time. There were 49 patients with complete exposure time information. The range of time from exposure to onset of symptoms (incubation period) was 0 to 21 days, with median (IQR) of 6 (4–10) days. Notably, 2 patients presented at 14 days, 1 patient at 16 days and 1 patient at 21 days.
Twenty-two patients (33.85%) had at least one concurrent disease (i.e. hypertension, diabetes, malignancy, endocrine disease, and tumor). Patients with any concurrent disease were significantly more likely to be diagnosed as critical cases (60% with any concurrent disease) as compared to other patients (mild 10%, general 25%, and severe 50%) (P = 0.0264).
Fever (86.15%) and dry cough (56.92%) were the most common symptoms. The mean (SD) maximum temperature was 38.04 (0.86)℃. There was no significant difference in fever and dry cough among patients with different clinical types (all P>0.05). The median (IQR) times of onset of symptom to laboratory confirmation and pneumonia were 5 (3–7) and 6 (3–8) days, respectively. The median (IQR) time of onset of symptom to virus RNA clearance was 15 (11–20) days. Virus RNA clearance time lengthened with disease progression (P = 0.0001). The most common pattern on chest X-ray and CT in patients with pneumonia was bilateral patchy shadowing and bilateral ground-glass opacity, respectively.
Laboratory Parameters of patients in different clinical types
Details of laboratory results are shown in table 2 and supplemental table 1. Laboratory abnormalities, including aspartate amino transferase (ALT), CK, potassium, creatine kinase-MB (CKMB), myocardial troponin I (cTnI), myoglobin, N-terminal pro-brain natriuretic peptide (NT-pro BNP), and lactate (LAC), were more common as disease worsened (All P<0.05). Nineteen (29.23%) patients had lymphopenia (lymphocyte count <1.0×10⁹/L). No mild type patients had lymphopenia, however, 73.33% of critical type did (P<0.0001). In lymphocyte subgroup analysis, 20(30.77%), 29(44.62%), 35(53.85%), and 18(27.69%) patients had decreased total, T, CD4, and CD8 lymphocyte, respectively. Decreases in all four indicators above were more common as the disease worsened (All P<0.05). Meanwhile, lymphocyte counts increased gradually as patients recovered. T lymphocyte count trends in different clinical types throughout the course of the disease are shown in figure 1a.
Forty-four (67.69%), 34 (52.31%), 33 (50.77%), and 35 (53.85%) patients had increased IL–6, ESR, serum ferritin, and CRP, respectively. Increases in all four infection-related biomarkers above were observed more often in critical patients (80%, 80%, 73.33%, and 73.33%, respectively) than in mild patients (10%, 10%, 10%, and 20%, respectively) (All P<0.05). During treatment, the IL–6 levels fluctuated and then decreased to normal as patients recovered. The fluctuation was probably influenced by corticosteroids use. IL–6 trends in different clinical types throughout the course of the disease are shown in figure 1b.
Treatment and clinical outcomes in different clinical types
Details of treatment and clinical outcomes are shown in table 1. Interferon-α (IFN-α) (59, 90.77%) and lopinavir/ritonavir (50, 76.92%) were the main antivirus medicines used in these patients. Oxygen therapy, high-flow nasal cannula (HFNC), non-invasive positive pressure ventilation (NIPPV), and invasive positive pressure ventilation (IPPV) were applied to 14 (21.54%), 8 (12.31%), 12(18.46%), and 3 (4.62%) patients, respectively. The median (IQR) time of onset of symptom to HFNC use and mechanical ventilation was 13 (11–19) days and 14.5 (11.5–17) days, respectively. CRRT was applied to 3 critical patients in AKI stage 2.
Methylprednisolone was applied to 31 (47.69%) patients with pneumonia, including 10 (31.25%) general, 8 (100%) severe, and 13 (86.67%) critical patients, respectively. The dosage and duration of methylprednisolone were prescribed individually according to the severity of pneumonia and/or PaO2/FiO2. Thirty of the 31 patients (96.77%) had stopped methylprednisolone due to improvement of pneumonia on chest X-ray/CT or PaO2/FiO2. One patient died of severe ARDS during methylprednisolone use. Another patient died of septic shock after methylprednisolone use had been stopped 11 days. The median (IQR) time of methylprednisolone use was 7 (5–9) days. The side effects of using methylprednisolone in these 31 patients were hypertension (8, 25.81%), hyperglycemia (11, 35.48%), hypokalemia (11, 35.48%), arrhythmia (4, 12.9%), neuropsychiatric symptoms (2, 6.45%), and gastrointestinal bleeding (1, 3.23%). All above side effects were relived after symptomatic treatment.
The median (IQR) of virus clearance time in patients without methylprednisolone (12.5, 6–17 days) was shorter than in patients with methylprednisolone (19, 14–22 days) (P = 0.0003). But in general type, there was no significant difference in virus clearance time between patients with (15, 12–19 days) and without methylprednisolone use (14.5, 11–18 days) (P = 0.7372).
A linear mixed model showed the relationship between clinical type or corticosteroids used and dynamic IL–6 values or T lymphocyte counts throughout the course of the disease (Table 3 and 4). The confounding variables included age, gender, and concurrent disease. IL–6 increase and T lymphocyte decrease were more common as the disease worsened (All P<0.05). Corticosteroids did not affect T lymphocyte counts (P = 0.0796) but inhibited IL–6 levels (P = 0.0215). In patients using corticosteroids, there was no significant difference in dynamic IL–6 values or T lymphocyte counts between patients using lower dose (≤2mg/kg.d) and higher dose(>2mg/kg.d) methylprednisolone throughout the course of the disease (All P>0.05). The IL–6 levels in patients using corticosteroids decreased quickly after using corticosteroids. Whereas in patients without corticosteroids use, the decrease of IL–6 levels was later and more gradual as patients recovered (Figure 1c). The T lymphocyte count trends had a similar pattern throughout the course of disease in patients with and without corticosteroids use (Figure 1d).
The number of patients discharged, admitted to the ICU, and deceased were 45 (69.23%), 4 (6.15%), and 2(3.08%), respectively. All 45 patients were discharged with full recovery. In these patients, the median (IQR) time from the symptom time to discharge was 19(14–30) days, including 13 (9–18) days for mild, 19 (15–27) days for general, 31(23.5–32) days for severe, and 31 (27–34) days for critical patients (P = 0.0003).