The epidemic trend and outline of the COVID-19
Located in the south of China with a population of 13.0 million, Shenzhen reported its first confirmed case on January, 19th 2020. Up until February 19th 2020, there were totally 417 cases confirmed according to the official reports. The epidemical trends of new cases, cumulative cases and remaining cases were shown in the Fig.1A, newly confirmed cases per day reached the peak in around 12 days after first case report and the remaining cases started to decrease after about 20 days. Since February 18th, there were barely new cases added and the situation turned better.
To explore the timeline and disease progression of COVID-19, we focus on the 333 confirmed cases in the first month (admission date from Jan 10th to Feb 10th) (Fig.1B). Most of the patients were admitted to hospital within 4 days after the disease onset, the median interval from disease onset to admission was 3 days (range 1-5 days) (Fig.1C). In the 254 confirmed cases who had clear and credible information of exposure contacts to calculate the incubation period, the median of incubation period was 7 days (range 4-12 days). During the hospitalization, 70(21.1%) mild-moderate cases progressed to severe condition in the median 5 days (range 2-8 days), 23(6.9%) cases were admitted to ICU in median 2 days (range 1-4 days) after progression and unfortunately 3(0.9%) patients died by the end of Feb 28th (Fig.1C).
The baseline clinical characteristics of disease progression
The median age of all the 323 patients was 46 years (IQR, 33-59; range, 8 months to 86 years), the age range and proportions were shown in the Fig 2A. A total of 333 patients were classified according to the criteria defined above. The proportion of patients with mild, moderate, severe and critical on admission were 7.5% (25/333), 89.5% (298/333), 2.1% (7/333), and 0.9% (3/333), respectively. The spectrum of severity of diseases changed as disease progressed, 70 mild-moderate cases progressed to severe condition (progressive group), while 253 patients did not (stable group), and another 10 patients were severe from the beginning of admission (Fig 2B).
Patients who later progressed to severe condition were more likely to have underlying comorbidities compared with the stable group (42.8% vs 16.6%, P<0.05). Of all, hypertension was the most common disease (35, 10.8%), followed by diabetes (5.0%), heart diseases (4.0%), pulmonary disease (2.8%), liver diseases (2.5%), malignancy (0.9%), cerebrovascular disease (0.3%) and other conditions (2.1%) (Fig. 2C).
The clinical characteristics of the progressive and stable groups
As shown in table 2, compared with the stable group, the progressive group was significantly older(P<0.001), there were no one under 18 years and patients over 65 years made up an evidently larger proportion (21.4%) in this group. More than half of the patients (169, 52.3%) were females, however, apparently more men (64.3%) ended up in progressive situation. Of all, 173(53.6%) patients had an exposure history related to Wuhan and 90(27.9%) cases were connected with other cities in Hubei province except Wuhan. Around 167(51.7%) patients lived in Shenzhen but had outside contacts with confirmed or suspected infections or experienced a short term trip outside, whereas only 11(3.4%) patients claimed no obvious exposure history. None of them were hospital-related transmission.
The most prevalent symptom was fever before admission (248,76.8%) and it was almost comparable between two groups (P=0.154). Nearly half of patients were presented with pneumonia symptoms and systemic manifestations, including cough (49.5%), expectoration (22.9%), fatigue or myalgia (21.4%), anorexia (12.4%), dizziness (8.0%), chest tightness (5.0%), dyspnea (2.8%), and all of those symptoms were significantly more common and frequent in the progressive group. Notably, diarrhea and abdominal discomfort occurred in 7.4% of the patients and were slightly different in the progressive and stable cohorts (10.0% vs 6.7%). As for the vital signs, the progressive group tended to have significantly higher temperature and systolic blood pressure, and prone to tachypnea and low oxygenation index compared to the stable one. Interestingly, 24 patients were asymptomatic on admission but still timely hospitalized due to an exposure history and a laboratory-confirmed positive nucleic acid result of COVID-19 virus.
All patients underwent chest CT on admission, 255 (79.0%) patients presented bilateral pneumonia and 35 (10.8%) patients presented unilateral involved, while 33(10.2%) patients showed almost no abnormalities. The progressive group displayed more lobes and segments involved, higher proportion of multiple ground-glass opacities, yet all 33 normal CT appeared only in the stable group.
The laboratory parameters of the progressive and stable patients
There were numerous differences in laboratory findings between the two groups. On admission, the progressive group presented slightly higher white blood cells and neutrophils (P=0.026) than the stable one. However, the counts of Lymphocytes, T lymphocytes, CD4+ cell, CD8+ cell and platelets were significantly lower in the progressive patients, resulting in comparatively high level of Neutrophil-to-Lymphocyte Ratio (NLR). Generally, the baseline parameters representing the function of liver (alanine aminotransferase, aspartate aminotransferase, gamma glutamyl transferase), kidney (Creatinine Cr, blood urea nitrogen BUN) and myocardial zymogram (Troponin T, LDH) were distinctly elevated in the progressive group, indicating the potential organ dysfunction at the beginning. The blood levels of sodium, potassium and PO2, PCO2, oxygenation index were statistically lower in progressive patients, while elevated level of the infection-related indexes, i.e. ESR, CRP, procalcitonin, interleukin-6 (IL-6) were significantly more prevalent in this group on admission, as with the D-Dimer level. The preliminary results of blood test had already altered visibly in the progressive patients at early stage.
Treatments and outcomes of all 333 patients
All of the 333 patients, most patients (71.7%) had oxygen therapy and all patients received antiviral treatment. For severe cases, there was a significantly higher proportion of patients used antibiotics (60.8%), corticosteroid and gamma globulin (both over 75%) for treatment compared with the non-severe one. All the severe patients had oxygen support. In addition, 23 patients were admitted to intensive care unit, 11 of them had to use the invasive mechanical ventilation and 5 patients switched to extracorporeal membrane oxygenation. The most common complication was acute respiratory distress syndrome (ARDS) which happened to 13 severe patients. Other included acute cardiac injury, acute renal injury, septic shock and multiple organ failure which led to death cases. All 3 death cases were males and over 60 years old, one coexisting with hypertension and another with chronic obstructive pulmonary disease. 2 of them were severe-critically ill at admission. Still, more than 240 patients were recovered and discharged from the hospital by February 28th.
The potential risk factors of disease progression
To predict the risk factors of disease progression based on the clinical features, we found that age, sex, history of exposure, comorbidities, radiology manifestation were significantly associated with the disease progression by the univariate logistic analysis. Furthermore, aged over 40 years, male sex, with comorbidities, a clear and certain exposure history and abnormal radiology manifestations were all risk factors for disease progression by the multivariate logistic analysis (Table 4).
As shown in the table 5 of laboratory parameters, the univariate logistic analysis suggested that the baseline levels of NLR, T lymphocyte, BUN, CRP, IL-6, ESR were significantly associated with the disease progression. However, the multivariate logistic analysis indicated that low T lymphocyte level and high levels of CRP, IL-6, NLR were risk factors for disease progression (Table 5).
The diagnosis value and predictors of disease progression
Furthermore, through the ROC curve test(Fig.3), the best cut-off point of age(AUC=0.767) was 53.5 years, with a specificity of 70% and a sensitivity of 28.1%. And the ROC curve of T lymphocyte(AUC=0.865) suggested that the best cut-off point was 825/ul with a specificity of 88.4% and a sensitivity of 26.3%. The ROC curve of CRP(AUC=0.0.768) suggested that the best cut-off point was 9.71 mg/ml with a specificity of 81.4% and a sensitivity of 41.2%.
Compared with the stable group, the length of disease progressing time was significantly different according to the age and sex by the Kaplan-Meier analysis(Fig.4). It can be inferred that the elderly and male patients were more likely to progress into severe-critically ill conditions.