56% of the patients enrolled in this multicenter study had never been to Wuhan and had been infected outside Wuhan. This suggests a gradual shift of initial infection to second-generation local infection which should be taken into account.
The percentage of male patients in our data was 48.5%, different from the male patient predominence reported in two studies on Wuhan cases (73% in Huang et al[13]
and 68% in Chen et al[14] ). In this study, the male-female ratio was approximately 1:1.06, with no difference between severe and non-severe cases. This finding is contradicting to the previous conclusion that men were more susceptible than women to SARS-CoV-2.[14, 15] This might be related to occupational exposures, for more men than women work as salesmen or market managers in seafood markets. As recorded, 66.0% of the patients in Huang’s report and 49% of the patients in Chen’s report had the history of exposure to the Huanan Seafood Market, and most of the affected patients were male workers.[13, 14] In contrast, no patient in our study had such exposure. All of these indicated a change of transmission mode outside Wuhan and that gender may not be a susceptible factor for COVID-19.
The median age of our patients was 46 years old, close to that of patients outside Wuhan as reported by Wu et al (46 years)[16] and Xu et al (41 years),[17] and younger than that of patients in Wuhan as reported by Wang et al (56 years)[18] and Chen et al (55 years).[14] Similarly, severe patients were much older than non-severe patients. This suggests that age may be an important risk factor for poor outcome. The role of age in COVID-19 seems to be similar to its role in SARS and MERS, which has been reported as an independent predictor of adverse outcome.[19, 20] T-cell and B-cell hypofunction and excessive production of type 2 cytokines in older people could lead to defect in inhibition of viral replication and stronger host innate responses with sustained cytokine storm, potentially leading to poor outcome.[21] Therefore, compromised immune function might be the major cause of higher mortality in older people infected by coronaviruses.
The proportion of severe cases in Shaanxi was close to that in Wuhan as reported by Wang et al.,[18] while the incidence of complications and mortality were considerably lower among Shaanxi patients than among the initially infected Wuhan patients.[13, 14, 18] Only two cases in our cohort needed mechanical ventilation. This might indicate that patients outside Wuhan had a much better prognosis than the first generation patients in Wuhan. What’s more, of the cases in Wuhan, those initially identified had a higher mortality than those confirmed and treated later (15%[13] vs. 11%[14] vs. 4.3%[18]). This phenomenon was similar to that during the transmission of MERS-CoV, in which the global mortality of the first-generation MERS-CoV was about 35.5%, while that of the second-generation was around 20%.[22] Furthermore, the median interval from symptom onset to hospital admission in Shaanxi cases was shorter than in Wuhan cases (4.5 vs. 7 days).[13, 18] and the Shaanxi patients were younger than those in Wuhan (46 vs. 55–62 years).[14, 15, 18] These may be reasons for the notable reduction in mortality in Shaanxi cases.
The percentage of cases having fever in our cohort was lower than that reported in Wuhan.[13, 14, 18] In this regard, patients with normal temperature may be missed if the surveillance case definition focused heavily on fever detection. Compared with non-severe patients, severe patients more commonly had symptoms and signs such as cough, sputum, chest stuffiness, dyspnea, temperature above 38℃, respiratory rate above 21 breaths per minute, and heart rate above 100 beats per minute. The onset of symptoms and signs may assist physicians in identifying patients with greater severity.
Based on the radiological data, the incidences of bilateral pneumonia and pleural effusion were higher in severe cases than in non-severe cases, which suggested greater disease severity. Similar to what was reported by Sun et al,[23] in 54.7% (69/126) of the pneumonia cases, pulmonary fibrosis was found in later chest CT images when shadowing had been resolved, and the phenomenon was more common in severe patients than in non-severe patients. These findings consistently suggest that pulmonary fibrosis can be one of the sequelaes of COVID-19. It is necessary and important to explore how to prevent and reduce the occurrence of pulmonary fibrosis and how to manage the situation whenever it occurs in the treatment of COVID-19.
In terms of laboratory tests, different from cases in Wuhan, most Shaanxi patients had lymphocytes within the normal range, and only 38.1% showed lymphopenia. The lymphocyte absolute count in our cohort of patients (1.1 × 109/L) was higher than that reported in Wuhan patients (0.6–0.8 × 109/L).[13, 18, 24] This may be another reason for the lower mortality of Shaanxi cases as compared with of Wuhan cases. In severe cases, the lymphocyte count was lower and the incidence of lymphopenia was higher than non severe cases. These findings suggest that SARS-CoV-2 might mainly act on lymphocytes, especially T lymphocytes, and the severity of lymphopenia might reflect the severity of the disease. Furthermore, levels of inflammatoryparameters, such as CRP and ESR elevated in COVID-19 patients and were even higher in severe patients. The changes of these laboratory parameters illustrated that the virus invaded through respiratory mucosa and spread in the body, triggering a series of immune responses and inducing severe inflammation and cytokine storm in vivo.[25]
Few patients in our study had abnormal levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and indirect bilirubin (IDBIL). The median level of DBIL in the patients elevated, and was even higher in severe patients. As reported, the potential mechanism of liver dysfunction in COVID-19 could be that the virus might directly bind to ACE2 positive bile duct cells.[26] Therefore, the liver abnormality of COVID-19 patients may not be caused by liver cell damage, but by bile duct cell dysfunction.
In addition, elevated glucose, LDH, HBDH, and pro-brain natriuretic peptide, as well as declined albumin, PaO2, and PaO2:FiO2, were more commonly seen in severe cases, suggesting greater disease severity.
The dynamic changes of six laboratory markers showed that baseline lymphocyte count was significantly higher in survivors than in the non-survivor patient, and it increased as the condition improved, but declined sharply when death occurred. Conversely, the IL-6 level displayed a downtrend in survivors, but continually rose to a very high level in the non-survivor patient. Hence, we assume that T cellular immune function might relate to mortality, and lymphocyte and IL-6 should be used as indicators for prognosis. Additionally, CRP, LDH, HBDH, and DBIL levels decreased as the condition improved in recovered patients, but increased rapidly as the condition worsened in the two critically ill cases. These may be related to cytokine storm and bile duct cell dysfunction induced by virus invasion.
Most patients (96.3%) in our study received antiviral therapy, including lopinavir/ritonavir (64.9%), interferon alpha inhalation (50.7%), arbidol (42.5%), ribaviron (32.8%) and chloroquine (2%). Up to now, no specific treatment has been recommended for COVID-19 infection except for optimal supportive care. A previous study showed that combination of lopinavir and ritonavir was associated with substantial clinical benefit for SARS infection.[27] Another study claimed that remdesivir had a good therapeutic effect on COVID-19.[28] Currently, randomised clinical trials for lopinavir/ritonavir (ChiCTR2000029308) and intravenous remdesivir (NCT04257656, NCT04252664) in treatment of COVID-19 are in progress.[24]Meanwhile, COVID-19 vaccine is highly expected. Ongoing efforts are needed to explore effective therapies for this emerging acute respiratory infection.