We retrospectively collected epidemiological, clinical, laboratory, and radiological data from electronic medical records of 150 COVID-19 patients admitted to the Ryan People's Hospital (46 patients) and the Wenzhou Central Hospital (104 patients) between January 17^th, 2020 and February 4^th, 2020. All patients were consecutively selected for this study. Both hospitals are designated tertiary hospitals for major infectious diseases in the Zhejiang province. These cases were divided into four generations according to the epidemic sources of the infection spread [9]: Generation one included patients that returned from Wuhan to Wenzhou within 14 days before symptom onset and were exposed to Huanan Seafood Wholesale Market; generation two were cases found in patients who returned from Wuhan within 14 days before symptom onset, but without exposure to Huanan Seafood Wholesale Market; generation three was comprised of patients who did not visit Wuhan or other cities between December, 2019 to February, 2020 but had known exposure to patients returned from Wuhan; generation four was for patients who did not visit Wuhan and had no suspected contact with patients returned from Wuhan. Cases of generation one and two were further categorized as imported cases and generation three and four as locally transmitted. This study was approved by the Ethics Review Boards of Wenzhou Central Hospital and Ryan People’s hospital. Consent to participate was explained to the patient and written consent was received.

Clinical management of COVID-19 followed the Guidelines on the Diagnosis and Treatment of COVID-19 Patients (Trial Version 4) of the National Health Commission of China [3]. All patients underwent Reverse Transcription Polymerase Chain Reaction (RT-PCR) tests, and chest CT scans (1.25mm in thickness, continuous scans, lung and mediastinum windows respectively). For RT-PCR tests, nasopharyngeal swabs were obtained and sent to Wenzhou Center for Disease Control for testing for SARS-CoV-2, influenza A and B, avian influenza, adenoviruses, swine-flu viruses, SARS-CoV and Middle East Respiratory Syndrome Coronavirus (MERS-CoV). Abnormal presentations on CT scans were categorized as ground-glass changes or pulmonary consolidation according to the density of pulmonary lesions, or as central or peripheral pulmonary diseases based on their distances to the hilum. Routine laboratory tests included complete blood counts and serum biochemical tests for kidney and liver function, levels of creatine kinase, lactic acid dehydrogenase and electrolytes. Patients with positive RT-PCR test results were isolated in negative pressure isolation wards. Antiviral (interferon alpha-2b, aerosol inhalation 5 million U, bid; lopinavir tablet, 500mg bid, arbidol tablet, 0.2 tid) and other supportive therapies were given. Nasal oxygen support and noninvasive ventilator were given to patients who became hypoxemic.

COVID-19 was diagnosed in both hospitals by following the same guidelines from the National Health Commission of China [3]. Diagnostic criteria for COVID-19 included 1) at least two of the following clinical manifestations, fever and respiratory tract symptoms, radiographic presentations suggesting pneumonia, normal or reduced white blood cell counts or reduced lymphocyte counts at the early stage of illness, and an epidemiological link to confirmed COVID-19 cases and, 2) apositive RT-PCR result for SARS-CoV-2.

The severity of COVID-19 cases was classified according to the same guidelines, as follows; 1) mild infections, with mild clinical symptoms and no abnormal changes onchest CT scans; 2) moderate infections, withfever, respiratory tract and other symptomsand a typical pneumonia presentation on CT images; 3) severe infections, if any of the following indicators presented, respiratory distress with a Respiratory Rate (RR) > 30 times/min,an oxygen saturation of 93% or below, and a PaO₂/ FiO₂ ratio (the ratio of arterial oxygen partial pressure to fractional inspired oxygen)< 300mmHg; 4) critical infections, including respiratory failure requiring a mechanical ventilator, shock, other organ failures that required intensive care.

The discharge criteria for COVID-19 used by both hospitals between January and April, 2020 included a normal temperature for longer than 72h, significant improvement of respiratory tract symptoms including cough, sputum production, chest tightness and shortness of breath, and two consecutive negative SARS-CoV-2 RT-PCR tests at least 24 h apart [3].

Data were entered into a standardized collection form. Median values representing continuous variables were compared using the non-parametric Mann-Whitney method. Categorical variables for clinical manifestations and severity of patients in different groups were compared using the Chi-squared (X2) test.To more accurately predict the contribution of different patient related risk factors to disease severity or its associated clinical presentations, a multivariable analysis was further conducted, using infection severity and lowered leucocyte counts as response variables respectively and selected risk factors as explanatory variables. Risk factors that might differ between the imported infection group and locally acquired infection group were selected, including gender, age, cough, length of hospitalization, and pulmonary consolidation. Mechanical ventilation was excluded as this treatment had an immediate link to disease severity. SPSS 24.0 Software was used for statistical analysis. A P-value less than 0.05 was considered to be statistically significant.

The first COVID-19 case in Wenzhou was diagnosed on January 16, 2020 and the patient was immediately admitted to the Wenzhou Central Hospital. Between January 17 and February 4, 2020, further 149 COVID-19 patients were hospitalized at the Ryan People's Hospital (46 patients) and the Wenzhou Central Hospital (104 patients in total). All 150 patients were permanent residents of Wenzhou. Among these patients, 77 visited Wuhan between December, 2019 and February, 2020 for business purposes.; three had visited the Huanan Wholesale Seafood Market. Seventy-three patients had no travel history to Wuhan or other cities in the Hubei province during the same period; 37 had confirmed exposure to patients who recently returned from Wuhan. All patients were successfully treated and discharged before April 2, 2020.

There was no significant difference between the imported case group and the local transmission group, in the birth origin of patients, gender (P=0.193), age (P=0.091), age distribution using 45 and 65 years of age as cutoff values (P=0.349), prevalence of underlying diseases (P=0.193), and the interval between symptom onset to hospitalization (P=0.702) (Table 1). Of 150 patients, 50.7% were male and 49.3% were female (Table 1). The median age of all patients was 45 years old (Table 1). Only 12% of patients had underlying diseases, with diabetes mellitus being the most prevalent (10/18). The median time from symptom onset to hospitalization for all patients was 5 days.

Epidemiological investigation of patients’ travel and residence history as well as contact tracing revealed a clear link between individual patients with the second generation of COVID-19 and several small clusters with the third generation of infections, suggesting possible viral passages from imported cases to locally transmitted cases.

There was no significant difference in the clinical manifestations, radiographic characteristics and laboratory test results between the locally transmitted and imported infection groups, with the exception of a slightly lower lymphocyte count in the latter (Table 1). Common symptoms of sympto- matic COVID-19 cases in Wenzhou included fever (129 cases, 86.0%), cough (99 cases, 66.0%), chest tightness and/or chest pain (21 cases, 14.0%) and diarrhea (25 cases, 16.7%) (Table 1). At the time of admission 136 patients (90.7%) had normal total white blood cell counts while 53 patients (35.3%) had a lowered lymphocyte count (<1 x10⁹/L). Leukopenia was more common in the imported infection group in comparison to the locally infected group (P = 0.02) (Table 1). An increase in the CRPlevel was observed in 97 patients (64.7%). One hundred forty patients (93.3%) had abnormal chest CT images, with 134 patients (89.3%) showing ground-glass opacities and 107 patients (71.3%) showing pulmonary consolidation. Such CT changes often involved multiple lobes in both lungs, with a median number of four lobes affected per patient. Thirty-seven locally infected patients who had a confirmed exposure to imported COVID-19 (the 3^rd generation) and 36 who had no known exposure (the 4^th generation) were further compared. No differences in the clinical presentations, radiographic characteristics, and laboratory test results were identified.

No specific treatments were allocated to patients in the imported and locally-acquired infection groups, with the exception of noninvasive mechanical ventilation that was given to two patients of 3^rd generation due to hypoxemia. Among 150 patients, 146 received interferon, 132 received Lopinavir/Ritonavir, and 77 patients received Umifenovir. Thirty-one patients also received empiric broad-spectrum antibiotics. Systemic steroids were not used in these patients.

There was no difference in the discharge rates and length of hospital stay between the two groups (Table 1). All patients were discharged before 2^nd of April, 2020 with a median hospital stay of 21 days. There were no deaths in this cohort of COVID-19 patients.

The main focus of this study was to compare the severity of COVID-19 in patients who were infected in Wuhan prior to returning to Wenzhou and those who were infected locally, via a known or unknown exposure. There were 13 cases of severe infections and three cases of critical infections among the 150 COVID-19 cases (Table 2). A majority of severe and critical patients were noted as being in the imported case group (12/16 cases, 75.0%). All three critical infections occurred in patients with imported COVID-19. A Chi-square test suggested that patients who returned from Wuhan were more likely to develop severe or critical conditions compared to those who acquired infections locally (15.6% versus 5.5%, X²=4.016, P= 0.045). Regression analysis indicated that there was a strong association between the severe condition of COVID-19 and the leukopenia status of patients (P=0.006). Multivariable analysis using disease severity as the single outcome suggested that gender and age were significant contributors. Importation from Wuhan also appeared to be a substantial risk factor, with a P value (0.074) reaching a significant level (Table 3). A parallel multivariable analysis using leukopenia as the single outcome suggested that importation from Wuhan, age, length of hospitalization, and pulmonary consolidation were all significant predictors of lowered lymphocyte counts (Table 4).