Multi-modal Representation Learning for Social Post Location Inference (preprint)

Inferring geographic locations via social posts is essential for many practical location-based applications such as product marketing, point-of-interest recommendation, and infector tracking for COVID-19. Unlike image-based location retrieval or social-post text embedding-based location inference, the combined effect of multi-modal information (i.e., post images, text, and hashtags) for social post positioning receives less attention. In this work, we collect real datasets of social posts with images, texts, and hashtags from Instagram and propose a novel Multi-modal Representation Learning Framework (MRLF) capable of fusing different modalities of social posts for location inference. MRLF integrates a multi-head attention mechanism to enhance location-salient information extraction while significantly improving location inference compared with single domain-based methods. To overcome the noisy user-generated textual content, we introduce a novel attention-based character-aware module that considers the relative dependencies between characters of social post texts and hashtags for flexible multi-model information fusion. The experimental results show that MRLF can make accurate location predictions and open a new door to understanding the multi-modal data of social posts for online inference tasks.

Prevalence of psychological symptoms and associated risk factors among nurses in 30 provinces during the COVID-19 pandemic in China

Summary Background With the outbreak of the coronavirus disease 2019 (COVID-19), nurses have won well-deserved recognition for their indispensable roles in providing humane and professional healthcare for patients. However, by the nature of their role working at the forefront of patient care, nurses are prone to experiencing mental health consequences. Therefore, we pay attention to measuring the magnitude of psychological symptoms and identifying associated factors among nurses in China. Methods We launched a nationwide, cross-sectional survey of nurses who worked in secondary or tertiary hospitals and public or private hospitals from 30 provinces in China. The prevalence and severity of symptoms of burnout, depression, and anxiety were investigated, respectively. Multivariable logistic regression analyses were performed to identify factors associated with each psychological symptom. Findings A total of 138 279 respondents who worked in 243 hospitals completed this survey. A substantial proportion of nurses reported symptoms of burnout (34%), depression (55·5%), and anxiety (41·8%). In line with the disproportionality of economic development, we noted that the middle or western region was an independent risk factor for depression and anxiety. Compared with those working in the secondary hospital, nurses who worked in tertiary hospitals were associated with a higher likelihood of burnout and depression. Interpretation Nurses are experiencing emotional, physical, and mental exhaustion during the COVID-19 epidemic. Governments and health policymakers need to draw attention to reinforcing prevention and ameliorating countermeasures to safeguard nurses' health. Funding The strategic consulting project of the Chinese Academy of Engineering [2021-32-5]. Advanced Institute of Infomation Technology, Peking University, Zhejiang Province [2020-Z-17]

Titers of antibodies the receptor-binding domain (RBD) of ancestral SARS-CoV-2 are predictive for levels of neutralizing antibodies to multiple variants (preprint)

Diagnostic assays currently used to monitor the efficacy of COVID-19 vaccines measure levels of antibodies to the receptor-binding domain of ancestral SARS-CoV-2 (RBDwt). However, the predictive value for protection against new variants of concern (VOCs) has not been firmly established. Here, we used bead-based arrays and flow cytometry to measure binding of antibodies to spike proteins and receptor-binding domains (RBDs) from VOCs in 12,000 sera. Effects of sera on RBD-ACE2 interactions were measured as a proxy for neutralizing antibodies. The samples were obtained from healthy individuals or patients on immunosuppressive therapy who had received two to four doses of COVID-19 vaccines and from COVID-19 convalescents. The results show that anti-RBDwt titers correlate with the levels of binding- and neutralizing antibodies against the Alpha, Beta, Gamma, Delta, Epsilon and Omicron variants. The benefit of multiplexed analysis lies in the ability to measure a wide range of anti-RBD titers using a single dilution of serum for each assay. The reactivity patterns also yield an internal reference for neutralizing activity and binding antibody units per milliliter (BAU/ml). Results obtained with sera from vaccinated healthy individuals and patients confirmed and extended results from previous studies on time-dependent waning of antibody levels and effects of immunosuppressive agents. We conclude that anti-RBDwt titers correlate with levels of neutralizing antibodies against VOCs and propose that our method may be implemented to enhance the precision and throughput of immunomonitoring.

Multiplexed measurement of binding- and neutralizing antibodies to SARS-CoV-2 variants in 12.000 post-vaccine sera (preprint)

Diagnostic assays currently used to monitor the efficacy of COVID-19 vaccines measure levels of antibodies to the receptor-binding domain of ancestral SARS-CoV-2 (RBDwt). However, the predictive value for protection against new variants of concern (VOCs) has not been firmly established. Here, we used bead-based arrays and flow cytometry to measure binding of antibodies to spike proteins and receptor-binding domains (RBDs) from VOCs in 12,000 sera. Effects of sera on RBD-ACE2 interactions were measured as a proxy for neutralizing antibodies. The samples were obtained from healthy individuals or patients on immunosuppressive therapy who had received two to four doses of COVID-19 vaccines and from COVID-19 convalescents. The results show that anti-RBDwt titers correlate with the levels of binding- and neutralizing antibodies against the Alpha, Beta, Gamma, Delta, Epsilon and Omicron variants. The benefit of multiplexed analysis lies in the ability to measure a wide range of anti-RBD titers using a single dilution of serum for each assay. The reactivity patterns also yield an internal reference for neutralizing activity and binding antibody units per milliliter (BAU/ml). Results obtained with sera from vaccinated healthy individuals and patients confirmed and extended results from previous studies on time-dependent waning of antibody levels and effects of immunosuppressive agents. We conclude that anti-RBDwt titers correlate with levels of neutralizing antibodies against VOCs and propose that our method may be implemented to enhance the precision and throughput of immunomonitoring.

Occupational Stigma Perception, Emotional Exhaustion State, and Professional Commitment Response: Understanding the Mechanisms Underlying Hotel Interns’ Perceptions of Career Prospects

This study uses an integrated model of resource conservation theory and social learning theory to explore the antecedents of hotel interns’ perceptions of occupational stigma and to explore the mechanisms inherent to retention willingness. This study first manipulated relevant subjects’ experimental materials through a contextual experiment and used a one-way ANOVA to test the effects of competence stereotypes and occupational stereotypes on hotel interns’ stigma perceptions, respectively, and then used partial least squares structural equation modeling (PLS-SEM) as a statistical tool and the SmartPLS 3.0 program to validate the model of hotel interns’ occupational stigma perceptions-intention. The effects of both competence stereotypes and occupational stereotypes on hotel interns’ perceptions of occupational stigma were significant. The results of the partial least squares structural equation model showed that hotel interns’ perceptions of occupational stigma significantly contributed to emotional exhaustion and that emotional exhaustion significantly influenced hotel interns’ retention willingness, hotel interns’ perceptions of occupational stigma had a significant effect on their retention willingness, while the role of emotional exhaustion as a mediating variable and occupational commitment as a moderator. The inner psychological and behavioral linkage mechanisms of hotel interns’ occupational stigma perceptions and their retention willingness under COVID-19 were explored, and the resource dynamics operating mechanism and professional commitment were also confirmed.

Comparative Study of HRCT Imaging Characteristics of Psittaci Pneumonia and COVID-19 Pneumonia (preprint)

Backgroud : Both Chlamydia psittaci and COVID-19 virus can cause lung inflammation, which manifests extremely similarly in clinical symptoms and imaging. Especially during the epidemic of COVID-19, psittacosis pneumonia is easily misdiagnosed as COVID-19 pneumonia. The identification of the chest imaging between the two diseases is of special significance when the epidemiological contact history is unclear, and the etiology and nucleic acid test results are not available. This study conducts to compare the imaging characteristics on chest high-resolution CTs (HRCT) between patients with psittaci pneumonia and COVID-19 pneumonia. Methods: : A retrospective analysis of the imaging characteristics on chest HRCTs of 10 psittaci pneumonia patients and 13 COVID-19 pneumonia patients. The similarities and differences in HRCT images of patients with psittaci pneumonia and COVID-19 pneumonia were analyzed. Results: : HRCT showed that among the 10 psittaci pneumonia patients, 8 cases (80.00%) had single lobe involvement, and 2 cases (20.00%) had multiple lobe involvement. Among the 13 COVID-19 pneumonia patients, 2 cases had single lobe involvement (15.38%), and 11 cases had multiple lobe involvement (84.62%). The types of lesions in 10 psittaci pneumonia patients included simple consolidation in 5 cases (50.00%), and ground-glass opacity (GGO) with consolidation in 5 cases (50.00%). The types of lesions in 13 COVID-19 pneumonia patients included simple GGO in 6 cases (46.15%), GGO with consolidation in 4 cases (30.77%), GGO with paving stone sign in 2 cases (15.38%), and simple consolidation in 1 case (7.69%). Lymphadenopathy was observed in 1 psittaci pneumonia patient (10.00%) and 1 COVID-19 pneumonia patient (7.69%). Among the 10 psittaci pneumonia patients, 8 cases (80.00%) had bronchial inflation, and 6 patients (60.00%) had pleural effusion. Among the 13 COVID-19 pneumonia patients, 5 patients (38.46%) showed signs of bronchial inflation, while no pleural effusion was observed in 13 patients. Conclusion: : Chest HRCTs can distinguish COVID-19 pneumonia from psittaci pneumonia, and can provide early diagnoses of these two diseases.

A novel Pan-viral prophylaxis strategy using vaccine adjuvant CAF09b protects against influenza virus infection (preprint)

ABSTRACT The SARS-CoV-2 pandemic caused a massive health and societal crisis, although the fast development of effective vaccines reduced some of the impact. To prepare for future pandemics, a pan-viral prophylaxis could be used to control the initial virus outbreak in the period prior to vaccine approval. The liposomal vaccine adjuvant CAF ® 09b contains the TLR3 agonist polyinosinic:polycytidylic acid, which induces a type I interferon (IFN-I) response and an antiviral state in the affected tissues. When testing CAF09b as a potential pan-viral prophylaxis, we observed that intranasal administration of CAF09b to mice resulted in an influx of innate immune cells into the nose and lungs and upregulation of IFN-I related gene expression. When CAF09b was administered prior to challenge with mouse-adapted influenza A/Puerto Rico/8/1934 virus, it protected from severe disease, although virus was still detectable in the lungs. However, when CAF09b was administered after influenza challenge, the mice had a similar disease course to controls. In conclusion, CAF09b may be a suitable candidate as a pan-viral prophylactic treatment for epidemic viruses, but must be administered prior to virus exposure to be effective.

Immune Features of COVID-19 Convalescent Individuals Revealed by a Single-cell RNA Sequencing (preprint)

Background: The immune response to SARS-CoV-2 is critical in both controlling primary infection and preventing re-infection. However, it remains unclear whether immune responses following natural infection can be sustained or potentially prove critical for long-term immune protection against SARS-CoV-2 reinfection. Here, we systematically mapped the phenotypic landscape of SARS-CoV-2-specific immune responses in peripheral blood samples of 4 healthy donors and 13 convalescent patients with COVID-19, including moderate and severe cases, by single-cell RNA sequencing. Results: : The relative percentage of the CD8+ effector memory subset was increased in both convalescent moderate and severe cases, but NKT-CD160 and maginal zone B clusters were decreased. Innate immune responses were attenuated reflected by decreased expression of genes involved in interferon-gamma, leukocyte migration and neutrophil mediated immune response in convalescent COVID-19 patients. Functions of T cell were strengthened in convalescent COVID-19 patients by clear endorsement of increased expression of genes involved in biological processes of regulation of T cell activation, differentiation and cell-cell adhesion. In addition, T cell mediated immune responses were enhanced with remarkable clonal expansions of TCR and increased transition of CD4+ effector memory and CD8+ effector-GNLY in severe subjects. B cell immune responses displayed sophisticated and dual functions during convalescence of COVID-19, providing a novel mechanism that B cell activation was observed especially in moderate while humoral immune response was weakened. Interestingly, HLA class I genes displayed downregulation while HLA class II genes upregulation in both T and B cell subsets in convalescent individuals. Notably, some unique IGV genes in severe patients may facilitate the design of vaccines. Conclusions: : Our collective dataset showed that innate immunity was declined but SARS-CoV-2-specific T cell responses were retained even strengthened whereas sophisticated and dual functions of B cells, including declined humoral immunity were presented at several months following infections, which provided insights into evaluation of possibility of reinfection of exposed individuals with COVID-19 and facilitation to design of effective therapeutics and vaccines.

Binding Strength and Hydrogen Bond Numbers between COVID-19 RBD and HVR of Antibody

The global battle against the COVID-19 pandemic relies strongly on the human defense of antibody, which is assumed to bind the antigen’s receptor binding domain (RBD) with its hypervariable region (HVR). Due to the similarity to other viruses such as SARS, however, our understanding of the antibody-virus interaction has been largely limited to the genomic sequencing, which poses serious challenges to containment and rapid serum testing. Based on the physical/chemical nature of the interaction, infrared spectroscopy was employed to reveal the binding disparity, the real cause of the antibody-virus specificity at the molecular level, which is inconceivable to be investigated otherwise. Temperature dependence was discovered in the absorption value from the 1550 cm−1 absorption band, attributed to the hydrogen bonds by carboxyl/amino groups, binding the SARS-CoV-2 spike protein and closely resembled SARS-CoV-2 or SARS-CoV-1 antibodies. The infrared absorption intensity, associated with the number of hydrogen bonds, was found to increase sharply between 27 °C and 31 °C, with the relative absorbance matching the hydrogen bonding numbers of the two antibody types (19 vs. 12) at 37 °C. Meanwhile, the ratio of bonds at 27 °C, calculated by thermodynamic exponentials, produces at least 5% inaccuracy. Beyond genomic sequencing, the temperature dependence, as well as the bond number match at 37 °C between relative absorbance and the hydrogen bonding numbers of the two antibody types, is not only of clinical significance in particular but also as a sample for the physical/chemical understanding of vaccine–antibody interactions in general.

Improving Google Flu Trends for COVID-19 Estimates Using Weibo Posts

While incomplete non-medical data has been integrated into prediction models for epidemics, the accuracy and the generalizability of the data are difficult to guarantee. To comprehensively evaluate the ability and applicability of using social media data to predict the development of COVID-19, a new confirmed case prediction algorithm improving the Google Flu Trends algorithm is established, called Weibo COVID-19 Trends (WCT), based on the post dataset generated by all users in Wuhan on Sina Weibo. A genetic algorithm is designed to select the keyword set for filtering COVID-19 related posts. WCT can constantly outperform the highest average test score in the training set between daily new confirmed case counts and the prediction results. It remains to produce the best prediction results among other algorithms when the number of forecast days increases from one to eight days with the highest correlation score from 0.98 (p < 0.01) to 0.86 (p < 0.01) during all analysis period. Additionally, WCT effectively improves the Google Flu Trends algorithm's shortcoming of overestimating the epidemic peak value. This study offers a highly adaptive approach for feature engineering of third-party data in epidemic prediction, providing useful insights for the prediction of newly emerging infectious diseases at an early stage.

Dynamic Blood Tests Reveal Predicted Week of Death And Tricky Week For Surveillance In Severely Ill COVID-19 Patients (preprint)

Background: Blood laboratory tests are the most reliable methods for the diagnosis and assessment of vital organs’ functions and the body’s response to infection. Herein, we compared the results of dynamic blood tests between the survivor and non-survivor group of patients with coronavirus disease 2019 (COVID-19) and aimed to determine the predicted and tricky week for death and surveillance.Methods: The survivor and non-survivor groups were compared using biochemical blood tests, routine blood tests, and coagulation blood tests over four weeks of investigation.Results: Blood urea nitrogen, creatinine, high-sensitivity C-reactive protein, total bile acid, neutrophil count, white blood cell count, D-dimer, fibrin and fibrinogen degradation product, and prothrombin time showed significantly higher levels in the non-survivor group than the survivor group. Only pre-albumin, eosinophil count, lymphocyte count, red blood cell count, platelet count, hemoglobin, and prothrombin activity tests were significantly higher in the survivor group than the non-survivor group. Generally, the third week of the non-survivor’s group could be regarded as the predicted week for death based on all tests except for creatinine, pre-albumin, total bile acid, monocyte count, white blood cell count, and prothrombin activity. The tricky week in the non-survivor group was the second week in all tests except for pre-albumin, basophil count, eosinophil count, lymphocyte count, platelet count, D-dimer, and fibrin and fibrinogen degradation product.Conclusions: Based on our study, specific attention should be given to some weeks with respect to their related tests as predicted or tricky for death or surveillance, respectively.

Long COVID affects home-isolated young patients (preprint)

Long-term complications following COVID-19 are common in hospitalised patients, but the spectrum of symptoms in milder cases remains unclear. In a Norwegian prospective cohort study of 312 patients, 61% of COVID-19 patients (247 home-isolated and 65 hospitalised) had persistent symptoms at six months, most commonly fatigue (37%), impaired concentration (26%), disturbed smell and/or taste (25%), memory loss (24%) and dyspnoea (21%). In young home-isolated adults, aged 16–30, 52% had symptoms at six months, with fatigue (21%), dyspnoea (13%), impaired concentration (13%) and memory loss (11%) posing particular challenges. Pre-existing chronic lung disease and high convalescent antibody titres predicted persistent symptoms, particularly fatigue. Our finding that young, home-isolated adults with mild COVID-19 are at risk of long-lasting dyspnoea and cognitive symptoms highlights the importance of infection control measures, such as vaccination.

High attack rates of SARS-CoV-2 infection through household-transmission: a prospective study (preprint)

Background: Household attack rates of SARS-CoV-2 ranging from 7% to 38% have been reported, using reverse transcription polymerase chain reaction (RT-PCR) of respiratory samples. Lower attack rates were described in children, but the importance of age in household transmission dynamics remains to be clarified. Methods: During the first month of the outbreak, we enrolled 112 households (291 participants) in a prospective case-ascertained study, collecting demographic and clinical data from index cases and household members. Sera were collected 6-8 weeks after index case symptom onset, to measure SARS-CoV-2-specific antibodies. Findings: The overall household attack rate was 45% assessed by seroconversion, and 47% when also including RT-PCR positives. Serology identified a significantly higher number of infected household members than RT-PCR. Attack rates were equally high in children (43%) and young adults (46%), but highest among household members aged [≥]60 years (72%). The attack rate was 16% in asymptomatic household members, and 42% in RT-PCR negative household members. Older adults generally had higher antibody titres than younger adults. The risk of household transmission was higher when the index case had fever or dyspnoea during acute illness but not associated with cough. Interpretation: Serological assays provide more accurate estimates of household secondary attack rate than RT-PCR, especially among children who have a lower RT-PCR positivity rate. Children are equally susceptible to infection as adults, but elderly show higher attack rates. Negative RT-PCR or lack of symptoms are not sufficient to rule out infection in household members.

Correlation Study of Chest CT Features of Severe/Critical type COVID-19 with Early Renal Damage and Clinical Prognosis (preprint)

Background: Among patients with confirmed severe/critical type COVID-19, we found that although the seurm creatinine (Cr) value is in normal range, patients might have occured early renal damage. For severe/critical type COVID-19 patients, whether some chest CT features can be used to predict the early renal damage or clinical prognosis.Methods: 162 patients with severe/critical type COVID-19 were reviewed retrospectively in 13 medical centers from China. According to the level of eGFR, 162 patients were divided into three groups, group A (eGFR < 60 ml/min/1.73m2), group B (60 ml/min/1.73m2 ≤ eGFR < 90 ml/min/1.73m2 group) and group C (eGFR ≥ 90 ml/min/1.73m2). All patients’ baseline clinical characteristics, laboratory data, CT features and clinical outcomes were collected and compared. The eGFR and CT features was assessed using univariate and multivariate Cox regression.Results: Baseline clinical characteristics showed that there were significant differences in age, hypertension, cough and fatigue among groups A, B and C. Laboratory data analysis revealed significant differences between the three groups of leukocyte count, platelet count, C-reactive protein, aspartate aminotransferase, creatine kinase. Chest CT features analysis indicated that crazy-paving pattern has significant statistical difference in groups A and B compared with group C. The eGFR of patients with crazy-paving pattern was significant lower than those without crazy-paving pattern (76.73 ± 30.50 vs. 101.69 ± 18.24 ml/min/1.73m2, p < 0.001), and eGFR (OR = 0.962, 95% CI = 0.940-0.985) was the independent risk factor of crazy-paving pattern. The eGFR (HR = 0.549, 95% CI = 0.331-0.909, p = 0.020) and crazy-paving pattern (HR = 2.996, 95% CI = 1.010-8.714, p = 0.048) were independent risk factors of mortality.Conclusions: In patients with severe/critical type COVID-19, the presence of crazy-paving pattern on chest CT are more likely occured the decline of eGFR and poor clinical prognosis. The crazy-paving pattern appeared could be used as an early warning indicator of renal damage and to guide clinicians to use drugs reasonably.

Prediction Model of Severe Coronavirus Disease 2019(COVID-19) Cases Shows the Leading Risk Factor of Hypocalcemia (preprint)

Background A striking characteristic of Coronavirus Disease 2019(COVID-19) is the coexistence of clinically mild and severe cases. A comprehensive analysis of multiple risk factors predicting progression to severity is clinically meaningful. Methods The patients were classified into moderate and severe groups. The univariate regression analysis was used to identify their epidemiological and clinical features related to severity, which were used as possible risk factors and were entered into a forward-stepwise multiple logistic regression analysis to develop a multiple factor prediction model for the severe cases.Results 255 patients (mean age, 49.1±SD 14.6) were included, consisting of 184 (72.2%) moderate cases and 71 (27.8%) severe cases. The common symptoms were dry cough (78.0%), sputum (62.7%), and fever (59.2%). The less common symptoms were fatigue (29.4%), diarrhea (25.9%), and dyspnea (20.8%). The univariate regression analysis determined 23 possible risk factors. The multiple logistic regression identified seven risk factors closely related to the severity of COVID-19, including dyspnea, exposure history in Wuhan, CRP (C-reactive protein), aspartate aminotransferase (AST), calcium, lymphocytes, and age. The probability model for predicting the severe COVID-19 was P=1/1+exp (-1.78+1.02×age+1.62×high-transmission-setting-exposure +1.77×dyspnea+1.54×CRP+1.03×lymphocyte+1.03×AST+1.76×calcium). Dyspnea (OR=5.91) and hypocalcemia (OR=5.79) were the leading risk factors, followed by exposure to a high-transmission setting (OR=5.04), CRP (OR=4.67), AST (OR=2.81), decreased lymphocyte count (OR=2.80), and age (OR=2.78). Conclusions This quantitative prognosis prediction model can provide a theoretical basis for the early formulation of individualized diagnosis and treatment programs and prevention of severe diseases.

Efficacy and Safety of Traditional Chinese Medicine in COVID-2019: A Study Protocol for an Open-Label, Multicenter Randomized Controlled Trial (preprint)

Introduction: The outbreak of a novel coronavirus (SARS-CoV-2) and associated COVID-19 disease in late December 2019 has led to a global pandemic. It directly leads to high morbidity and mortality, but also results in a devastating effect on the global economy. Unfortunately, there are no effective therapies or vaccines for it. Hence, we designed a randomized trial to evaluate the efficacy and safety of Traditional Chinese Medicine for treating patients with COVID-19.Methods and analysis: This is an open-label, multicenter randomized controlled clinical trial. One hundred and twelve patients infected by SARS-CoV-2 will be randomly assigned to the experimental or the control group in an equal ratio. The patients in control group will accept routine supportive clinically care including the therapies of anti-viral, anti-bacterial and ameliorating the related symptoms, while patients in the experimental group will be asked to take traditional Chinese medicine depending on the different stages of the disease for consecutive 14 days in addition to supportive care. All data will be gathered at baseline and on days 3, 7, 10 and 14. The primary outcome measures will be the time of Reverse Transcription PCR testing of respiratory tract sample turns to be negative. Secondary outcome measures will include Murray lung injury score, MuLBSTA score and TCM ( Traditional Chinese Medicine ) Syndrome Scoring System. A laboratory test will be taken before and after treatment to assess the safety of TCM.Discussion: The study may help to identify the the efficacy and safety of Traditional Chinese Medicine in treating COVID-2019.Trial registration: Chinese Clinical Trial Registry, ChiCTR2000030759.Registered on March 13th 2020-Retrospectively registered, http://www.chictr.org.cn/.

Risk Factors for In-hospital Progression of Ordinary COVID-19 in Wuhan, China: A Retrospective Cohort Study (preprint)

Objective: To describe the clinical characteristics and outcomes of ordinary COVID-19 when admitted, to describe how these patients were treated and risk factors for in-hospital progression.Methods: In this retrospective study, we included 291 adult patients diagnosed as ordinary COVID-19 on admission who had been discharged or had died between Jan 20, 2020 and Mar 16, 2020 from General Hospital of Central Theatre Command (Wuhan, China).Results: Of the 291 patients diagnosed as ordinary COVID-19 when admitted, 65 (22.34%) had been recorded COVID-19 progressing at least once, and 226 (77.66%) had been recorded COVID-19 improving during hospitalization. The median time from admission to disease progressed was 5.0 days (2.0-7.0). Multivariable regression showed increasing odds of in-hospital progression associated with male (odds ratio 2.333, 95% CI 1.135-4.395; P=0.020), preexisting cardiovascular diseases (2.433, 1.044-5.671; P=0.039), and lymphopenia (3.482, 1.783-6.799; P<0.001), elevated IL-6 (2.669, 1.084-6.574; P=0.033), d-dimer (2.829, 1.420-5.636; P=0.003) and lactate dehydrogenase (2.855, 1.458-5.591; P= 0.002) on admission.Conclusions: The potential risk factors of male, preexisting cardiovascular disease, lymphopenia, elevated IL-6, and lactate dehydrogenase, d-dimer could help clinicians to identify in-hospital progression among ordinary COVID-19 at early stage to optimize medical treatment.