Transcription factor Dmrt1 triggers the SPRY1-NF-κB pathway to maintain testicular immune homeostasis and male fertility.

Bacterial or viral infections, such as Brucella, mumps virus, herpes simplex virus, and Zika virus, destroy immune homeostasis of the testes, leading to spermatogenesis disorder and infertility. Of note, recent research shows that SARS-CoV-2 can infect male gonads and destroy Sertoli and Leydig cells, leading to male reproductive dysfunction. Due to the many side effects associated with antibiotic therapy, finding alternative treatments for inflammatory injury remains critical. Here, we found that Dmrt1 plays an important role in regulating testicular immune homeostasis. Knockdown of Dmrt1 in male mice inhibited spermatogenesis with a broad inflammatory response in seminiferous tubules and led to the loss of spermatogenic epithelial cells. Chromatin immunoprecipitation sequencing (ChIP-seq) and RNA sequencing (RNA-seq) revealed that Dmrt1 positively regulated the expression of Spry1, an inhibitory protein of the receptor tyrosine kinase (RTK) signaling pathway. Furthermore, immunoprecipitation-mass spectrometry (IP-MS) and co-immunoprecipitation (Co-IP) analysis indicated that SPRY1 binds to nuclear factor kappa B1 (NF-κB1) to prevent nuclear translocation of p65, inhibit activation of NF-κB signaling, prevent excessive inflammatory reaction in the testis, and protect the integrity of the blood-testis barrier. In view of this newly identified Dmrt1- Spry1-NF-κB axis mechanism in the regulation of testicular immune homeostasis, our study opens new avenues for the prevention and treatment of male reproductive diseases in humans and livestock.

Computed tomography–based COVID–19 triage through a deep neural network using mask–weighted global average pooling

Background There is an urgent need to find an effective and accurate method for triaging coronavirus disease 2019 (COVID-19) patients from millions or billions of people. Therefore, this study aimed to develop a novel deep-learning approach for COVID-19 triage based on chest computed tomography (CT) images, including normal, pneumonia, and COVID-19 cases. Methods A total of 2,809 chest CT scans (1,105 COVID-19, 854 normal, and 850 non-3COVID-19 pneumonia cases) were acquired for this study and classified into the training set (n = 2,329) and test set (n = 480). A U-net-based convolutional neural network was used for lung segmentation, and a mask-weighted global average pooling (GAP) method was proposed for the deep neural network to improve the performance of COVID-19 classification between COVID-19 and normal or common pneumonia cases. Results The results for lung segmentation reached a dice value of 96.5% on 30 independent CT scans. The performance of the mask-weighted GAP method achieved the COVID-19 triage with a sensitivity of 96.5% and specificity of 87.8% using the testing dataset. The mask-weighted GAP method demonstrated 0.9% and 2% improvements in sensitivity and specificity, respectively, compared with the normal GAP. In addition, fusion images between the CT images and the highlighted area from the deep learning model using the Grad-CAM method, indicating the lesion region detected using the deep learning method, were drawn and could also be confirmed by radiologists. Conclusions This study proposed a mask-weighted GAP-based deep learning method and obtained promising results for COVID-19 triage based on chest CT images. Furthermore, it can be considered a convenient tool to assist doctors in diagnosing COVID-19.

Computed tomography-based COVID-19 triage through a deep neural network using mask-weighted global average pooling.

Background: There is an urgent need to find an effective and accurate method for triaging coronavirus disease 2019 (COVID-19) patients from millions or billions of people. Therefore, this study aimed to develop a novel deep-learning approach for COVID-19 triage based on chest computed tomography (CT) images, including normal, pneumonia, and COVID-19 cases. Methods: A total of 2,809 chest CT scans (1,105 COVID-19, 854 normal, and 850 non-3COVID-19 pneumonia cases) were acquired for this study and classified into the training set (n = 2,329) and test set (n = 480). A U-net-based convolutional neural network was used for lung segmentation, and a mask-weighted global average pooling (GAP) method was proposed for the deep neural network to improve the performance of COVID-19 classification between COVID-19 and normal or common pneumonia cases. Results: The results for lung segmentation reached a dice value of 96.5% on 30 independent CT scans. The performance of the mask-weighted GAP method achieved the COVID-19 triage with a sensitivity of 96.5% and specificity of 87.8% using the testing dataset. The mask-weighted GAP method demonstrated 0.9% and 2% improvements in sensitivity and specificity, respectively, compared with the normal GAP. In addition, fusion images between the CT images and the highlighted area from the deep learning model using the Grad-CAM method, indicating the lesion region detected using the deep learning method, were drawn and could also be confirmed by radiologists. Conclusions: This study proposed a mask-weighted GAP-based deep learning method and obtained promising results for COVID-19 triage based on chest CT images. Furthermore, it can be considered a convenient tool to assist doctors in diagnosing COVID-19.

COVID-19 detection based on self-supervised transfer learning using chest X-ray images.

PURPOSE: Considering several patients screened due to COVID-19 pandemic, computer-aided detection has strong potential in assisting clinical workflow efficiency and reducing the incidence of infections among radiologists and healthcare providers. Since many confirmed COVID-19 cases present radiological findings of pneumonia, radiologic examinations can be useful for fast detection. Therefore, chest radiography can be used to fast screen COVID-19 during the patient triage, thereby determining the priority of patient's care to help saturated medical facilities in a pandemic situation. METHODS: In this paper, we propose a new learning scheme called self-supervised transfer learning for detecting COVID-19 from chest X-ray (CXR) images. We compared six self-supervised learning (SSL) methods (Cross, BYOL, SimSiam, SimCLR, PIRL-jigsaw, and PIRL-rotation) with the proposed method. Additionally, we compared six pretrained DCNNs (ResNet18, ResNet50, ResNet101, CheXNet, DenseNet201, and InceptionV3) with the proposed method. We provide quantitative evaluation on the largest open COVID-19 CXR dataset and qualitative results for visual inspection. RESULTS: Our method achieved a harmonic mean (HM) score of 0.985, AUC of 0.999, and four-class accuracy of 0.953. We also used the visualization technique Grad-CAM++ to generate visual explanations of different classes of CXR images with the proposed method to increase the interpretability. CONCLUSIONS: Our method shows that the knowledge learned from natural images using transfer learning is beneficial for SSL of the CXR images and boosts the performance of representation learning for COVID-19 detection. Our method promises to reduce the incidence of infections among radiologists and healthcare providers.

Boosting automatic COVID-19 detection performance with self-supervised learning and batch knowledge ensembling.

PROBLEM: Detecting COVID-19 from chest X-ray (CXR) images has become one of the fastest and easiest methods for detecting COVID-19. However, the existing methods usually use supervised transfer learning from natural images as a pretraining process. These methods do not consider the unique features of COVID-19 and the similar features between COVID-19 and other pneumonia. AIM: In this paper, we want to design a novel high-accuracy COVID-19 detection method that uses CXR images, which can consider the unique features of COVID-19 and the similar features between COVID-19 and other pneumonia. METHODS: Our method consists of two phases. One is self-supervised learning-based pertaining; the other is batch knowledge ensembling-based fine-tuning. Self-supervised learning-based pretraining can learn distinguished representations from CXR images without manually annotated labels. On the other hand, batch knowledge ensembling-based fine-tuning can utilize category knowledge of images in a batch according to their visual feature similarities to improve detection performance. Unlike our previous implementation, we introduce batch knowledge ensembling into the fine-tuning phase, reducing the memory used in self-supervised learning and improving COVID-19 detection accuracy. RESULTS: On two public COVID-19 CXR datasets, namely, a large dataset and an unbalanced dataset, our method exhibited promising COVID-19 detection performance. Our method maintains high detection accuracy even when annotated CXR training images are reduced significantly (e.g., using only 10% of the original dataset). In addition, our method is insensitive to changes in hyperparameters. CONCLUSION: The proposed method outperforms other state-of-the-art COVID-19 detection methods in different settings. Our method can reduce the workloads of healthcare providers and radiologists.

The rapid and efficient strategy for SARS-CoV-2 Omicron transmission control: analysis of outbreaks at the city level.

BACKGROUND: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron (B.1.1.529) variant is highly transmissible with potential immune escape. Hence, control measures are continuously being optimized to guard against large-scale coronavirus disease 2019 (COVID-19) outbreaks. This study aimed to explore the relationship between the intensity of control measures in response to different SARS-CoV-2 variants and the degree of outbreak control at city level. METHODS: A retrospective study was conducted in 49 cities with COVID-19 outbreaks between January 2020 and June 2022. Epidemiological data on COVID-19 were extracted from the National Health Commission, People's Republic of China, and the population flow data were sourced from the Baidu migration data provided by the Baidu platform. Outbreak control was quantified by calculating the degree of infection growth and the time-varying reproduction number ([Formula: see text]). The intensity of the outbreak response was quantified by calculating the reduction in population mobility during the outbreak period. Correlation and regression analyses of the intensity of the control measures and the degree of outbreak control for the Omicron variant and non-Omicron mutants were conducted, respectively. RESULTS: Overall, 65 outbreaks occurred in 49 cities in China from January 2020 to June 2022. Of them, 66.2% were Omicron outbreaks and 33.8% were non-Omicron outbreaks. The intensity of the control measures was positively correlated with the degree of outbreak control (r = 0.351, P = 0.03). The degree of reduction in population mobility was negatively correlated with the Rt value (r = - 0.612, P < 0.01). Therefore, under the same control measure intensity, the number of new daily Omicron infections was 6.04 times higher than those attributed to non-Omicron variants, and the Rt value of Omicron outbreaks was 2.6 times higher than that of non-Omicron variants. In addition, the duration of non-Omicron variant outbreaks was shorter than that of the outbreaks caused by the Omicron variant (23.0 ± 10.7, 32.9 ± 16.3, t = 2.243, P = 0.031). CONCLUSIONS: Greater intensity of control measures was associated with more effective outbreak control. Thus, in response to the Omicron variant, the management to restrict population movement should be used to control its spread quickly, especially in the case of community transmission occurs widely. Faster than is needed for non-Omicron variants, and decisive control measures should be imposed and dynamically adjusted in accordance with the evolving epidemic situation.

Prevalence of antibiotic resistance pathogens in online fresh-cut fruit from Chongqing, China and controlling Enterococcus faecalis by bacteriocin GF-15

Consuming fresh-cut fruits from online platform has become a common lifestyle in China, especially under the background of COVID-2019 pandemic. To investigate microbial safety of online fresh-cut fruits, pathogens were isolated and their antibiotic resistance was characterized. Results showed that Enterococcus faecalis, Enterococcus faecium, Morganella morganii, Klebsiella pneumoniae, Klebsiella aerogenes, Klebsiella variicola, Proteus mirabilis, Stenotrophomonas maltophilia, Microbacterium paraoxydans, and Achromobacter xylosoxidans were isolated from online fresh-cut fruits. These pathogens showed resistance to tetracyclines, aminoglycosides, beta-lactams, quinolones, macrolides, etc. After genome sequencing, antibiotic resistance genes lsa(A), tet(M), aph(3′)-IIc, aac(6′)-Ii, qnrD1, OqxA, bla(CMY), msr(C), sul2, dfrA14, mdf(A), etc. were identified in these pathogens. A novel bacteriocin GF-15 was used to control E. faecalis which occupied the highest abundance in online fresh-cut fruits. MIC values of GF-15 to E. faecalis strains were 8–16 μM. The growth of E. faecalis can be totally inhibited by ≥ 4 × MIC GF-15 treatment. GF-15 had bactericidal mode and damaged cell envelope of E. faecalis. Tracing by FITC labelled GF-15 showed that it located both on cell envelope and in cytoplasm. On cell envelope, GF-15 led to great pore-formation and strong membrane depolarization. After entering cytoplasm, it induced reactive oxygen species production.

Potential urinary biomarkers in young adults with short-term exposure to particulate matter and bioaerosols identified using an unbiased metabolomic approach.

Numerous epidemiological studies have shown a close relationship between outdoor air pollution and increased risks for cancer, infection, and cardiopulmonary diseases. However, very few studies have investigated the potential health effects of coexposure to airborne particulate matter (PM) and bioaerosols through the transmission of infectious agents, particularly under the current circumstances of the coronavirus disease 2019 pandemic. In this study, we aimed to identify urinary metabolite biomarkers that might serve as clinically predictive or diagnostic standards for relevant diseases in a real-time manner. We performed an unbiased gas/liquid chromatography-mass spectroscopy (GC/LC-MS) approach to detect urinary metabolites in 92 samples from young healthy individuals collected at three different time points after exposure to clean air, polluted ambient, or purified air, as well as two additional time points after air repollution or repurification. Subsequently, we compared the metabolomic profiles between the two time points using an integrated analysis, along with Kyoto Encyclopedia of Genes and Genomes-enriched pathway and time-series analysis. We identified 33 and 155 differential metabolites (DMs) associated with PM and bioaerosol exposure using GC/LC-MS and follow-up analyses, respectively. Our findings suggest that 16-dehydroprogesterone and 4-hydroxyphenylethanol in urine samples may serve as potential biomarkers to predict or diagnose PM- or bioaerosol-related diseases, respectively. The results indicated apparent differences between PM- and bioaerosol-associated DMs at five different time points and revealed dynamic alterations in the urinary metabolic profiles of young healthy humans with cyclic exposure to clean and polluted air environments. Our findings will help in investigating the detrimental health effects of short-term coexposure to airborne PM and bioaerosols in a real-time manner and improve clinically predictive or diagnostic strategies for preventing air pollution-related diseases.

Dynamic changes in liver function parameters in patients with coronavirus disease 2019: a multicentre, retrospective study.

BACKGROUND: Liver injuries have been reported in patients with coronavirus disease 2019 (COVID-19). This study aimed to investigate the clinical role played by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). METHODS: In this multicentre, retrospective study, the parameters of liver function tests in COVID-19 inpatients were compared between various time-points in reference to SARS-CoV-2 shedding, and 3 to 7 days before the first detection of viral shedding was regarded as the reference baseline. RESULTS: In total, 70 COVID-19 inpatients were enrolled. Twenty-two (31.4%) patients had a self-medication history after illness. At baseline, 10 (14.3%), 7 (10%), 9 (12.9%), 2 (2.9%), 15 (21.4%), and 4 (5.7%) patients already had abnormal alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP), albumin, and total bilirubin (TBIL) values, respectively. ALT and AST abnormal rates and levels did not show any significant dynamic changes during the full period of viral shedding (all p > 0.05). The GGT abnormal rate (p = 0.008) and level (p = 0.033) significantly increased on day 10 of viral shedding. Meanwhile, no simultaneous significant increases in abnormal ALP rates and levels were observed. TBIL abnormal rates and levels significantly increased on days 1 and 5 of viral shedding (all p < 0.05). Albumin abnormal decrease rates increased, and levels decreased consistently from baseline to SARS-CoV-2 clearance day (all p < 0.05). Thirteen (18.6%) patients had chronic liver disease, two of whom died. The ALT and AST abnormal rates and levels did not increase in patients with chronic liver disease during SARS-CoV-2 shedding. CONCLUSIONS: SARS-CoV-2 does not directly lead to elevations in ALT and AST but may result in elevations in GGT and TBIL; albumin decreased extraordinarily even when SARS-CoV-2 shedding ended.

Remarkable gastrointestinal and liver manifestations of COVID-19: A clinical and radiologic overview.

The coronavirus disease 2019 (COVID-19) raging around the world still has not been effectively controlled in most countries and regions. As a severe acute respiratory syndrome coronavirus, in addition to the most common infectious pneumonia, it can also cause digestive system disease such as diarrhea, nausea, vomiting, liver function damage, etc. In medical imaging, it manifests as thickening of the intestinal wall, intestinal perforation, pneumoperitoneum, ascites and decreased liver density. Angiotensin-converting enzyme 2 has great significance in COVID-19-related digestive tract diseases. In this review, we summarized the data on the clinical and imaging manifestations of gastrointestinal and liver injury caused by COVID-19 so far and explored its possible pathogenesis.

First known case of successful pressure ulcer treatment in a lung transplant patient with post-COVID-19 pneumonia.

Given the current COVID-19 crisis, multiple clinical manifestations and related complications of COVID-19 disease, especially in lung transplant patients following post-COVID-19 pneumonia, are a major challenge. Herein, we report the therapeutic course of the first reported case of sacrococcyx pressure ulcers (PU) in a 65-year-old male COVID-19 patient who underwent lung transplantation and developed a PU following surgery. We used a combination of regulated negative pressure-assisted wound therapy system (RNPT, six treatment courses, five days per treatment course), a skin tension-relief system (an intraoperative aid in minimising wounds caused by sacrococcygeal PUs) and a gluteus maximus myocutaneous flap to repair sacrococcygeal wounds. This successfully treated case provides a reference point for the treatment of similar cases.

Prone position combined with high-flow nasal oxygen could benefit spontaneously breathing, severe COVID-19 patients: A case report.

BACKGROUND: Since the outbreak of coronavirus disease 2019 (COVID-19) in Wuhan, China in December 2019, the overall fatality rate of severe and critical patients with COVID-19 is high and the effective therapy is limited. CASE SUMMARY: In this case report, we describe a case of the successful combination of the prone position (PP) and high-flow nasal oxygen (HFNO) therapy in a spontaneously breathing, severe COVID-19 patient who presented with fever, fatigue and hypoxemia and was diagnosed by positive throat swab COVID-19 RNA testing. The therapy significantly improved the patient's clinical symptoms, oxygenation status, and radiological characteristics of lung injury during hospitalization, and the patient showed good tolerance and avoided intubation. Additionally, we did not find that medical staff wearing optimal airborne personal protective equipment (PPE) were infected by the new coronavirus in our institution. CONCLUSION: We conclude that the combination of PP and HFNO could benefit spontaneously breathing, severe COVID-19 patients. The therapy does not increase risk of healthcare workers wearing optimal airborne PPE to become infected with virus particles.

Expression and purification of recombinant SARS-CoV-2 nucleocapsid protein in inclusion bodies and its application in serological detection.

The current standard for the diagnosis of COVID-19 is the nucleic acid test of SARS-CoV-2 RNA, however, virus antibody detection has the advantages of convenient sample collection, high throughout, and low cost. When combining detection with nucleic acid detection, antibody detection can effectively compensate for nucleic acid detection. Virus infection always induce high antibody titer against SARS-CoV-2 nucleocapsid protein (N protein), which can be used to detect COVID-19 at both infected and convalescent patients. In this study we reported the expression and purification of N protein in E.coli from inclusion bodies by a combination of two cation exchange chromatography, and the yield of N protein was around 50 mg/L fermentation broth with more than 90% purity. A corresponding colloidal gold detection kit prepared with our purified N protein was used to verify the efficiency and accuracy our N protein in antibody detection method. Of the 58 COVID-19 PCR positive patients' inactivated serum samples, 40 samples were IgM positive (69.0%), and 42 samples were IgG positive (72.4%), and all 95 COVID-19 negative patients' inactivated serum samples were both IgM and IgG negative. Our results indicates that the refolded soluble N protein could be used for the preliminary detection of IgG and IgM antibodies against SARS-CoV- 2.

Efficacy of the combination of modern medicine and traditional Chinese medicine in pulmonary fibrosis arising as a sequelae in convalescent COVID-19 patients: a randomized multicenter trial.

BACKGROUND: The coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has led to a significant number of mortalities worldwide. COVID-19 poses a serious threat to human life. The clinical manifestations of COVID-19 are diverse and severe and 20% of infected patients are reported to be in a critical condition. A loss in lung function and pulmonary fibrosis are the main manifestations of patients with the severe form of the disease. The lung function is affected, even after recovery, thereby greatly affecting the psychology and well-being of patients, and significantly reducing their quality of life. METHODS: Participants must meet the following simultaneous inclusion criteria: over 18 years of age, should have recovered from severe or critical COVID-19 cases, should exhibit pulmonary fibrosis after recovery, and should exhibit Qi-Yin deficiency syndrome as indicated in the system of traditional Chinese medicine (TCM). The eligible candidates will be randomized into treatment or control groups. The treatment group will receive modern medicine (pirfenidone) plus TCM whereas the control group will be administered modern medicine plus TCM placebo. The lung function index will be continuously surveyed and recorded. By comparing the treatment effect between the two groups, the study intend to explore whether TCM can improve the effectiveness of modern medicine in patients with pulmonary fibrosis arising as a sequelae after SARS-CoV-2 infection. DISCUSSION: Pulmonary fibrosis is one of fatal sequelae for some severe or critical COVID-19 cases, some studies reveal that pirfenidone lead to a delay in the decline of forced expiratory vital capacity, thereby reducing the mortality partly. Additionally, although TCM has been proven to be efficacious in treating pulmonary fibrosis, its role in treating pulmonary fibrosis related COVID-19 has not been explored. Hence, a multicenter, parallel-group, randomized controlled, interventional, prospective clinical trial has been designed and will be conducted to determine if a new comprehensive treatment for pulmonary fibrosis related to COVID-19 is feasible and if it can improve the quality of life of patients. TRIAL REGISTRATION: This multicenter, parallel-group, randomized controlled, interventional, prospective trial was registered at the Chinese Clinical Trial Registry (ChiCTR2000033284) on 26th May 2020 (prospective registered).

A five-layer deep convolutional neural network with stochastic pooling for chest CT-based COVID-19 diagnosis.

Till August 17, 2020, COVID-19 has caused 21.59 million confirmed cases in more than 227 countries and territories, and 26 naval ships. Chest CT is an effective way to detect COVID-19. This study proposed a novel deep learning model that can diagnose COVID-19 on chest CT more accurately and swiftly. Based on traditional deep convolutional neural network (DCNN) model, we proposed three improvements: (i) We introduced stochastic pooling to replace average pooling and max pooling; (ii) We combined conv layer with batch normalization layer and obtained the conv block (CB); (iii) We combined dropout layer with fully connected layer and obtained the fully connected block (FCB). Our algorithm achieved a sensitivity of 93.28% ± 1.50%, a specificity of 94.00% ± 1.56%, and an accuracy of 93.64% ± 1.42%, in identifying COVID-19 from normal subjects. We proved using stochastic pooling yields better performance than average pooling and max pooling. We compared different structure configurations and proved our 3CB + 2FCB yields the best performance. The proposed model is effective in detecting COVID-19 based on chest CT images.

Clinical outcomes of COVID-19 in Wuhan, China: a large cohort study.

BACKGROUND: Since December 2019, an outbreak of Coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-Cov-2) initially emerged in Wuhan, China, and has spread worldwide now. Clinical features of patients with COVID-19 have been described. However, risk factors leading to in-hospital deterioration and poor prognosis in COVID-19 patients with severe disease have not been well identified. METHODS: In this retrospective, single-center cohort study, 1190 adult inpatients (≥ 18 years old) with laboratory-confirmed COVID-19 and determined outcomes (discharged or died) were included from Wuhan Infectious Disease Hospital from December 29, 2019 to February 28, 2020. The final follow-up date was March 2, 2020. Clinical data including characteristics, laboratory and imaging information as well as treatments were extracted from electronic medical records and compared. A multivariable logistic regression model was used to explore the potential predictors associated with in-hospital deterioration and death. RESULTS: 1190 patients with confirmed COVID-19 were included. Their median age was 57 years (interquartile range 47-67 years). Two hundred and sixty-one patients (22%) developed a severe illness after admission. Multivariable logistic regression demonstrated that higher SOFA score (OR 1.32, 95% CI 1.22-1.43, per score increase, p < 0.001 for deterioration and OR 1.30, 95% CI 1.11-1.53, per score increase, p = 0.001 for death), lymphocytopenia (OR 1.81, 95% CI 1.13-2.89 p = 0.013 for deterioration; OR 4.44, 95% CI 1.26-15.87, p = 0.021 for death) on admission were independent risk factors for in-hospital deterioration from not severe to severe disease and for death in severe patients. On admission D-dimer greater than 1 µg/L (OR 3.28, 95% CI 1.19-9.04, p = 0.021), leukocytopenia (OR 5.10, 95% CI 1.25-20.78), thrombocytopenia (OR 8.37, 95% CI 2.04-34.44) and history of diabetes (OR 11.16, 95% CI 1.87-66.57, p = 0.008) were also associated with higher risks of in-hospital death in severe COVID-19 patients. Shorter time interval from illness onset to non-invasive mechanical ventilation in the survivors with severe disease was observed compared with non-survivors (10.5 days, IQR 9.25-11.0 vs. 16.0 days, IQR 11.0-19.0 days, p = 0.030). Treatment with glucocorticoids increased the risk of progression from not severe to severe disease (OR 3.79, 95% CI 2.39-6.01, p < 0.001). Administration of antiviral drugs especially oseltamivir or ganciclovir is associated with a decreased risk of death in severe patients (OR 0.17, 95% CI 0.05-0.64, p < 0.001). CONCLUSIONS: High SOFA score and lymphocytopenia on admission could predict that not severe patients would develop severe disease in-hospital. On admission elevated D-dimer, leukocytopenia, thrombocytopenia and diabetes were independent risk factors of in-hospital death in severe patients with COVID-19. Administration of oseltamivir or ganciclovir might be beneficial for reducing mortality in severe patients.