The clinical and virological features of two children's coinfections with human adenovirus type 7 and human coronavirus-229E virus

Objective Human adenovirus (HAdV) coinfection with other respiratory viruses is common, but adenovirus infection combined with human coronavirus-229E (HCoV-229E) is very rare. Study design and setting Clinical manifestations, laboratory examinations, and disease severity were compared between three groups: one coinfected with HAdV-Ad7 and HCoV-229E, one infected only with adenovirus (mono-adenovirus), and one infected only with HCoV-229E (mono-HCoV-229E). Results From July to August 2019, there were 24 hospitalized children: two were coinfected with HAdV-Ad7 and HCoV-229E, and 21 were infected with a single adenovirus infection. Finally, one 14-year-old boy presented with a high fever, but tested negative for HAdV-Ad7 and HCoV-229E. Additionally, three adult asymptotic cases with HCoV-229E were screened. No significant difference in age was found in the coinfection and mono-adenovirus groups (11 vs. 8 years, p = 0.332). Both groups had the same incubation period (2.5 vs. 3 days, p = 0.8302), fever duration (2.5 vs. 2.9 days, p = 0.5062), and length of hospital stay (7 vs. 6.76 days, p = 0.640). No obvious differences were found in viral loads between the coinfection and mono-adenovirus groups (25.4 vs. 23.7, p = 0.570), or in the coinfection and mono-HCoV-229E groups (32.9 vs. 30.06, p = 0.067). All cases recovered and were discharged from the hospital. Conclusion HAdV-Ad7 and HCoV-229E coinfection in healthy children may not increase the clinical severity or prolong the clinical course. The specific interaction mechanism between the viruses requires further study.

A New Feature Based Deep Attention Sales Forecasting Model for Enterprise Sustainable Development

In recent years, with the rise of the Internet, e-commerce has become an important field of commodity sales. However, e-commerce is affected by many factors, and the wrong judgment of supply and marketing relationships will bring huge losses to operators. Therefore, it is of great significance to establish a model that can effectively achieve high precision sales prediction for ensuring the sustainable development of e-commerce enterprises. In this paper, we propose an e-commerce sales forecasting model that considers the features of many aspects of correlation. In the first layer of the model, the temporal convolutional network (TCN) is used to extract the deep temporal characteristics of univariate sales historical data, which ensures the integrity of temporal information of sales characteristics. In the second layer, the feature selection method based on reinforcement learning is used to filter the effective correlation feature set and combine it with the temporal feature after processing, which not only improves the amount of effective information input by the model, but also avoids the high feature dimension. The third layer of the reformer model learns all the features and pays different attention to the features with different degrees of importance, ensuring the stability of the sales forecast. In the experimental part, we compare the proposed model with the current advanced sales forecasting model, and we can find that the proposed model has higher stability and accuracy.

Effectiveness of inactivated and Ad5-nCoV COVID-19 vaccines against SARS-CoV-2 Omicron BA. 2 variant infection, severe illness, and death (preprint)

ABSTRACT Background Limited data are available on effectiveness of inactivated and Ad5-nCoV COVID-19 vaccines in real-world use - especially against Omicron variants in SARS-CoV-2 infection-naïve population. During an outbreak in Shanghai’s SARS-CoV-2 infection-naïve population, we evaluated vaccine effectiveness (VE) against Omicron infection, severe or critical COVID-19, and COVID-19-related death. Methods A matched case-control study was conducted among people aged ≥3 years between 2 December 2021 through 13 May 2022. Cases were SARS-CoV-2 infected individuals, individuals with severe/critical COVID-19, or COVID-19-related deaths. Controls were selected from consecutively test negative individuals at the same time as cases were diagnosed and were exact-matched on year-of-age, gender, birthplace, illness onset date, and residency district in ratios of 1:1 with infected individuals and 4:1 with severe/critical COVID-19 and COVID-19-related deaths. Results Our study included 612597 documented SARS-CoV-2 infections, among which 1485 progressed to severe or critical illness and 568 died. Inactivated vaccine was 16.3% (95% CI: 15.4%-17.2%) effective against infection, 88.6% (95% CI: 85.8%-90.9%) effective against severe/critical COVIID-19 and 91.7% (95% CI: 86.9%-94.7%) against COVID-19 death. Ad5-vectored vaccine was 13.2% (95% CI: 10.9%-15.5%) effective against infection and 77.9% (95% CI: 15.6%-94.2%) effective against severe/critical COVIID-19. Booster vaccination with inactivated vaccines enhanced protection against severe COVID-19 (92.7%, 95% CI: 90.1%-94.6%) and COVID-19 death (95.9%, 95% CI: 91.4%-98.1%). Inactivated VE against infection began to wane 12 weeks after the last dose but two- and three-dose sustained high protection levels (>80%) against severe/critical illness and death. Conclusions Our real-world study found high and durable two- and three-dose inactivated VE against Omicron-associated severe/critical illness and death across all age groups, but lower effectiveness against Omicron infection. High direct protection from severe/fatal Omicron COVID-19 provided by inactivated vaccines, and a consequent potential reduction in health-care utilization, reinforces the critical importance of full-series vaccination and timely booster dose administration for all eligible individuals.

Computational simulations reveal the binding dynamics between human ACE2 and the receptor binding domain of SARS-CoV-2 spike protein

Background: A novel coronavirus (the SARS-CoV-2) has been identified in January 2020 as the causal pathogen for COVID-19 , a pandemic started near the end of 2019. The Angiotensin converting enzyme 2 protein (ACE2) utilized by the SARS-CoV as a receptor was found to facilitate the infection of SARS-CoV-2, initiated by the binding of the spike protein to human ACE2. Methods: Using homology modeling and molecular dynamics (MD) simulation methods, we report here the detailed structure and dynamics of the ACE2 in complex with the receptor binding domain (RBD) of the SARS-CoV-2 spike protein. Results: The predicted model is highly consistent with the experimentally determined structures, validating the homology modeling results. Besides the binding interface reported in the crystal structures, novel binding poses are revealed from all-atom MD simulations. The simulation data are used to identify critical residues at the complex interface and provide more details about the interactions between the SARS-CoV-2 RBD and human ACE2. Conclusion: Simulations reveal that RBD binds to both open and closed state of ACE2. Two human ACE2 mutants and rat ACE2 are modeled to study the mutation effects on RBD binding to ACE2. The simulations show that the N-terminal helix and the K353 are very important for the tight binding of the complex, the mutants are found to alter the binding modes of the CoV2-RBD to ACE2.

Therapeutic of four antiviral drugs in treatment of COVID-19: A protocol for systematic review and Network Meta-analysis (preprint)

Introduction The COVID-19 pandemic has resulted in a global crisis in public health, however, there are still no safe and effective drugs to resist COVID-19 until now. We will use network meta-analysis to analysis available evidence from RCTs to compare the safety and efficacy of four antiviral drugs (including Ribavirin, Arbidol, Chloroquine Phosphate and Interferon) alone or in combination, in patients with COVID-19 on the basis of standard treatment, to reveal the robustness and strength of evidence for relative efficacy against COVID-19, which will provide better evidence for future clinical decision-making. Methods Using English and Chinese search strategies to search 8 databases including PubMed, Web of Science, Embase, the Cochrane Library, CNKI, CBM, WANFANG Database and VIP. In addition, manual search for references in publications has been supplemented by electronic search. To enhance the effectiveness of this study, only randomized controlled trials of four antiviral drugs (Ribavirin, Arbidol, Chloroquine Phosphate, Interferon) used alone or in combination with the primary therapy shall be included. Analysis The nucleic acid turning negative, complete absorption of lung inflammation, adverse reactions, aggravation and death shall be the primary outcome measures; whereas temperature return to normal, hospitalization, and positive rate after discharge will be the secondary outcomes. To ensure the quality of the systematic evaluation of this study, study screening, data extraction and quality evaluation will be carried out independently by two reviewers, and any differences will be resolved through consultation between them or by a third reviewer. Ethics and dissemination This systematic review will evaluate the efficacy of four antiviral drugs (Ribavirin, Arbidol, Chloroquine Phosphate and Interferon) for Covid-19 in adults. Since all included data will be obtained from published articles, it does not require ethical approval and will be published in a peer-reviewed journal. PROSPERO registration number: CRD42022300104.

SARS‐CoV‐2 Omicron variant: Immune escape and vaccine development

New genetic variants of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) constantly emerge through unmitigated spread of the virus in the ongoing Coronavirus disease 2019 pandemic. Omicron (B.1.1.529), the latest variant of concern (VOC), has so far shown exceptional spread and infectivity and has established itself as the dominant variant in recent months. The SARS‐CoV‐2 spike glycoprotein is a key component for the recognition and binding to host cell angiotensin‐converting enzyme 2 receptors. The Omicron variant harbors a cluster of substitutions/deletions/insertions, and more than 30 mutations are located in spike. Some noticeable mutations, including K417N, T478K, N501Y, and P681H, are shared with the previous VOCs Alpha, Beta, Gamma, or Delta variants and have been proven to be associated with higher transmissibility, viral infectivity, and immune evasion potential. Studies have revealed that the Omicron variant is partially resistant to the neutralizing activity of therapeutic antibodies and convalescent sera, which poses significant challenges for the clinical effectiveness of the current vaccines and therapeutic antibodies. We provide a comprehensive analysis and summary of the epidemiology and immune escape mechanisms of the Omicron variant. We also suggest some therapeutic strategies against the Omicron variant. This review, therefore, aims to provide information for further research efforts to prevent and contain the impact of new VOCs during the ongoing pandemic. Omicron (B.1.1.529), the latest variant of concern, is partially resistant to the neutralizing activity of therapeutic antibodies and convalescent sera, which poses significant challenges for the clinical effectiveness of the current vaccines and therapeutic antibodies. We provide a comprehensive analysis and summary of the epidemiology and immune escape mechanisms of the Omicron variant. We also suggest some therapeutic strategies against the Omicron variant.

Long-Term Outcomes of COVID-19 Survivors and an Exploratory Analysis of Factors Associated with Sequela-Free Survival of Patients Treated at Leishenshan Hospital in Wuhan, China (preprint)

Objective: To determine key factors associated with sequela-free survival of patients with COVID-19.Design: Retrospective–prospective observational study.Setting: Leishenshan Hospital in Wuhan, China.Participants: Patients with COVID-19 admitted at Leishenshan Hospital in Wuhan, China between 15 February 2020 and 1 April 2020 and discharged.Main Outcomes: MeasuresSequela-free survival is defined as having none of the long-term sequelae measured with the Activity of Daily Living (ADL) Scale, Modified Medical Research Council (mMRC) Dyspnoea Scale, New York Heart Association (NYHA) classification, Self-rating Depression Scale (SDS), Carcinologic Handicap Index (CHI) and Modified Telephone Interview for Cognitive Status (TICS-M). Multivariable logistic regression was used to examine the key factors of sequela-free survival.ResultsThe follow-up period for the 464 patients was 10 August to 30 September 2020. Of these, 424 patients completed the assessment of all scales. 30.2% (128 of 424) were categorised as sequela-free at follow-up. The most common sequelae were psychosocial problems (254 [57.7%]), respiratory function abnormality (149 [32.6%]) and cardiac function abnormality (98 [21.5%]). Risk factors associated with COVID-19 sequelae were anaemia on admission, longer duration from the onset of symptoms to admission and advanced age, whereas anti-anaemia treatment was a protective factor against sequelae. A haemoglobin level of <113 g/l for female patients or <92 g/l for male patients at admission discriminated a higher risk of any COVID-19 sequelae.Conclusions: Only one third COVID-19 survivors were sequela-free. Anaemia on admission, longer duration from the onset of symptoms to admission and advanced age were the risk factors of any long-term sequela. Anti-anaemia treatment was associated with beneficial outcomes in COVID-19 survivors.Funding: This study was funded by grants to DSS from the National Natural Science Foundation of China (Nos.81771133, 81970995) (DSD grants), Shanghai Shenkang Hospital Development Center Founding (SHDC12017X11), Renji Hospital Clinical Innovation Foundation (PYMDT-007), Shanghai municipal Education Commission-Gaofeng Clinical Medicine Support (20191903), State Key Laboratory of Neuroscience(SKLN－201803), the National Natural Science Foundation of China (No.81701358) (DH grants), the Shanghai Municipal Commission of Health and Family Planning Funding (20184Y0205) (CYC grants) and Renji Hospital founding (RJZZ18-019) (LLH grants).Declaration of Interest: None to declare. Ethical Approval: This study was approved by the Institutional Research Ethics Committee of Renji Hospital, School of Medicine, Shanghai Jiaotong University (ethics committee approval No. KY2020-34).

Investigation and precautions of SARS-CoV-2 infection among healthcare workers in Emergency Center

Objective: To analyze the causes of SARS-CoV-2 nosocomial infection among healthcare workers (HCWs) and explore the effective precaution strategies in Emergency Center.

Willingness of Medical Undergraduates to Work as Volunteers for anti-epidemic of COVID-19: Reflections on Medical Education in the post-epidemic era of China (preprint)

Objective: To identify the willingness of medical undergraduates to work as volunteers for fighting COVID-19 and effect factors of willingness to volunteer. Methods: A cross-sectional online questionnaire survey which consisted of three sections with 26 items was conducted. The primary outcome was the willingness to volunteer. Descriptive analysis was used for the results of characteristics, a univariant analysis was performed by chi-square test and a multivariate analysis was performed by multivariate logistic regression. Results: A total of 5499 questionnaires were issued and 5379 effective questionnaires were finally recovered. There were 1797 males (33.4%) and 3582 females (66.6%), with an average age of 20±1.5 years old. 3553(66.1%) students wanted to volunteer. Male was associated with less adjusted OR (0.68, 0.56 to 0.81) for willingness. The undergraduates in school of public health was most (3.81, 2.26 to 6.40) willing to work. The undergraduates who had often participated in voluntary activities but not regularly was 3.23(2.03-5.16) times more willing than those that never participate. Conclusion: It suggested that medical educators should emphasize voluntary activities in undergraduate medical education and continue to insistence of medical humanities and public health education to improve the social responsibility and professional identity of medical undergraduates.

Imaging features of COVID-19: What we can learn from SARS and MERS (Review)

Coronavirus disease 2019 (COVID-19) is a highly infectious type of pneumonia caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that has rapidly become a global pandemic. COVID-19, SARS and Middle East respiratory syndrome (MERS) are all caused by members of the Coronaviridae family. As expected, emerging genetic and clinical evidence from patients with COVID-19 has indicated that the pathway of infection is similar to that of SARS and MERS. Additionally, much like SARS and MERS, chest imaging serves an important role in the diagnosis, management and follow-up of patients with COVID-19. Although these related viruses present a similar pneumonic pathogenesis, the imaging results have distinguishable features. The current review evaluated the imaging results of patients with SARS and MERS and explored the potential similarities and differences among patients with COVID-19, SARS and MERS at early, progressive, severe and recovery stages, with the aim of improving our understanding of SARS-CoV-2 infections by comparing the features of COVID-19 images with those of SARS and MERS. The current review assessed whether imaging results had implications for the administration of corticosteroids as treatment for COVID-19. Whether corticosteroids can inhibit inflammatory cytokine storms and reduce the mortality of patients with viral pneumonia remains controversial. However, his review may help radiologists and clinicians to identify viral pneumonia and guide appropriate COVID-19 treatment. [ABSTRACT FROM AUTHOR] Copyright of Experimental & Therapeutic Medicine is the property of Spandidos Publications UK Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Administration of Statins Can Improve Survival and Coagulation Function in Patients with Severe COVID-19 in Intensive Care Units: A Retrospective Study (preprint)

Background: COVID-19 is a global epidemic posing threat to public health. Without specific drugs and therapies, it is important to evaluate the effectiveness of currently applicable drugs.This study was a retrospective analysis of the effect of the administration of statins on survival in patients with severe COVID-19.Design: This study was a retrospective analysis of data collected in the ICU of Tongji Hospital affiliated with Tongji Medical College of Huazhong University of Science and Technology (HUST) during the COVID-19 outbreak.Data from69 patients with severe COVID-19 treated in the ICU from February to March 2020 were collected for the analysis.Patients with severe COVID-19 were treated with the standard of care at the time, and some patients were also treated with statins.Results: Sixty-four patients with complete records were enrolled in the final stage of analysis. Statin administration had a beneficial effect on the overall survival of the patients. The statin administration cohorthad a significantly shorter activated partial thromboplastin time (aPTT) and prothrombin time (PT) according to t-tests. The aPTT and PT were also stable, as shown by locally weighted smoothing (LOESS) analysis. The use of statins combined with corticosteroids, anti-coagulants and anti-hypertensive drugs had a relatively greater effect on survival.Conclusions: Statins can be therapeutic agents that may promote patient survival, possibly by improving coagulation function. Trial registration: HIRB 2020-1119 September 11th, 2020, retrospectively registered.

Long-Term Outcomes and Sequelae for 464 COVID-19 Patients Discharged from Leishenshan Hospital in Wuhan, China (preprint)

Background: While coronavirus disease 2019 (COVID-19) has emerged as a global pandemic, millions of COVID-19 patients have recovered and returned to their families and work, although the long-term outcomes remain unknown.Methods: This retrospective–prospective study analysed data for COVID-19 patients discharged from Leishenshan Hospital in Wuhan, China. Long-term outcomes were measured by Activity of Daily Living (ADL) Scale, Modified Medical Research Council (mMRC) Dyspnoea Scale, New York Heart Association (NYHA) classification, Self-rating Depression Scale (SDS), Carcinologic Handicap Index (CHI), and Modified Telephone Interview for Cognitive Status (TICS-M). Multivariable logistic regression was used to examine the risk factors of long-term outcomes.Findings: The follow-up period for the 464 patients was August 10–September 30, 2020. The most common sequelae were psychosocial problems (254 [57·7%]), respiratory function abnormality (149 [32·6%]), and cardiac function abnormality (98 [21·5%]). Rare sequelae were ADL disability (61 [13·3%]); pain (55 [12·5%]); feeding difficulties (54 [12·2%]); dysphonia (46 [10·4%]); and hyposmia (27 [6·1%]), as well as impairment in hearing (40 [9·1%]), vison (37 [8·4%]), swallowing (30 [6·8%]), and gustation (18 [4·1%]). Almost all patients (98·2%) had normal cognitive function. Risk factors associated with ADL disability were advanced age, intensive care unit (ICU) stay, and cancer, which were also associated with respiratory function abnormality. Risk factors associated with cardiac function abnormality were long hospital stay, cancer, and respiratory diseases. Advanced age, ICU stay, and nonischemic heart diseases were associated with psychosocial problems. Compared with female patients, male patients had decreased odds of declined respiratory, cardiac function, depression and anxiety, and pain.Interpretation: Nearly normal ADL, moderate cardiopulmonary function and psychosocial issues, and minor sensory abnormalities were observed in COVID-19 survivors. Advanced age, ICU stay, cancer, and female sex were adverse risk factors in long-term sequelae. These data provide a generalisable estimate of long-term outcomes for COVID-19.Trial Registration: This study was registered in the ClinicalTrial.gov database (NCT04508712).Funding: National Natural Science Foundation of China, Shanghai Shenkang Hospital Development Center, Shanghai municipal Education Commission, State Key Laboratoy of Neuroscience, and Renji Hospital.Declaration of Interests: All authors declare no competing interests.Ethics Approval Statement: This study was approved by the Institutional Research Ethics Committee of Renji Hospital, School of Medicine, Shanghai Jiaotong University (Ethical Committee approval number: KY2020-34). Oral consent was acquired from patients or from their immediate family members in cases of communication disorder or death.

The Association of Incidence of Face-Touching Behaviors with Masks Wearing Rates During the Coronavirus Disease 2019 Pandemic (preprint)

Background: Mandatory mask wearing policy for general population in public areas were the most controversial mitigative measure for the coronavirus disease 2019 (COVID-19) pandemic. Thus, it was imperative to investigate its influence on the incidence of face touching behaviors. Methods: Videos displaying mask-wearing and face-touching behaviors of general population in public areas were analyzed. Period before the COVID-19 epidemic were defined as January 2018 to October 2019, and those during the pandemic were from February 2020 to August 2020 in East Asia, and March to August 2020 in Europe and United States (US). Findings: 37 videos (4699 individuals) before the pandemic with 135 videos (8217 individuals) were included. The mask wearing rates all increased significantly during the pandemic. However, the incidence of face touching behaviors maintained. The incidence of eyes, nose and mouth touching behaviors decreased in East Asia and Europe, instead of US. Mask wearing rates was negatively related to incidences of face touching behaviors, especially in East Asia. Surprisingly, when mask wearing rates were >0% and <75%, mask wearing rates was positively related to the incidence of face touching behaviors in East Asia significantly (p=0.017). Interpretation: The incidence of face touching behaviors of general population in public areas was negatively associated with mask wearing rates. However, Mandatory mask wearing polices were risky in population with low adherence to masks, among whom, the face touching behaviors in public areas might increase with mask wearing rates rise. Funding: This study was supported by Guangzhou Science and Technology Project (201904010461).Declaration of Interests: The authors declare no competing interests.Ethics Approval Statement: This study was approved by the Clinical Ethics Review Board of the Third Affiliated Hospital of Sun Yat-sen University and Sun Yat-sen University. Informed consent was waived according to ethical review of biomedical research involving humans by Order of the National Health and Family Planning Commission of the People’s Republic of China No. 11.

Epidemic characteristics of respiratory viruses in hospitals in a Chinese city during the SARS-CoV-2 epidemic (preprint)

The SARS-CoV-2 virus first broke out in China in early 2020. The early symptoms of COVID-19 are similar to those of influenza. Therefore, during the epidemic, patients with similar symptoms will be tested for multiple pathogens at the same time. In order to control the spread of SARS-CoV-2, China has taken many measures. Under this special situation, have the types and epidemic characteristics of respiratory viruses changed? The nucleic acid test results of influenza A virus, influenza B virus and respiratory syncytial virus, as well as the antibody test results of 8 common respiratory viruses of Jinan Central Hospital were collected before and after the occurrence of SARS-CoV-2, and age distribution and time distribution characteristics were statisticed. Furthermore the epidemiological characteristics of this new virus before and after the SARS-CoV-2 epidemic was compared. In the early stage of the SARS-CoV-2 epidemic, influenza A, influenza B and respiratory syncytial virus nucleic acid test samples were large, and the positive rate of the three viruses was high. After that, the sample size and positive rate decreased significantly. No co-infection of SARS-CoV-2 and other viruses was found in our hospital. The sample size before the SARS-CoV-2 outbreak was larger than that after the outbreak, but the positive rate of the outbreak was lower than that after the outbreak. And the infection rate of children decreased in the middle and late stages of the epidemic. This is because since January 23, in order to prevent the spread of the new crown epidemic, my country has adopted measures such as wearing masks, not gathering together, and quarantining at home. This not only prevents the spread of the new crown virus, but also prevents the common respiratory tract. The spread of the virus has reduced the incidence of residents.

Early transfusion of a large cohort of COVID-19 patients with high titer anti-SARS-CoV-2 spike protein IgG convalescent plasma confirms a signal of significantly decreased mortality (preprint)

Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 remains a global threat with few proven efficacious treatments. Transfusion of convalescent plasma collected from donors who have recovered from COVID-19 disease has emerged as a promising therapy and has been granted emergency use authorization by the U.S. Food and Drug Administration (FDA). We recently reported results from interim analysis of a propensity-score matched study suggesting that early treatment of COVID-19 patients with convalescent plasma containing high titer anti-spike protein receptor binding domain (RBD) IgG significantly decreases mortality. We here present results from 60-day follow up of our cohort of 351 transfused hospitalized patients. Prospective determination of ELISA anti-RBD IgG titer facilitated selection and transfusion of the highest titer units available. Retrospective analysis by the Ortho VITROS IgG assay revealed a median signal/cutoff (S/C) ratio of 24.0 for transfused units, a value far exceeding the recently FDA-required cutoff of 12.0 for designation of high titer convalescent plasma. With respect to altering mortality, our analysis identified an optimal window of 44 hours post-hospitalization for transfusing COVID-19 patients with high titer convalescent plasma. In the aggregate, the analysis confirms and extends our previous preliminary finding that transfusion of COVID-19 patients soon after hospitalization with high titer anti-spike protein RBD IgG present in convalescent plasma significantly reduces mortality.

Interpretable Machine Learning for COVID-19: An Empirical Study on Severity Prediction Task (preprint)

The black-box nature of machine learning models hinders the deployment of some high-accuracy models in medical diagnosis. It is risky to put one's life in the hands of models that medical researchers do not fully understand. However, through model interpretation, black-box models can promptly reveal significant biomarkers that medical practitioners may have overlooked due to the surge of infected patients in the COVID-19 pandemic. This research leverages a database of 92 patients with confirmed SARS-CoV-2 laboratory tests between 18th Jan. 2020 and 5th Mar. 2020, in Zhuhai, China, to identify biomarkers indicative of severity prediction. Through the interpretation of four machine learning models, decision tree, random forests, gradient boosted trees, and neural networks using permutation feature importance, Partial Dependence Plot (PDP), Individual Conditional Expectation (ICE), Accumulated Local Effects (ALE), Local Interpretable Model-agnostic Explanations (LIME), and Shapley Additive Explanation (SHAP), we identify an increase in N-Terminal pro-Brain Natriuretic Peptide (NTproBNP), C-Reaction Protein (CRP), and lactic dehydrogenase (LDH), a decrease in lymphocyte (LYM) is associated with severe infection and an increased risk of death, which is consistent with recent medical research on COVID-19 and other research using dedicated models. We further validate our methods on a large open dataset with 5644 confirmed patients from the Hospital Israelita Albert Einstein, at S\~ao Paulo, Brazil from Kaggle, and unveil leukocytes, eosinophils, and platelets as three indicative biomarkers for COVID-19.

The Prediction for Development of COVID-19 in Global Major Epidemic Areas Through Empirical Trends in China by Utilizing State Transition Matrix Model (preprint)

Background: Since pneumonia caused by coronavirus disease 2019 (COVID-19) broke out in Wuhan, Hubei province, China, tremendous infected cases has risen all over the world attributed to its high transmissibility. We aimed to mathematically forecast the inflection point (IFP) of new cases in South Korea, Italy, and Iran, utilizing the transcendental model from China. Methods: : Data from reports released by the National Health Commission of the People’s Republic of China (Dec 31, 2019 to Mar 5, 2020) and the World Health Organization (Jan 20, 2020 to Mar 5, 2020) were extracted as the training set and the data from Mar 6 to 9 as the validation set. New close contacts, newly confirmed cases, cumulative confirmed cases, non-severe cases, severe cases, critical cases, cured cases, and death were collected and analyzed. We analyzed the data above through the State Transition Matrix model. Results: : The optimistic scenario (non-Hubei model, daily increment rate of -3.87%), the cautiously optimistic scenario (Hubei model, daily increment rate of -2.20%), and the relatively pessimistic scenario (adjustment, daily increment rate of -1.50%) were inferred and modeling from data in China. The IFP of time in South Korea would be Mar 6 to 12, Italy Mar 10 to 24, and Iran Mar 10 to 24. The numbers of cumulative confirmed patients will reach approximately 20k in South Korea, 209k in Italy, and 226k in Iran under fitting scenarios, respectively. However, with the adoption of different diagnosis criteria, the variation of new cases could impose various influences in the predictive model. If that happens, the IFP of increment will be earlier than predicted above. Conclusion: The end of the pandemic is still inapproachable, and the number of confirmed cases is still escalating. With the augment of data, the world epidemic trend could be further predicted, and it is imperative to consummate the assignment of global medical resources to curb the development of COVID-19.