Deep learning attention-guided radiomics for COVID-19 chest radiograph classification.

Background: Accurate assessment of coronavirus disease 2019 (COVID-19) lung involvement through chest radiograph plays an important role in effective management of the infection. This study aims to develop a two-step feature merging method to integrate image features from deep learning and radiomics to differentiate COVID-19, non-COVID-19 pneumonia and normal chest radiographs (CXR). Methods: In this study, a deformable convolutional neural network (deformable CNN) was developed and used as a feature extractor to obtain 1,024-dimensional deep learning latent representation (DLR) features. Then 1,069-dimensional radiomics features were extracted from the region of interest (ROI) guided by deformable CNN's attention. The two feature sets were concatenated to generate a merged feature set for classification. For comparative experiments, the same process has been applied to the DLR-only feature set for verifying the effectiveness of feature concatenation. Results: Using the merged feature set resulted in an overall average accuracy of 91.0% for three-class classification, representing a statistically significant improvement of 0.6% compared to the DLR-only classification. The recall and precision of classification into the COVID-19 class were 0.926 and 0.976, respectively. The feature merging method was shown to significantly improve the classification performance as compared to using only deep learning features, regardless of choice of classifier (P value <0.0001). Three classes' F1-score were 0.892, 0.890, and 0.950 correspondingly (i.e., normal, non-COVID-19 pneumonia, COVID-19). Conclusions: A two-step COVID-19 classification framework integrating information from both DLR and radiomics features (guided by deep learning attention mechanism) has been developed. The proposed feature merging method has been shown to improve the performance of chest radiograph classification as compared to the case of using only deep learning features.

Artificial intelligence-assisted multistrategy image enhancement of chest X-rays for COVID-19 classification.

Background: The coronavirus disease 2019 (COVID-19) led to a dramatic increase in the number of cases of patients with pneumonia worldwide. In this study, we aimed to develop an AI-assisted multistrategy image enhancement technique for chest X-ray (CXR) images to improve the accuracy of COVID-19 classification. Methods: Our new classification strategy consisted of 3 parts. First, the improved U-Net model with a variational encoder segmented the lung region in the CXR images processed by histogram equalization. Second, the residual net (ResNet) model with multidilated-rate convolution layers was used to suppress the bone signals in the 217 lung-only CXR images. A total of 80% of the available data were allocated for training and validation. The other 20% of the remaining data were used for testing. The enhanced CXR images containing only soft tissue information were obtained. Third, the neural network model with a residual cascade was used for the super-resolution reconstruction of low-resolution bone-suppressed CXR images. The training and testing data consisted of 1,200 and 100 CXR images, respectively. To evaluate the new strategy, improved visual geometry group (VGG)-16 and ResNet-18 models were used for the COVID-19 classification task of 2,767 CXR images. The accuracy of the multistrategy enhanced CXR images was verified through comparative experiments with various enhancement images. In terms of quantitative verification, 8-fold cross-validation was performed on the bone suppression model. In terms of evaluating the COVID-19 classification, the CXR images obtained by the improved method were used to train 2 classification models. Results: Compared with other methods, the CXR images obtained based on the proposed model had better performance in the metrics of peak signal-to-noise ratio and root mean square error. The super-resolution CXR images of bone suppression obtained based on the neural network model were also anatomically close to the real CXR images. Compared with the initial CXR images, the classification accuracy rates of the internal and external testing data on the VGG-16 model increased by 5.09% and 12.81%, respectively, while the values increased by 3.51% and 18.20%, respectively, for the ResNet-18 model. The numerical results were better than those of the single-enhancement, double-enhancement, and no-enhancement CXR images. Conclusions: The multistrategy enhanced CXR images can help to classify COVID-19 more accurately than the other existing methods.

Radiomics-Based Detection of COVID-19 from Chest X-ray Using Interpretable Soft Label-Driven TSK Fuzzy Classifier.

The COVID-19 pandemic has posed a significant global public health threat with an escalating number of new cases and death toll daily. The early detection of COVID-related CXR abnormality potentially allows the early isolation of suspected cases. Chest X-Ray (CXR) is a fast and highly accessible imaging modality. Recently, a number of CXR-based AI models have been developed for the automated detection of COVID-19. However, most existing models are difficult to interpret due to the use of incomprehensible deep features in their models. Confronted with this, we developed an interpretable TSK fuzzy system in this study for COVID-19 detection using radiomics features extracted from CXR images. There are two main contributions. (1) When TSK fuzzy systems are applied to classification tasks, the commonly used binary label matrix of training samples is transformed into a soft one in order to learn a more discriminant transformation matrix and hence improve classification accuracy. (2) Based on the assumption that the samples in the same class should be kept as close as possible when they are transformed into the label space, the compactness class graph is introduced to avoid overfitting caused by label matrix relaxation. Our proposed model for a multi-categorical classification task (COVID-19 vs. No-Findings vs. Pneumonia) was evaluated using 600 CXR images from publicly available datasets and compared against five state-of-the-art AI models in aspects of classification accuracy. Experimental findings showed that our model achieved classification accuracy of over 83%, which is better than the state-of-the-art models, while maintaining high interpretability.

Ultrafine particles exposure is associated with specific operative procedures in a multi-chair dental clinic.

Air quality in dental clinics is critical, especially in light of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) pandemic, given that dental professionals and patients are at risk of regular exposure to aerosols and bioaerosols in dental clinics. High levels of ultrafine particles (UFP) may be produced by dental procedures. This study aimed to quantify ultrafine particles (UFP) concentrations in a real multi-chair dental clinic and compare the levels of UFP produced by different dental procedures. The efficiency of a high-volume evacuator (HVE) in reducing the UFP concentrations during dental procedures was also assessed. UFP concentrations were measured both inside and outside of a dental clinic in Shanghai, China during a 12-day period from July to September 2020. Dental activities were recorded during working hours. The mean (±standard deviation) concentrations of indoor and outdoor UFP during the sampling period were 8,209 (±4,407) counts/cm3 and 15,984 (±7,977) counts/cm3, respectively. The indoor UFP concentration was much higher during working hours (10,057 ± 5,725 counts/cm3) than during non-working hours (7,163 ± 2,972 counts/cm3). The UFP concentrations increased significantly during laser periodontal treatment, root canal filling, tooth drilling, and grinding, and were slightly elevated during ultrasonic scaling or tooth extraction by piezo-surgery. The highest UFP concentration (241,136 counts/cm3) was observed during laser periodontal treatment, followed by root canal filling (75,034 counts/cm3), which showed the second highest level. The use of an HVE resulted in lower number concentration of UFP when drilling and grinding teeth with high-speed handpieces, but did not significantly reduce UFP measured during laser periodontal therapy. we found that many dental procedures can generate high concentration of UFP in dental clinics, which may have a great health impact on the dental workers. The use of an HVE may help reduce the exposure to UFP during the use of high-speed handpieces.

The effect of COVID-19 restrictions on atmospheric new particle formation in Beijing

During the COVID-19 lockdown, the dramatic reduction of anthropogenic emissions provided a unique opportunity to investigate the effects of reduced anthropogenic activity and primary emissions on atmospheric chemical processes and the consequent formation of secondary pollutants. Here, we utilize comprehensive observations to examine the response of atmospheric new particle formation (NPF) to the changes in the atmospheric chemical cocktail. We find that the main clustering process was unaffected by the drastically reduced traffic emissions, and the formation rate of 1.5 nm particles remained unaltered. However, particle survival probability was enhanced due to an increased particle growth rate (GR) during the lockdown period, explaining the enhanced NPF activity in earlier studies. For GR at 1.5–3 nm, sulfuric acid (SA) was the main contributor at high temperatures, whilst there were unaccounted contributing vapors at low temperatures. For GR at 3–7 and 7–15 nm, oxygenated organic molecules (OOMs) played a major role. Surprisingly, OOM composition and volatility were insensitive to the large change of atmospheric NOx concentration;instead the associated high particle growth rates and high OOM concentration during the lockdown period were mostly caused by the enhanced atmospheric oxidative capacity. Overall, our findings suggest a limited role of traffic emissions in NPF.

Development and validation of bone-suppressed deep learning classification of COVID-19 presentation in chest radiographs.

Background: Coronavirus disease 2019 (COVID-19) is a pandemic disease. Fast and accurate diagnosis of COVID-19 from chest radiography may enable more efficient allocation of scarce medical resources and hence improved patient outcomes. Deep learning classification of chest radiographs may be a plausible step towards this. We hypothesize that bone suppression of chest radiographs may improve the performance of deep learning classification of COVID-19 phenomena in chest radiographs. Methods: Two bone suppression methods (Gusarev et al. and Rajaraman et al.) were implemented. The Gusarev and Rajaraman methods were trained on 217 pairs of normal and bone-suppressed chest radiographs from the X-ray Bone Shadow Suppression dataset (https://www.kaggle.com/hmchuong/xray-bone-shadow-supression). Two classifier methods with different network architectures were implemented. Binary classifier models were trained on the public RICORD-1c and RSNA Pneumonia Challenge datasets. An external test dataset was created retrospectively from a set of 320 COVID-19 positive patients from Queen Elizabeth Hospital (Hong Kong, China) and a set of 518 non-COVID-19 patients from Pamela Youde Nethersole Eastern Hospital (Hong Kong, China), and used to evaluate the effect of bone suppression on classifier performance. Classification performance, quantified by sensitivity, specificity, negative predictive value (NPV), accuracy and area under the receiver operating curve (AUC), for non-suppressed radiographs was compared to that for bone suppressed radiographs. Some of the pre-trained models used in this study are published at (https://github.com/danielnflam). Results: Bone suppression of external test data was found to significantly (P<0.05) improve AUC for one classifier architecture [from 0.698 (non-suppressed) to 0.732 (Rajaraman-suppressed)]. For the other classifier architecture, suppression did not significantly (P>0.05) improve or worsen classifier performance. Conclusions: Rajaraman suppression significantly improved classification performance in one classification architecture, and did not significantly worsen classifier performance in the other classifier architecture. This research could be extended to explore the impact of bone suppression on classification of different lung pathologies, and the effect of other image enhancement techniques on classifier performance.

Increased levels of lactate dehydrogenase and hypertension are associated with severe illness of COVID-19.

BACKGROUND: Coronavirus disease 2019 (COVID-19) is amid an ongoing pandemic. It has been shown that patients with cardiovascular comorbidities are at higher risk of severe illness of COVID-19. AIM: To find out the relationship between cardiovascular comorbidities and severe illness of COVID-19. METHODS: The clinical data of 140 COVID-19 patients treated from January 22, 2020 to March 3, 2020 at our hospital were retrospectively collected. The clinical characteristics were compared between patients with mild illness and those with severe illness. RESULTS: There were 75 male patients and 65 female patients (53.6% vs 46.4%). The mean age was 45.4 ± 14.6 years (range, 2-85 years). Most of the patients had mild illness (n = 114, 81.4%) and 26 patients had severe illness (18.6%). The most common symptom was fever (n = 110, 78.6%), followed by cough (n = 82, 58.6%) and expectoration (n = 51, 36.4%). Eight patients were asymptomatic but were positive for severe acute respiratory syndrome coronavirus 2 RNA. Patients with severe illness were significantly more likely to be hypertensive than those with mild illness [(10/26, 38.4%) vs (22/114, 19.3%), P = 0.036]. The levels of lactate dehydrogenase were significantly higher in the severe illness group than in the mild illness group (299.35 ± 68.82 vs 202.94 ± 63.87, P < 0.001). No patient died in either the severe illness or the mild illness group. CONCLUSION: Hypertension and elevated levels of lactate dehydrogenase may be associated with severe illness of COVID-19.

Comparison of the community-acquired pneumonia and COVID-19 at the early stage: findings from two cohort studies.

BACKGROUND: The current focus is largely on whole course medical management of coronavirus disease-19 (COVID-19) with real-time polymerase chain reaction (RT-PCR) and radiological features, while the mild cases are usually missed. Thus, combination of multiple diagnostic methods is urgent to understand COVID-19 fully and to monitor the progression of COVID-19. METHODS: laboratory variables of 40 mild COVID-19 patients, 30 patients with community-acquired pneumonia (CAP) and 32 healthy individuals were analyzed by principal component analysis (PCA), Kruskal test, Procrustes test, the vegan package in R, CCA package and receiver operating characteristic to investigate the characteristics of the laboratory variables and their relationships in COVID-19. RESULTS: The correlations between the laboratory variables presented a variety of intricate linkages in the COVID-19 group compared with the healthy group and CAP patient group. The prediction probability of the combination of lymphocyte count (LY), eosinophil (EO) and platelets (PLT) was 0.847, 0.854 for the combination of lactate (LDH), creatine kinase isoenzyme (CK-MB), and C-reactive protein (CRP), 0.740 for the combination of EO, white blood cell count (WBC) and neutrophil count (NEUT) and 0.872 for the combination of CK-MB and P. CONCLUSIONS: The correlations between the laboratory variables in the COVID-19 group could be a unique characteristic showing promise as a method for COVID-19 prediction and monitoring progression of COVID-19 infection.

[Correlation between the early dynamic changes of lymphocyte and severity of disease in coronavirus disease 2019 patients].

OBJECTIVE: To explore the relationship between the dynamic changes of lymphocytes in the early stage (within 72 hours of admission) and the severity of disease in patients with coronavirus disease 2019 (COVID-19). METHODS: A retrospective study was conducted. The clinical data of COVID-19 patients admitted in Wenzhou Central Hospital from January 17, 2020 to February 14, 2020 were collected and analyzed. According to whether there was lymphopenia on the first day of admission [lymphocyte count (LYM) < 0.8×109/L], whether the difference between LYM on the third day and the first day of admission (ΔLYM) was less than 0, the patients were divided into four groups: the first group was LYM normal on the first day of admission, ΔLYM ≥ 0; the second group was LYM normal on the first day of admission, ΔLYM < 0; the third group was lymphopenia on the first day of admission, ΔLYM ≥ 0; the fourth group was lymphopenia on the first day of admission, ΔLYM < 0. The study endpoint was the development of severe/critically ill patients within 30 days after admission. Severe/critical standard referred to classification of Diagnosis and treatment protocol for coronavirus disease 2019 (trial version 5, revised edition). The differences in general information, laboratory results, and probability of developing severe/critical were compared among the four groups. Cox regression analysis was used to analyze the correlation between the early dynamic changes of lymphocytes and the probability of severe illness; and the Kaplan-Meier survival curve was drawn to assess the probability of severe illness in patients with different LYM groups. RESULTS: A total of 104 patients with COVID-19 were enrolled, and 21 patients developed to severe/critical cases within 30 days of onset (accounting for 20.2%; 17 severe cases and 4 critical cases). There were significant differences in age (F = 5.061, P = 0.003), white blood cell count (WBC) on the first day of admission (Z = 10.850, P = 0.013), C-reactive protein (CRP) on the first day of admission (Z = -4.449, P < 0.001), LYM on the first day of admission (Z = 43.132, P < 0.001), LYM on the third day of admission (Z = 40.340, P < 0.001), and the occurrence of severe/critical illness (χ2 = 18.645, P < 0.001) among the four groups. Patients in groups 3 and 4 were older; patients in group 3 had the lowest WBC and LYM on the first day of admission; patients in group 4 had the highest CRP on the first day of admission, the lowest LYM on the third day of admission, and high proportion of severe/critical cases. Regarding the probability of severe/critically ill patients within 30 days of admission, univariate Cox regression analysis showed that the probability of severe/critical patients in group 4 was 12.7 times higher than that in group 1 [hazard ratio (HR) = 12.732, 95% confidence interval (95%CI) was 3.951-41.025, P < 0.001]; age, CRP, albumin (ALB) and lymphocyte grouping were included in multivariate Cox regression analysis, the probability of severe/critically ill patients in group 4 was 6.4 times that of group 1 (HR = 6.398, 95%CI was 1.757-23.301, P = 0.005); however, there was no difference in the probability among the group 1, 2 and 3. Kaplan-Meier survival curve showed that the probability of severe/critically ill patients in group 4 was significantly higher than that in groups 1, 2 and 3 (Log-Rank test: χ2 = 42.617, P < 0.001). CONCLUSIONS: Early lymphocyte dynamics change is related to the severity of patients with COVID-19. Patients with low LYM on the first day and continued decrease within 72 hours of admission have a higher probability to develop into severe/critically cases.

Development and clinical evaluation of a rapid antibody lateral flow assay for the diagnosis of SARS-CoV-2 infection.

BACKGROUND: The novel coronavirus disease 2019 (COVID-19) is an infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which has quickly spread worldwide since its outbreak in December 2019. One of the primary measures for controlling the spread of SARS-CoV-2 infection is an accurate assay for its diagnosis. SARS-CoV-2 real-time PCR kits suffer from some limitations, including false-negative results in the clinic. Therefore, there is an urgent need for the development of a rapid antibody test kit for COVID-19 diagnosis. METHODS: The nuclear capsid protein (N) and spike protein 1 (S1) fragments of SARS-CoV-2 were expressed in Escherichia coli, and rapid antibody-based tests for the diagnosis of SARS-CoV-2 infection were developed. To evaluate their clinical applications, the serum from COVID-19 patients, suspected COVID-19 patients, recovering COVID-19 patients, patients with general fever or pulmonary infection, doctors and nurses who worked at the fever clinic, and health professionals was analyzed by the rapid antibody test kits. The serum from patients infected with Mycoplasma pneumoniae and patients with respiratory tract infection was further analyzed to test its cross-reactivity with other respiratory pathogens. RESULTS: A 47 kDa N protein and 67 kDa S1 fragment of SARS-CoV-2 were successfully expressed, purified, and renatured. The rapid antibody test with recombinant N protein showed higher positive rate than the rapid IgM antibody test with recombinant S1 protein. Clinical evaluation showed that the rapid antibody test kit with recombinant N protein had 88.56 % analytical sensitivity and 97.42 % specificity for COVID-19 patients, 53.48 % positive rate for suspected COVID-19 patients, 57.14 % positive rate for recovering COVID-19 patients, and 0.5-0.8 % cross-reactivity with other respiratory pathogens. The analytical sensitivity of the kit did not significantly differ in COVID-19 patients with different disease courses (p < 0.01). CONCLUSIONS: The rapid antibody test kit with recombinant N protein has high specificity and analytical sensitivity, and can be used for the diagnosis of SARS-CoV-2 infection combined with RT-PCR.

Resumption of the treatment of non-COVID-19 gynecologic patients after lifting lockdown: Triage and infection prevention experiences from Wuhan.

AIM: To share our experiences of resuming the treatments for gynecologic patients after lifting the lockdown in a hotspot area for the Coronavirus Disease 2019 (COVID-19) pandemic. METHODS: The triage process used to resume medical activities for gynecologic patients at the Wuhan Union Hospital after a 76-day lockdown of the city is described, and its effectiveness in preventing COVID-19 nosocomial transmission is shown. RESULTS: Nonemergency patients were pretriaged based on their contact history and body temperature at an outpatient clinic, and negative COVID-19 screening test results were required for admission to the buffering rooms at the gynecologic department. The buffering lasted at least 3 days for symptom monitoring, and a second round of COVID-19 testing was required before patients could be transferred to the regular gynecologic wards. For patients who needed emergency surgery, the first screening was completed at the isolation wards after surgery, followed by buffering at the gynecologic department. We received 19 298 outpatient visits, admitted 326 patients, and performed 223 operations in the first 2 months after the lockdown was lifted. No COVID-19 cases occurred in the hospitalized patients, while the proportion of potentially high-risk patients with cancer and severe anemia was increased in comparison to that observed during the same period in 2019 and the first 2 months of 2020 before the lockdown. CONCLUSIONS: We provide an effective triage system with buffering at two levels to guarantee safe and timely treatment for non-COVID-19 gynecologic patients in the postlockdown phase.

Malnutrition is associated with hyperinflammation and immunosuppression in COVID-19 patients: A prospective observational study.

BACKGROUND: Coronavirus disease 2019 (COVID-19) is spreading globally and has caused many deaths. This study investigated, for the first time, COVID-19 patients' nutrition status and its effects on their inflammatory and immune responses. METHODS: Forty-seven COVID-19 patients were recruited for this prospective study. According to the subjective global assessment at admission, patients were divided into the normal nutrition (NN), risk of malnutrition (RMN), or MN group. Serum cytokines and whole blood T-cell subpopulations were measured to assess the inflammatory and immune responses in COVID-19 patients. Analysis of covariance and χ2 tests were used. RESULTS: On admission, the incidences of MN and the RMN in COVID-19 patients were 17.0% and 38.3%, respectively. The MN group had a higher proportion with severe/critical COVID-19 and a longer hospitalization duration than the NN group. Serum interleukin (IL) 6 concentrations were elevated in 97.9% of the patients and were the highest in malnourished patients. The IL-4 and IL-10 levels were elevated in 46.8% and 48.9% of the patients, respectively. The proportion of CD8+ T cells was significantly lower in the MN group than in the NN group. CONCLUSION: A high proportion of COVID-19 patients are malnourished or at risk of malnuourishment, especially those with severe disease. MN is associated with hyperinflammation and immunosuppression in COVID-19 patients, and it may contribute to disease progression.

Clinical features and viral RNA shedding of imported and local cases with COVID-19 in Wenzhou, China.

ABSTRACT: Wenzhou had the highest number of confirmed novel coronavirus 2019 (COVID-19) cases outside the Hubei province. The aim of this study was to identify the difference in clinical features and viral RNA shedding between the imported and local COVID-19 cases in Wenzhou.All patients with confirmed COVID-19 admitted to Wenzhou Sixth People's Hospital, Wenzhou Central Hospital Medical Group, from January 17 to February 11, 2020, were enrolled in this study. Data was analyzed and compared for the imported and local cases with regard to epidemiological, demographic, clinical, radiological features, and laboratory findings. Outcomes for the enrolled participants were followed up until May 7, 2020.Of the 136 cases, 50 were imported from Wuhan. The median age was 45 years and 73 (53.7%) were men. The most common symptoms at onset were fever (104 [76.5%]) and cough (85[62.5%]). Pleural effusion was more common among imported cases compared to local cases. The white blood cell count, neutrophil count, lymphocyte count and platelet count of the imported cases were significantly lower than those of the local cases, while the prothrombin time was significantly longer than that of the local cases. Severe and critically ill patients accounted for 15.4% and 2.9%, respectively. The median duration of SARS-CoV-2 RNA shedding from symptom onset was 26 days (IQR 17-32.3 days) and there were no significant differences in duration of viral RNA shedding between the two groups.The study findings suggest that imported cases from Wuhan were more likely to be severe compared to the local cases in Wenzhou. However, there was no difference between imported and local cases on the viral shedding among the COVID patients.

Do post-COVID-19 symptoms exist? A longitudinal study of COVID-19 sequelae in Wenzhou, China.

OBJECTIVE: We conducted a cohort of tracing discharge patients of COVID-19. MATERIALS AND METHODS: We used the Mann-Whitney U test, χ2 test, or Fisher's exact test to compare differences between age groups and gender groups where appropriate. RESULTS: Our study provides insights into the nature and severity of medical conditions specific to survivors of COVID-19. CONCLUSIONS: It also highlights the potential mental health issues resulting from infectious disease outbreaks within communities.

OBJECTIFS: Nous avons suivi une cohorte de patients à la sortie du COVID-19. MATÉRIAUX ET MÉTHODES: Nous avons utilisé les test de Mann­Whitney U, de Fisher ou du Chi2 pour comparer les différences entre les groupes d'âge et de genre, le cas échéant. RÉSULTATS: Notre étude fournit un aperçu de la nature et de la gravité des troubles médicaux propres aux survivants du COVID-19. CONCLUSIONS: Elle met également en lumière les problèmes de santé mentale potentiels découlant des éclosions de maladies infectieuses dans les collectivités.

Corticosteroid treatment in severe COVID-19 patients with acute respiratory distress syndrome.

BACKGROUNDCorticosteroids are widely used in patients with COVID 19, although their benefit-to-risk ratio remains controversial.METHODSPatients with severe COVID-19-related acute respiratory distress syndrome (ARDS) were included from December 29, 2019 to March 16, 2020 in 5 tertiary Chinese hospitals. Cox proportional hazards and competing risks analyses were conducted to analyze the impact of corticosteroids on mortality and SARS-CoV-2 RNA clearance, respectively. We performed a propensity score (PS) matching analysis to control confounding factors.RESULTSOf 774 eligible patients, 409 patients received corticosteroids, with a median time from hospitalization to starting corticosteroids of 1.0 day (IQR 0.0-3.0 days) . As compared with usual care, treatment with corticosteroids was associated with increased rate of myocardial (15.6% vs. 10.4%, P = 0.041) and liver injury (18.3% vs. 9.9%, P = 0.001), of shock (22.0% vs. 12.6%, P < 0.001), of need for mechanical ventilation (38.1% vs. 19.5%, P < 0.001), and increased rate of 28-day all-cause mortality (44.3% vs. 31.0%, P < 0.001). After PS matching, corticosteroid therapy was associated with 28-day mortality (adjusted HR 1.46, 95% CI 1.01-2.13, P = 0.045). High dose (>200 mg) and early initiation (≤3 days from hospitalization) of corticosteroid therapy were associated with a higher 28-day mortality rate. Corticosteroid use was also associated with a delay in SARS-CoV-2 coronavirus RNA clearance in the competing risk analysis (subhazard ratio 1.59, 95% CI 1.17-2.15, P = 0.003).CONCLUSIONAdministration of corticosteroids in severe COVID-19-related ARDS is associated with increased 28-day mortality and delayed SARS-CoV-2 coronavirus RNA clearance after adjustment for time-varying confounders.FUNDINGNone.

No significant association between dipeptidyl peptidase-4 inhibitors and adverse outcomes of COVID-19.

BACKGROUND: Dipeptidyl peptidase-4 (DPP4) is commonly targeted to achieve glycemic control and has potent anti-inflammatory and immunoregulatory effects. Recent structural analyses indicated a potential tight interaction between DPP4 and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), raising a promising hypothesis that DPP4 inhibitor (DPP4i) drugs might be an optimal strategy for treating coronavirus disease 2019 (COVID-19) among patients with diabetes. However, there has been no direct clinical evidence illuminating the associations between DPP4i use and COVID-19 outcomes. AIM: To illuminate the associations between DPP4i usage and the adverse outcomes of COVID-19. METHODS: We conducted a multicenter, retrospective analysis including 2563 patients with type 2 diabetes who were hospitalized due to COVID-19 at 16 hospitals in Hubei Province, China. After excluding ineligible individuals, 142 patients who received DPP4i drugs and 1115 patients who received non-DPP4i oral anti-diabetic drugs were included in the subsequent analysis. We performed a strict propensity score matching (PSM) analysis where age, sex, comorbidities, number of oral hypoglycemic agents, heart rate, blood pressure, pulse oxygen saturation (SpO2) < 95%, CT diagnosed bilateral lung lesions, laboratory indicators, and proportion of insulin usage were matched. Finally, 111 participants treated with DPP4i drugs were successfully matched to 333 non-DPP4i users. Then, a linear logistic model and mixed-effect Cox model were applied to analyze the associations between in-hospital DPP4i use and adverse outcomes of COVID-19. RESULTS: After rigorous matching and further adjustments for imbalanced variables in the linear logistic model and Cox adjusted model, we found that there was no significant association between in-hospital DPP4i use (DPP4i group) and 28-d all-cause mortality (adjusted hazard ratio = 0.44, 95%CI: 0.09-2.11, P = 0.31). Likewise, the incidences and risks of secondary outcomes, including septic shock, acute respiratory distress syndrome, or acute organ (kidney, liver, and cardiac) injuries, were also comparable between the DPP4i and non-DPP4i groups. The performance of DPP4i agents in achieving glucose control (e.g., the median level of fasting blood glucose and random blood glucose) and inflammatory regulation was approximately equivalent in the DPP4i and non-DPP4i groups. Furthermore, we did not observe substantial side effects such as uncontrolled glycemia or acidosis due to DPP4i application relative to the use of non-DPP4i agents in the study cohort. CONCLUSION: Our findings demonstrated that DPP4i use is not significantly associated with poor outcomes of COVID-19 or other adverse effects of anti-diabetic treatment. The data support the continuation of DPP4i agents for diabetes management in the setting of COVID-19.

Epidemiological characteristics and the entire evolution of coronavirus disease 2019 in Wuhan, China.

BACKGROUND: Coronavirus Disease 2019 (COVID-19) spread rapidly around the world. We aimed to describe the epidemiological characteristics and the entire evolution of COVID-19 in Wuhan, and to evaluate the effect of non-pharmaceutical intervention by the government. METHODS: The information of COVID-19 cases until Mar 18, 2020 in Wuhan were collected from the national infectious disease surveillance system in Hubei province. RESULTS: A total of 49,973 confirmed cases were reported until Mar 18, 2020 in Wuhan. Among whom, 2496 cases died and the overall mortality was 5.0%. Most confirmed cases (25,619, 51.3%) occurred during Jan 23 to Feb 4, with a spike on Feb 1 (new cases, 3374). The number of daily new cases started to decrease steadily on Feb 19 (new cases, 301) and decreased greatly on Mar 1 (new cases, 57). However, the mortality and the proportion of severe and critical cases has been decreasing over time, with the lowest of 2.0 and 10.1% during Feb 16 to Mar 18, 2020, respectively. The percentage of severe and critical cases among all cases was 19.6%, and the percentage of critical and dead cases aged over 60 was 70.1 and 82.0%, respectively. CONCLUSION: The number of new cases has dropped significantly after the government taking the isolation of four types of personnel and the community containment for 14 days. Our results indicate that the mortality and proportion of severe and critical cases gradually decreased over time, and critical and dead cases are more incline to be older individuals.

Towards an Artificial Intelligence Framework for Data-Driven Prediction of Coronavirus Clinical Severity

The virus SARS-CoV2, which causes coronavirus disease (COVID-19) has become a pandemic and has spread to every inhabited continent. Given the increasing caseload, there is an urgent need to augment clinical skills in order to identify from among the many mild cases the few that will progress to critical illness. We present a first step towards building an artificial intelligence (AI) framework, with predictive analytics (PA) capabilities applied to real patient data, to provide rapid clinical decision-making support. COVID-19 has presented a pressing need as a) clinicians are still developing clinical acumen to this novel disease and b) resource limitations in a surging pandemic require difficult resource allocation decisions. The objectives of this research are: (1) to algorithmically identify the combinations of clinical characteristics of COVID-19 that predict outcomes, and (2) to develop a tool with AI capabilities that will predict patients at risk for more severe illness on initial presentation. The predictive models learn from historical data to help predict who will develop acute respiratory distress syndrome (ARDS), a severe outcome in COVID-19. Our results, based on data from two hospitals in Wenzhou, Zhejiang, China, identified features on initial presentation with COVID-19 that were most predictive of later development of ARDS. A mildly elevated alanine aminotransferase (ALT) (a liver enzyme), the presence of myalgias (body aches), and an elevated hemoglobin (red blood cells), in this order, are the clinical features, on presentation, that are the most predictive. The predictive models that learned from historical data of patients from these two hospitals achieved 70% to 80% accuracy in predicting severe cases.

Predictive Model and Risk Factors for Case Fatality of COVID-19: A Cohort of 21,392 Cases in Hubei, China.

An increasing number of patients are being killed by coronavirus disease 2019 (COVID-19), however, risk factors for the fatality of COVID-19 remain unclear. A total of 21,392 COVID-19 cases were recruited in the Hubei Province of China between December 2019 and February 2020, and followed up until March 18, 2020. We adopted Cox regression models to investigate the risk factors for case fatality and predicted the death probability under specific combinations of key predictors. Among the 21,392 patients, 1,020 (4.77%) died of COVID-19. Multivariable analyses showed that factors, including age (≥60 versus <45 years, hazard ratio [HR] = 7.32; 95% confidence interval [CI], 5.42, 9.89), sex (male versus female, HR = 1.31; 95% CI, 1.15, 1.50), severity of the disease (critical versus mild, HR = 39.98; 95% CI, 29.52, 48.86), comorbidity (HR = 1.40; 95% CI, 1.23, 1.60), highest body temperature (>39°C versus <39°C, HR = 1.28; 95% CI, 1.09, 1.49), white blood cell counts (>10 × 109/L versus (4-10) × 109/L, HR = 1.69; 95% CI, 1.35, 2.13), and lymphocyte counts (<0.8 × 109/L versus (0.8-4) × 109/L, HR = 1.26; 95% CI, 1.06, 1.50) were significantly associated with case fatality of COVID-19 patients. Individuals of an older age, who were male, with comorbidities, and had a critical illness had the highest death probability, with 21%, 36%, 46%, and 54% within 1-4 weeks after the symptom onset. Risk factors, including demographic characteristics, clinical symptoms, and laboratory factors were confirmed to be important determinants of fatality of COVID-19. Our predictive model can provide scientific evidence for a more rational, evidence-driven allocation of scarce medical resources to reduce the fatality of COVID-19.