Development and external evaluation of predictions models for mortality of COVID-19 patients using machine learning method.

To predict the mortality of patients with coronavirus disease 2019 (COVID-19). We collected clinical data of COVID-19 patients between January 18 and March 29 2020 in Wuhan, China . Gradient boosting decision tree (GBDT), logistic regression (LR) model, and simplified LR were built to predict the mortality of COVID-19. We also evaluated different models by computing area under curve (AUC), accuracy, positive predictive value (PPV), and negative predictive value (NPV) under fivefold cross-validation. A total of 2924 patients were included in our evaluation, with 257 (8.8%) died and 2667 (91.2%) survived during hospitalization. Upon admission, there were 21 (0.7%) mild cases, 2051 (70.1%) moderate case, 779 (26.6%) severe cases, and 73 (2.5%) critically severe cases. The GBDT model exhibited the highest fivefold AUC, which was 0.941, followed by LR (0.928) and LR-5 (0.913). The diagnostic accuracies of GBDT, LR, and LR-5 were 0.889, 0.868, and 0.887, respectively. In particular, the GBDT model demonstrated the highest sensitivity (0.899) and specificity (0.889). The NPV of all three models exceeded 97%, while their PPV values were relatively low, resulting in 0.381 for LR, 0.402 for LR-5, and 0.432 for GBDT. Regarding severe and critically severe cases, the GBDT model also performed the best with a fivefold AUC of 0.918. In the external validation test of the LR-5 model using 72 cases of COVID-19 from Brunei, leukomonocyte (%) turned to show the highest fivefold AUC (0.917), followed by urea (0.867), age (0.826), and SPO2 (0.704). The findings confirm that the mortality prediction performance of the GBDT is better than the LR models in confirmed cases of COVID-19. The performance comparison seems independent of disease severity. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at(10.1007/s00521-020-05592-1).

Association Between Ischemic Stroke and COVID-19 in China: A Population-Based Retrospective Study.

Background and Purpose: To investigate the effect of prior ischemic stroke on the outcomes of patients hospitalized with coronavirus disease 2019 (COVID-19), and to describe the incidence, clinical features, and risk factors of acute ischemic stroke (AIS) following COVID-19. Methods: In this population-based retrospective study, we included all the hospitalized positive patients with COVID-19 at Wuhan City from December 29, 2019 to April 15, 2020. Clinical data were extracted from administrative datasets coordinated by the Wuhan Health Commission. The propensity score matching and multivariate logistic regression analyses were used to adjust the confounding factors. Results: There are 36,358 patients in the final cohort, in which 1,160 (3.2%) had a prior stroke. After adjusting for available baseline characteristics, patients with prior stroke had a higher proportion of severe and critical illness and mortality. We found for the first time that the premorbid modified Rankin Scale (MRS) grouping (odds ratio [OR] = 1.796 [95% CI 1.334-2.435], p < 0.001) and older age (OR = 1.905 [95% CI 1.211-3.046], p = 0.006) imparted increased risk of death. AIS following COVID-19 occurred in 124 (0.34%) cases, and patients with prior stroke had a much higher incidence of AIS (3.4%). Logistic regression analyses confirmed an association between the severity of COVID-19 with the incidence of AIS. COVID-19 patients with AIS had a significantly higher mortality compared with COVID-19 patients without stroke and AIS patients without COVID-19. Conclusions: Coronavirus disease 2019 patients with prior stroke, especially those with the higher premorbid MRS or aged, have worse clinical outcomes. Furthermore, COVID-19 increases the incidence of AIS, and the incidence is positively associated with the severity of COVID-19.

Association Between Ischemic Stroke and COVID-19 in China: A Population-Based Retrospective Study

Background and Purpose To investigate the effect of prior ischemic stroke on the outcomes of patients hospitalized with coronavirus disease 2019 (COVID-19), and to describe the incidence, clinical features, and risk factors of acute ischemic stroke (AIS) following COVID-19. Methods In this population-based retrospective study, we included all the hospitalized positive patients with COVID-19 at Wuhan City from December 29, 2019 to April 15, 2020. Clinical data were extracted from administrative datasets coordinated by the Wuhan Health Commission. The propensity score matching and multivariate logistic regression analyses were used to adjust the confounding factors. Results There are 36,358 patients in the final cohort, in which 1,160 (3.2%) had a prior stroke. After adjusting for available baseline characteristics, patients with prior stroke had a higher proportion of severe and critical illness and mortality. We found for the first time that the premorbid modified Rankin Scale (MRS) grouping (odds ratio [OR] = 1.796 [95% CI 1.334–2.435], p < 0.001) and older age (OR = 1.905 [95% CI 1.211–3.046], p = 0.006) imparted increased risk of death. AIS following COVID-19 occurred in 124 (0.34%) cases, and patients with prior stroke had a much higher incidence of AIS (3.4%). Logistic regression analyses confirmed an association between the severity of COVID-19 with the incidence of AIS. COVID-19 patients with AIS had a significantly higher mortality compared with COVID-19 patients without stroke and AIS patients without COVID-19. Conclusions Coronavirus disease 2019 patients with prior stroke, especially those with the higher premorbid MRS or aged, have worse clinical outcomes. Furthermore, COVID-19 increases the incidence of AIS, and the incidence is positively associated with the severity of COVID-19.

Age-Related Risk Factors and Complications of Patients With COVID-19: A Population-Based Retrospective Study

Objective: To study the differences in clinical characteristics, risk factors, and complications across age-groups among the inpatients with the coronavirus disease 2019 (COVID-19). Methods: In this population-based retrospective study, we included all the positive hospitalized patients with COVID-19 at Wuhan City from December 29, 2019 to April 15, 2020, during the first pandemic wave. Multivariate logistic regression analyses were used to explore the risk factors for death from COVID-19. Canonical correlation analysis (CCA) was performed to study the associations between comorbidities and complications. Results: There are 36,358 patients in the final cohort, of whom 2,492 (6.85%) died. Greater age (odds ration [OR] = 1.061 [95% CI 1.057–1.065], p < 0.001), male gender (OR = 1.726 [95% CI 1.582–1.885], p < 0.001), alcohol consumption (OR = 1.558 [95% CI 1.355–1.786], p < 0.001), smoking (OR = 1.326 [95% CI 1.055–1.652], p = 0.014), hypertension (OR = 1.175 [95% CI 1.067–1.293], p = 0.001), diabetes (OR = 1.258 [95% CI 1.118–1.413], p < 0.001), cancer (OR = 1.86 [95% CI 1.507–2.279], p < 0.001), chronic kidney disease (CKD) (OR = 1.745 [95% CI 1.427–2.12], p < 0.001), and intracerebral hemorrhage (ICH) (OR = 1.96 [95% CI 1.323–2.846], p = 0.001) were independent risk factors for death from COVID-19. Patients aged 40–80 years make up the majority of the whole patients, and them had similar risk factors with the whole patients. For patients aged <40 years, only cancer (OR = 17.112 [95% CI 6.264–39.73], p < 0.001) and ICH (OR = 31.538 [95% CI 5.213–158.787], p < 0.001) were significantly associated with higher odds of death. For patients aged >80 years, only age (OR = 1.033 [95% CI 1.008–1.059], p = 0.01) and male gender (OR = 1.585 [95% CI 1.301–1.933], p < 0.001) were associated with higher odds of death. The incidence of most complications increases with age, but arrhythmias, gastrointestinal bleeding, and sepsis were more common in younger deceased patients with COVID-19, with only arrhythmia reaching statistical difference (p = 0.039). We found a relatively poor correlation between preexisting risk factors and complications. Conclusions: Coronavirus disease 2019 are disproportionally affected by age for its clinical manifestations, risk factors, complications, and outcomes. Prior complications have little effect on the incidence of extrapulmonary complications.

Immunologic memory to SARS-CoV-2 in convalescent COVID-19 patients at 1 year postinfection.

BACKGROUND: Understanding the complexities of immune memory to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is key to gain insights into the durability of protective immunity against reinfection. OBJECTIVE: We sought to evaluate the immune memory to SARS-CoV-2 in convalescent patients with longer follow-up time. METHODS: SARS-CoV-2-specific humoral and cellular responses were assessed in convalescent patients with coronavirus disease 2019 (COVID-19) at 1 year postinfection. RESULTS: A total of 78 convalescent patients with COVID-19 (26 moderate, 43 severe, and 9 critical) were recruited after 1 year of recovery. The positive rates of both anti-receptor-binding domain and antinucleocapsid antibodies were 100%, whereas we did not observe a statistical difference in antibody levels among different severity groups. Accordingly, the prevalence of neutralizing antibodies (nAbs) reached 93.59% in convalescent patients. Although nAb titers displayed an increasing trend in convalescent patients with increased severity, the difference failed to achieve statistical significance. Notably, there was a significant correlation between nAb titers and anti-receptor-binding domain levels. Interestingly, SARS-CoV-2-specific T cells could be robustly maintained in convalescent patients, and their number was positively correlated with both nAb titers and anti-receptor-binding domain levels. Amplified SARS-CoV-2-specific CD4+ T cells mainly produced a single cytokine, accompanying with increased expression of exhaustion markers including PD-1, Tim-3, TIGIT, CTLA-4, and CD39, while the proportion of multifunctional cells was low. CONCLUSIONS: Robust SARS-CoV-2-specific humoral and cellular responses are maintained in convalescent patients with COVID-19 at 1 year postinfection. However, the dysfunction of SARS-CoV-2-specific CD4+ T cells supports the notion that vaccination is needed in convalescent patients for preventing reinfection.

Deep learning-based triage and analysis of lesion burden for COVID-19: a retrospective study with external validation.

Background: Prompt identification of patients suspected to have COVID-19 is crucial for disease control. We aimed to develop a deep learning algorithm on the basis of chest CT for rapid triaging in fever clinics. Methods: We trained a U-Net-based model on unenhanced chest CT scans obtained from 2447 patients admitted to Tongji Hospital (Wuhan, China) between Feb 1, 2020, and March 3, 2020 (1647 patients with RT-PCR-confirmed COVID-19 and 800 patients without COVID-19) to segment lung opacities and alert cases with COVID-19 imaging manifestations. The ability of artificial intelligence (AI) to triage patients suspected to have COVID-19 was assessed in a large external validation set, which included 2120 retrospectively collected consecutive cases from three fever clinics inside and outside the epidemic centre of Wuhan (Tianyou Hospital [Wuhan, China; area of high COVID-19 prevalence], Xianning Central Hospital [Xianning, China; area of medium COVID-19 prevalence], and The Second Xiangya Hospital [Changsha, China; area of low COVID-19 prevalence]) between Jan 22, 2020, and Feb 14, 2020. To validate the sensitivity of the algorithm in a larger sample of patients with COVID-19, we also included 761 chest CT scans from 722 patients with RT-PCR-confirmed COVID-19 treated in a makeshift hospital (Guanggu Fangcang Hospital, Wuhan, China) between Feb 21, 2020, and March 6, 2020. Additionally, the accuracy of AI was compared with a radiologist panel for the identification of lesion burden increase on pairs of CT scans obtained from 100 patients with COVID-19. Findings: In the external validation set, using radiological reports as the reference standard, AI-aided triage achieved an area under the curve of 0·953 (95% CI 0·949-0·959), with a sensitivity of 0·923 (95% CI 0·914-0·932), specificity of 0·851 (0·842-0·860), a positive predictive value of 0·790 (0·777-0·803), and a negative predictive value of 0·948 (0·941-0·954). AI took a median of 0·55 min (IQR: 0·43-0·63) to flag a positive case, whereas radiologists took a median of 16·21 min (11·67-25·71) to draft a report and 23·06 min (15·67-39·20) to release a report. With regard to the identification of increases in lesion burden, AI achieved a sensitivity of 0·962 (95% CI 0·947-1·000) and a specificity of 0·875 (95 %CI 0·833-0·923). The agreement between AI and the radiologist panel was high (Cohen's kappa coefficient 0·839, 95% CI 0·718-0·940). Interpretation: A deep learning algorithm for triaging patients with suspected COVID-19 at fever clinics was developed and externally validated. Given its high accuracy across populations with varied COVID-19 prevalence, integration of this system into the standard clinical workflow could expedite identification of chest CT scans with imaging indications of COVID-19. Funding: Special Project for Emergency of the Science and Technology Department of Hubei Province, China.

COVID-19 in Wuhan: Sociodemographic characteristics and hospital support measures associated with the immediate psychological impact on healthcare workers.

BACKGROUND: The outbreak of COVID-19 has laid unprecedented threats and challenges to health workers (HWs) in Wuhan, China. We aimed to assess the sociodemographic characteristics and hospital support measures associated with the immediate psychological impact on HWs at Tongji Hospital in Wuhan during COVID-19 outbreak. METHODS: We conducted a single-center, cross-sectional survey of HWs via online questionnaires between February 8th and 10th, 2020. We evaluated stress, depression and anxiety by IES-R, PHQ-9, and GAD-7, respectively. We also designed a questionnaire to assess the perceptions of threat of COVID-19, and the satisfactions of the hospital's support measures. Multivariate logistic regressions were used to identify associated variables of acute stress, depression, and anxiety. FINDINGS: We received 5062 completed questionnaires (response rate, 77.1%). 29.8%, 13.5% and 24.1% HWs reported stress, depression and anxiety symptoms. Women (odds ratio [OR], 1.31; 95% CI, 0.47-0.97; p = 0.032), years of working >10 years (OR, 2.02; 95% CI, 1.47-2.79; p<0.001), concomitant chronic diseases (OR, 1.51; 95% CI, 1.27-1.80; p<0.001), history of mental disorders (OR, 3.27; 95% CI, 1.77-6.05; p<0.001), family members or relatives confirmed or suspected (OR, 1.23; 95% CI, 1.02-1.48; p = 0.03), hospital-based and department-based care (OR, 0.76; 95% CI, 0.60-0.97; p = 0.024) and full coverage of all departments for avoiding nosocomial infection (OR, 0.69; 95% CI, 0.53-0.89; p = 0.004) were associated with stress. INTERPRETATION: Women and those who have more than 10 years of working, concomitant chronic diseases, history of mental disorders, and family members or relatives confirmed or suspected are susceptible to stress, depression and anxiety among HWs during the pandemic. In addition, since HWs often have a greater stigma against mental problems than the general public, it is worthwhile to address the needs of the HWs during this pandemic and to provide appropriate psychological supports for those people at high risk of mental problems.

Safety, tolerability, and immunogenicity of a recombinant adenovirus type-5 vectored COVID-19 vaccine: a dose-escalation, open-label, non-randomised, first-in-human trial.

BACKGROUND: A vaccine to protect against COVID-19 is urgently needed. We aimed to assess the safety, tolerability, and immunogenicity of a recombinant adenovirus type-5 (Ad5) vectored COVID-19 vaccine expressing the spike glycoprotein of a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) strain. METHODS: We did a dose-escalation, single-centre, open-label, non-randomised, phase 1 trial of an Ad5 vectored COVID-19 vaccine in Wuhan, China. Healthy adults aged between 18 and 60 years were sequentially enrolled and allocated to one of three dose groups (5 × 1010, 1 × 1011, and 1·5 × 1011 viral particles) to receive an intramuscular injection of vaccine. The primary outcome was adverse events in the 7 days post-vaccination. Safety was assessed over 28 days post-vaccination. Specific antibodies were measured with ELISA, and the neutralising antibody responses induced by vaccination were detected with SARS-CoV-2 virus neutralisation and pseudovirus neutralisation tests. T-cell responses were assessed by enzyme-linked immunospot and flow-cytometry assays. This study is registered with ClinicalTrials.gov, NCT04313127. FINDINGS: Between March 16 and March 27, 2020, we screened 195 individuals for eligibility. Of them, 108 participants (51% male, 49% female; mean age 36·3 years) were recruited and received the low dose (n=36), middle dose (n=36), or high dose (n=36) of the vaccine. All enrolled participants were included in the analysis. At least one adverse reaction within the first 7 days after the vaccination was reported in 30 (83%) participants in the low dose group, 30 (83%) participants in the middle dose group, and 27 (75%) participants in the high dose group. The most common injection site adverse reaction was pain, which was reported in 58 (54%) vaccine recipients, and the most commonly reported systematic adverse reactions were fever (50 [46%]), fatigue (47 [44%]), headache (42 [39%]), and muscle pain (18 [17%]. Most adverse reactions that were reported in all dose groups were mild or moderate in severity. No serious adverse event was noted within 28 days post-vaccination. ELISA antibodies and neutralising antibodies increased significantly at day 14, and peaked 28 days post-vaccination. Specific T-cell response peaked at day 14 post-vaccination. INTERPRETATION: The Ad5 vectored COVID-19 vaccine is tolerable and immunogenic at 28 days post-vaccination. Humoral responses against SARS-CoV-2 peaked at day 28 post-vaccination in healthy adults, and rapid specific T-cell responses were noted from day 14 post-vaccination. Our findings suggest that the Ad5 vectored COVID-19 vaccine warrants further investigation. FUNDING: National Key R&D Program of China, National Science and Technology Major Project, and CanSino Biologics.