Prediction of Disease Progression of COVID-19 Based upon Machine Learning.

BACKGROUND: Since December 2019, COVID-19 has spread throughout the world. Clinical outcomes of COVID-19 patients vary among infected individuals. Therefore, it is vital to identify patients at high risk of disease progression. METHODS: In this retrospective, multicenter cohort study, COVID-19 patients from Huoshenshan Hospital and Taikang Tongji Hospital (Wuhan, China) were included. Clinical features showing significant differences between the severe and nonsevere groups were screened out by univariate analysis. Then, these features were used to generate classifier models to predict whether a COVID-19 case would be severe or nonsevere based on machine learning. Two test sets of data from the two hospitals were gathered to evaluate the predictive performance of the models. RESULTS: A total of 455 patients were included, and 21 features showing significant differences between the severe and nonsevere groups were selected for the training and validation set. The optimal subset, with eleven features in the k-nearest neighbor model, obtained the highest area under the curve (AUC) value among the four models in the validation set. D-dimer, CRP, and age were the three most important features in the optimal-feature subsets. The highest AUC value was obtained using a support vector-machine model for a test set from Huoshenshan Hospital. Software for predicting disease progression based on machine learning was developed. CONCLUSION: The predictive models were successfully established based on machine learning, and achieved satisfactory predictive performance of disease progression with optimal-feature subsets.

Association of Early-Phase In-Hospital Glycemic Fluctuation With Mortality in Adult Patients With Coronavirus Disease 2019.

OBJECTIVE: To investigate the association of in-hospital early-phase glycemic control with adverse outcomes among inpatients with coronavirus disease 2019 (COVID-19) in Wuhan, China. RESEARCH DESIGN AND METHODS: The study is a large case series, and data were obtained regarding consecutive patients hospitalized with COVID-19 in the Central Hospital of Wuhan between 2 January and 15 February 2020. All patients with definite outcomes (death or discharge) were included. Demographic, clinical, treatment, and laboratory information were extracted from electronic medical records. We collected daily fasting glucose data from standard morning fasting blood biochemistry to determine glycemic status and fluctuation (calculated as the square root of the variance of daily fasting glucose levels) during the 1st week of hospitalization. RESULTS: A total of 548 patients were included in the study (median age 57 years; 298 [54%] were women, and n = 99 had diabetes [18%]), 215 suffered acute respiratory distress syndrome (ARDS), 489 survived, and 59 died. Patients who had higher mean levels of glucose during their 1st week of hospitalization were older and more likely to have a comorbidity and abnormal laboratory markers, prolonged hospital stays, increased expenses, and greater risks of severe pneumonia, ARDS, and death. Compared with patients with the lowest quartile of glycemic fluctuation, those who had the highest quartile of fluctuation magnitude had an increased risk of ARDS (risk ratio 1.97 [95% CI 1.01, 4.04]) and mortality (hazard ratio 2.73 [95% CI 1.06, 7.73]). CONCLUSIONS: These results may have implications for optimizing glycemic control strategies in COVID-19 patients during the early phase of hospitalization.

Association between iron status and the risk of adverse outcomes in COVID-19.

BACKGROUND & AIMS: Iron is an essential trace element to almost all organism, and the delicate balance between host defend system and viral proliferation plays an important role in infective conditions. While the association of the iron metabolism with the prognosis of COVID-19 remains poorly understood. We aimed to estimate the associations of systemic iron metabolism parameters with the severity and risks of adverse outcomes in COVID-19. METHODS: In this retrospective cohort study, we included 158 confirmed COVID-19 patients in Tongji Hospital, Wuhan, China (27 January to 5 April, 2020). Demographic data, comorbidities, laboratory examinations, treatments, and clinical outcomes were all collected. Multivariable Poisson regression was used to estimate the association of iron parameter levels with the severity and risks of adverse outcomes in COVID-19 patients. RESULTS: We identified 60 (38%) severe cases in 158 COVID-19 patients. The median age was 63 years (interquartile range [IQR]: 54-73) and the median length of hospital stay was 28 days (IQR: 17-40). After adjusting for age, sex, IL-6, and pre-existing comorbidities, all iron parameters were associated with the severity of COVID-19 with adjusted risk ratio of 0.42 [95% CI: 0.22-0.83], 4.38 [95% CI: 1.86-10.33], 0.19 [95% CI: 0.08-0.48], and 0.25 [95% CI: 0.10-0.58] for serum iron, ferritin, transferrin, and total iron-binding capacity, respectively. These iron indices were also related to the risk of ARDS, coagulopathy, acute cardiac injury, acute liver injury, and acute kidney injury in COVID-19 patients and high cytokine concentrations. CONCLUSIONS: Patients with low serum iron status likely suffered from severe condition and multiple-organ injury in COVID-19. The iron metabolism parameters might be risk factors and clinical biomarkers for COVID-19 prognosis.

Comparison of clinical characteristics between patients with coronavirus disease 2019 (COVID-19) who retested RT-PCR positive versus negative: a retrospective study of data from Nanjing.

BACKGROUND: The epidemiological and clinical characteristics of patients with coronavirus disease 2019 (COVID-19) have been reported. However, the prevalence of retesting positive by RT-PCR for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the associated patient characteristics, remain unclear. METHODS: We included 90 confirmed cases of COVID-19 treated in the Nanjing Public Health Center from January 20, 2020 to February 16, 2020 in this retrospective study. All patients completed treatment for COVID-19 and were retested by RT-PCR for SARS-CoV-2 4-20 days after completion of therapy. The clinical characteristics between patients with who retested positive versus negative by RT-PCR were compared, and the factors predictive of positive retesting were analyzed. Positive retesting was modeled with the area under the receiver operating characteristic curve (AUC). RESULTS: The age range of the study population was 0.8-97 years, and all patients were cured or showed improvement. A total of 10 (11%) patients retested positive by RT-PCR 4-20 days after completion of therapy. As compared with patients who retested negative, those who retested positive had a lower percentage of pre-admission fever, a higher percentage of post-admission fever, a lower percentage of bilateral lung infection, higher white blood cell (WBC) count and creatine phosphokinase, and lower hypersensitive c-reactive protein (hs-CRP), interleukin-6 and erythrocyte sedimentation rates (all P<0.05). Logistic regression analysis of the above eight key variables showed that lower hs-CRP and higher WBC were independently associated with positive retesting by RT-PCR. A combination of hs-CRP and WBC were predictive of positive retesting, with an AUC of 0.859. CONCLUSIONS: Patients with COVID-19 who retested positive by RT-PCR for SARS-CoV-2 had mild symptoms and better blood testing results. A combination of hs-CRP and WBC may predict positive retesting by RT-PCR; however, the sensitivity and specificity should be studied further.

Clinical Use of Short-Course and Low-Dose Corticosteroids in Patients With Non-severe COVID-19 During Pneumonia Progression.

Background: The emerging coronavirus disease 2019 (COVID-19) has become a serious public health concern with a high number of fatalities. It is unclear whether corticosteroids could be a candidate for an early intervention strategy for patients with COVID-19. Methods: In this retrospective cohort study, we analyzed data from 28 corticosteroid-treated patients with non-severe but advanced COVID-19, in which short-course and low-dose corticosteroids were administered because of unremitting or worsening clinical conditions during hospitalization. To compare the effect of corticosteroids on viral clearance, 44 corticosteroid-untreated patients were included as controls. Results: At the time of admission, corticosteroid-treated patients (n = 28) had a more advanced baseline illness compared with corticosteroid-untreated patients (n = 44), as reflected by poorer blood laboratory parameters (lymphocytes, C-reactive protein, and lactate dehydrogenase) and more extensive chest computed tomography (CT) abnormalities. Corticosteroids were given because of radiological evidence of pneumonia progression (26/28) and/or unremitting fever (22/28) after admission. The median time from illness onset to corticosteroid treatment was 9 days (IQR, 7-10). The median duration and accumulated dose of corticosteroid treatment were 4.5 days [interquartile range (IQR), 3-5] and 140 mg of methylprednisolone (IQR, 120-200). Intravenous immunoglobulin (20 g per day for 3-5 days) was co-administered with corticosteroids. With the corticosteroid treatment, all patients achieved an abatement of fever within 1 day, and 78.6% (22/28) of the patients achieved radiological remission when evaluated about 3 days later. Only one (3.6%) patient progressed to severe COVID-19, and all patients recovered and were discharged without any sequela. The median time from illness onset to viral clearance was similar, as compared with 44 corticosteroid-untreated patients with relatively milder disease [18 (IQR 14.3-23.5) days vs. 17 (IQR, 12-20) days, p = 0.252]. When adjusted for age, sex, underlying comorbidities, baseline blood laboratory parameters, viral load, and chest radiological findings, the causal hazard ratio of corticosteroid treatment for the viral clearance was 0.79 (95%CI, 0.48-1.30, p = 0.34). Conclusion: Short-course and low-dose applications of corticosteroids, when co-administered with intravenous immunoglobulin, in non-severe COVID-19 patients during the stage of clinical deterioration may possibly prevent disease progression, while having a negligible impact on the viral clearance.