As of March 24, 2020, novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been responsible for 379,661 infection cases with 16,428 deaths globally, and the number is still increasing rapidly. Herein, we present four critically ill patients with SARS-CoV-2 infection who received supportive care and convalescent plasma. Although all four patients (including a pregnant woman) recovered from SARS-CoV-2 infection eventually, randomized trials are needed to eliminate the effect of other treatments and investigate the safety and efficacy of convalescent plasma therapy.

We read great interest in the risk factors of critical or mortal COVID-19 cases, recently reported by Ye et al. in this journal. 1 Here we paid more attention about the long-term lung sequelae among survivors of severe COVID-19. With more than 21 million people worldwide recovered from COVID-19, early analysis suggested a high rate of patients had residual abnormal lung function and fibrotic remodeling on CT, especially in survivors of severe SARS-CoV-2 associated pneumonia. 2 , 3 These might contribute to longterm impairment of lung function or even lung transplants. The early identification of patients at higher risk of lung injury and fibrotic damage is critical. 4 Therefore, we performed an observational cohort study that compared fibrosis and non-fibrosis group to investigate the potential indicators for post-fibrosis. The two-center retrospective study was approved by the institutional review board of Xianning Central Hospital and Huangshi Central Hospital, both in Hubei Province. The informed consent was waived. From December 19, 2019 to March 5, 2020, a total of 430 consecutive patients with positive RT-PCR were reviewed. Finally 81 survivors who recovered from severe COVID-19 pneumonia were enrolled. The median hospitalization was 26 days; all had at least three follow-up CT scans after discharge, and the median period between the discharge and the latest CT scan was 58 days (IQR: 25-46). Pulmonary fibrosis was diagnosed based on the extensive and persistent fibrotic changes, including parenchymal bands, irregular interfaces, reticular opacities, and traction bronchiectasis with or without honeycombing on the follow-up CT scans. CT scores were evaluated by two experienced cardiothoracic radiologists independently, and quantified by the percentage of high attenuation area using thresholds with pixels between 0 and −700 HU via Chest Imaging Platform (http://chestimagingplatform.org/). Fibrosis grouping was reached by consensus. Comparative analysis were performed with R software, covering age, sex, prior medical history, signs and symptoms, laboratory data, oxygen supply, ICU admission, and treatments. The statistical difference was assessed with the unpaired, 2-tailed chisquare test for categorical variables and t-test or Mann-Whitney for continuous variables. P < 0.05 indicated a statistically significant difference.

Background: Ultrasound (US) examination is helpful in the differential diagnosis of thyroid nodules (malignant vs. benign), but its accuracy relies heavily on examiner experience. Therefore, the aim of this study was to develop a less subjective diagnostic model aided by machine learning. Methods: A total of 2064 thyroid nodules (2032 patients, 695 male; M age = 45.25-13.49 years) met all of the following inclusion criteria: (i) hemi-or total thyroidectomy, (ii) maximum nodule diameter 2.5 cm, (iii) examination by conventional US and real-time elastography within one month before surgery, and (iv) no previous thyroid surgery or percutaneous thermotherapy. Models were developed using 60% of randomly selected samples based on nine commonly used algorithms, and validated using the remaining 40% of cases. All models function with a validation data set that has a pretest probability of malignancy of 10%. The models were refined with machine learning that consisted of 1000 repetitions of derivatization and validation, and compared to diagnosis by an experienced radiologist. Sensitivity, specificity, accuracy, and area under the curve (AUC) were calculated. Results: A random forest algorithm led to the best diagnostic model, which performed better than radiologist diagnosis based on conventional US only (AUC = 0.924 [confidence interval (CI) 0.895-0.953] vs. 0.834 [CI 0.815-0.853]) and based on both conventional US and real-time elastography (AUC = 0.938 [CI 0.914-0.961] vs. 0.843 [CI 0.829-0.857]). Conclusions: Machine-learning algorithms based on US examinations, particularly the random forest classifier, may diagnose malignant thyroid nodules better than radiologists.