Kidney-transplant recipients appear to be at particularly high risk for critical Covid-19 illness due to chronic immunosuppression and coexisting conditions.¹ At Montefiore Medical Center, we identified 36 consecutive adult kidney-transplant recipients who tested positive for Covid-19 between March 16 and April 1, 2020. A total of 26 recipients (72%) were male, and the median age was 60 years (range, 32 to 77). Fourteen recipients (39%) were black, and 15 recipients (42%) were Hispanic. Twenty-seven recipients (75%) had received a deceased-donor kidney; 34 recipients (94%) had hypertension, 25 (69%) had diabetes mellitus, 13 (36%) had a history of smoking tobacco or were current smokers, and 6 (17%) had heart disease. Thirty-five of the patients (97%) were receiving tacrolimus, 34 (94%) were receiving prednisone, and 31 (86%) were receiving mycophenolate mofetil or mycophenolic acid.

BACKGROUND: Coronavirus disease 2019 (COVID-19) is more benign in children compared with adults for unknown reasons. This contrasts with other respiratory viruses where disease manifestations are often more severe in children. We hypothesize that a more robust early innate immune response to SARS coronavirus 2 (SARS-CoV-2) protects against severe disease. METHODS: Clinical outcomes, SARS-CoV-2 viral copies, and cellular gene expression were compared in nasopharyngeal swabs obtained at the time of presentation to the emergency department from 12 children and 27 adults using bulk RNA sequencing and quantitative reverse-transcription PCR. Total protein, cytokines, and anti–SARS-CoV-2 IgG and IgA were quantified in nasal fluid. RESULTS: SARS-CoV-2 copies, angiotensin-converting enzyme 2, and TMPRSS2 gene expression were similar in children and adults, but children displayed higher expression of genes associated with IFN signaling, NLRP3 inflammasome, and other innate pathways. Higher levels of IFN-α2, IFN-γ, IP-10, IL-8, and IL-1β protein were detected in nasal fluid in children versus adults. Children also expressed higher levels of genes associated with immune cells, whereas expression of those associated with epithelial cells did not differ in children versus adults. Anti–SARS-CoV-2 IgA and IgG were detected at similar levels in nasal fluid from both groups. None of the children required supplemental oxygen, whereas 7 adults did ( P = 0.03); 4 adults died. CONCLUSION: These findings provide direct evidence of a more vigorous early mucosal immune response in children compared with adults and suggest that this contributes to favorable clinical outcomes. FUNDING: NIH grants R01 AI134367, UL1 TR002556, T32 AI007501, T32GM007288, P30 AI124414; an Albert Einstein College of Medicine Dean’s COVID-19 Pilot Research Award; and the Eric J. Heyer, MD, PhD Translational Research Pilot Project Award.

Background: During a pandemic, it is important for clinicians to stratify patients and decide who receives limited medical resources. Machine learning models have been proposed to accurately predict COVID-19 disease severity. Previous studies have typically tested only one machine learning algorithm and limited performance evaluation to area under the curve analysis. To obtain the best results possible, it may be important to test different machine learning algorithms to find the best prediction model. Objective: In this study, we aimed to use automated machine learning (autoML) to train various machine learning algorithms. We selected the model that best predicted patients’ chances of surviving a SARS-CoV-2 infection. In addition, we identified which variables (ie, vital signs, biomarkers, comorbidities, etc) were the most influential in generating an accurate model. Methods: Data were retrospectively collected from all patients who tested positive for COVID-19 at our institution between March 1 and July 3, 2020. We collected 48 variables from each patient within 36 hours before or after the index time (ie, real-time polymerase chain reaction positivity). Patients were followed for 30 days or until death. Patients’ data were used to build 20 machine learning models with various algorithms via autoML. The performance of machine learning models was measured by analyzing the area under the precision-recall curve (AUPCR). Subsequently, we established model interpretability via Shapley additive explanation and partial dependence plots to identify and rank variables that drove model predictions. Afterward, we conducted dimensionality reduction to extract the 10 most influential variables. AutoML models were retrained by only using these 10 variables, and the output models were evaluated against the model that used 48 variables. Results: Data from 4313 patients were used to develop the models. The best model that was generated by using autoML and 48 variables was the stacked ensemble model (AUPRC=0.807). The two best independent models were the gradient boost machine and extreme gradient boost models, which had an AUPRC of 0.803 and 0.793, respectively. The deep learning model (AUPRC=0.73) was substantially inferior to the other models. The 10 most influential variables for generating high-performing models were systolic and diastolic blood pressure, age, pulse oximetry level, blood urea nitrogen level, lactate dehydrogenase level, D-dimer level, troponin level, respiratory rate, and Charlson comorbidity score. After the autoML models were retrained with these 10 variables, the stacked ensemble model still had the best performance (AUPRC=0.791). Conclusions: We used autoML to develop high-performing models that predicted the survival of patients with COVID-19. In addition, we identified important variables that correlated with mortality. This is proof of concept that autoML is an efficient, effective, and informative method for generating machine learning–based clinical decision support tools.

The COVID-19 pandemic caused by the new SARS-CoV-2 coronavirus has imposed severe challenges on laboratories in their effort to achieve sufficient diagnostic testing capability for identifying infected individuals. In this study, we report the analytical and clinical performance characteristics of a new, high-throughput, fully automated nucleic acid amplification test system for the detection of SARS-CoV-2. The assay utilizes target capture, transcription-mediated amplification, and acridinium ester-labeled probe chemistry on the automated Panther system to directly amplify and detect two separate target sequences in the open reading frame 1ab (ORF1ab) region of the SARS-CoV-2 RNA genome. The probit 95% limit of detection of the assay was determined to be 0.004 50% tissue culture infective dose (TCID₅₀)/ml using inactivated virus and 25 copies/ml (c/ml) using synthetic in vitro transcript RNA targets. Analytical sensitivity (100% detection) was confirmed to be 83 to 194 c/ml using three commercially available SARS-CoV-2 nucleic acid controls. No cross-reactivity or interference was observed with testing of six related human coronaviruses, as well as 24 other viral, fungal, and bacterial pathogens, at high titers. Clinical nasopharyngeal swab specimen testing (n = 140) showed 100%, 98.7%, and 99.3% positive, negative, and overall agreement, respectively, with a validated reverse transcription-PCR nucleic acid amplification test (NAAT) for SARS-CoV-2 RNA. These results provide validation evidence for a sensitive and specific method for pandemic-scale automated molecular diagnostic testing for SARS-CoV-2.