Background: The outbreak of coronavirus disease 2019 (COVID-19) has globally strained medical resources and caused significant mortality.Objective: To develop and validate a machine-learning model based on clinical features for severity risk assessment and triage for COVID-19 patients at hospital admission.Method: 725 patients were used to train and validate the model. This included a retrospective cohort from Wuhan, China of 299 hospitalised COVID-19 patients from 23 December 2019 to 13 February 2020, and five cohorts with 426 patients from eight centres in China, Italy and Belgium from 20 February 2020 to 21 March 2020. The main outcome was the onset of severe or critical illness during hospitalisation. Model performances were quantified using the area under the receiver operating characteristic curve (AUC) and metrics derived from the confusion matrix.Results: In the retrospective cohort, the median age was 50 years and 137 (45.8%) were male. In the five test cohorts, the median age was 62 years and 236 (55.4%) were male. The model was prospectively validated on five cohorts yielding AUCs ranging from 0.84 to 0.93, with accuracies ranging from 74.4% to 87.5%, sensitivities ranging from 75.0% to 96.9%, and specificities ranging from 55.0% to 88.0%, most of which performed better than the pneumonia severity index. The cut-off values of the low-, medium- and high-risk probabilities were 0.21 and 0.80. The online calculators can be found atwww.covid19risk.ai.Conclusion: The machine-learning model, nomogram and online calculator might be useful to access the onset of severe and critical illness among COVID-19 patients and triage at hospital admission.

SARS-CoV-2 is associated with a 3.4% mortality rate in patients with severe disease. The pathogenesis of severe cases remains unknown. We performed an in-depth prospective analysis of immune and inflammation markers in two patients with severe COVID-19 disease from presentation to convalescence. Peripheral blood from 18 SARS-CoV-2-infected patients, 9 with severe and 9 with mild COVID-19 disease, was obtained at admission and analyzed for T-cell activation profile, myeloid-derived suppressor cells (MDSCs) and cytokine profiles. MDSC functionality was tested in vitro. In four severe and in four mild patients, a longitudinal analysis was performed daily from the day of admission to the early convalescent phase. Early after admission severe patients showed neutrophilia, lymphopenia, increase in effector T cells, a persisting higher expression of CD95 on T cells, higher serum concentration of IL-6 and TGF-β, and a cytotoxic profile of NK and T cells compared with mild patients, suggesting a highly engaged immune response. Massive expansion of MDSCs was observed, up to 90% of total circulating mononuclear cells in patients with severe disease, and up to 25% in the patients with mild disease; the frequency decreasing with recovery. MDSCs suppressed T-cell functions, dampening excessive immune response. MDSCs decline at convalescent phase was associated to a reduction in TGF-β and to an increase of inflammatory cytokines in plasma samples. Substantial expansion of suppressor cells is seen in patients with severe COVID-19. Further studies are required to define their roles in reducing the excessive activation/inflammation, protection, influencing disease progression, potential to serve as biomarkers of disease severity, and new targets for immune and host-directed therapeutic approaches.

Several recent case reports have described common early chest imaging findings of lung pathology caused by 2019 novel Coronavirus (SARS-COV2) which appear to be similar to those seen previously in SARS-CoV and MERS-CoV infected patients. We present some remarkable imaging findings of the first two patients identified in Italy with COVID-19 infection travelling from Wuhan, China. The follow-up with chest X-Rays and CT scans was also included, showing a progressive adult respiratory distress syndrome (ARDS). Moderate to severe progression of the lung infiltrates, with increasing percentage of high-density infiltrates sustained by a bilateral and multi-segmental extension of lung opacities, were seen. During the follow-up, apart from pleural effusions, a tubular and enlarged appearance of pulmonary vessels with a sudden caliber reduction was seen, mainly found in the dichotomic tracts, where the center of a new insurgent pulmonary lesion was seen. It could be an early alert radiological sign to predict initial lung deterioration. Another uncommon element was the presence of mediastinal lymphadenopathy with short-axis oval nodes. Although only two patients have been studied, these findings are consistent with the radiological pattern described in literature. Finally, the pulmonary vessels enlargement in areas where new lung infiltrates develop in the follow-up CT scan, could describe an early predictor radiological sign of lung impairment.

Data concerning the transmission of the novel severe acute respiratory syndrome coronavirus (SARS-CoV-2) in paucisymptomatic patients are lacking. We report an Italian paucisymptomatic case of coronavirus disease 2019 with multiple biological samples positive for SARS-CoV-2. This case was detected using the World Health Organization protocol on cases and contact investigation. Current discharge criteria and the impact of extra-pulmonary SARS-CoV-2 samples are discussed.