Dataset Information

Unbiased identification of clinical characteristics predictive of COVID-19 severity.

ABSTRACT: There is currently limited clinical ability to identify COVID-19 patients at risk for severe outcomes. To unbiasedly identify metrics associated with severe outcomes in COVID-19 patients, we conducted a retrospective study of 835 COVID-19 positive patients at a single academic medical center between March 10, 2020 and October 13, 2020. As of December 1, 2020, 656 (79%) patients required hospitalization and 149 (18%) died. Unbiased comparisons of all clinical characteristics and mortality revealed that abnormal pH (OR 8.54, 95% CI 5.34-13.6), abnormal creatinine (OR 6.94, 95% CI 4.22-11.4), and abnormal PTT (OR 4.78, 95% CI 3.11-7.33) were most significantly associated with mortality. Correlation with ordinal severity scores confirmed these associations, in addition to associations between respiratory rate (Spearman's rho = -0.56), absolute neutrophil count (Spearman's rho = -0.5), and C-reactive protein (Spearman's rho = 0.59) with disease severity. Unsupervised principal component analysis and machine learning model classification of patient demographics, laboratory results, medications, comorbidities, signs and symptoms, and vitals are capable of separating patients on the basis of COVID-19 mortality (AUC 0.82). This retrospective analysis identifies laboratory and clinical metrics most relevant to predict COVID-19 severity.

SUBMITTER: Akama-Garren EH

PROVIDER: S-EPMC8178667 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ImportanceSince the emergence of COVID-19 in central China, sub-Saharan African countries, with the exception of South Africa, have been relatively spared during the COVID-19 pandemic. Consequently, few descriptive studies from this region are available.ObjectiveTo describe the clinical characteristics and outcomes of patients with COVID-19 infection in Gabon, from March to June 2020.Design, setting, and participantsA single-center, cross-sectional study of 837 patients with COVID-19 was conducted from March to June 2020 in the Armed Forces Hospital in Libreville, Gabon.Main outcomes and measuresDemographic and clinical characteristics and imaging findings of hospitalized patients with COVID-19.ResultsOf the 837 patients enrolled, 572 (68.3%) were men, and 264 (31.5%) were women (male to female ratio, 2:1); the median (interquartile range [IQR]) age was 35 (30-45) years (mean [SD] age, 38.0 [12.2] years. The mortality rate associated with COVID-19 was low (1.4%). Of these 837 patients, 524 (62.6%) were categorized as having no symptoms, 282 (33.7%) as having mild symptoms, and 31 (3.7%) as having severe symptoms. Patients with severe symptoms were older (mean [SD] age, 46.1 [14.7] years) than patients with mild symptoms (mean [SD] age, 41.3 [12.5] years) and those with no symptoms (mean [SD] age, 35.7 [11.3] years) (Kruskal-Wallis χ22 = 53.5; P < .001). History of diabetes was the principal risk factor associated with both severe symptoms in 5 of 31 patients (16.1%) and mild symptoms in 11 of 282 (3.9%) compared with no symptoms in 5 of 524 (0.9%) (Pearson χ22 = 30.9; P < .001). Patients with severe symptoms and a fatal outcome were older (mean [SD] age, 53.4 [15.1] years) than survivors (mean [SD] age, 41.5 [12.9] years) (t20.83 = 2.2; P = .03).Conclusions and relevanceIn this single-center, cross-sectional study in Libreville, Gabon, the mortality rate associated with COVID-19 infection from March to June 2020 was low, and patients who died of COVID-19 infection were younger on average than reported elsewhere, possibly reflecting a smaller elderly population in Gabon.

Project description:Using RNA-seq and high-resolution mass spectrometry we performed a comprehensive systems analysis on 128 plasma and leukocyte samples from hospitalized patients with or without COVID-19 (n=102 and 26 respectively) and with differing degrees of disease severity. We generated abundance measurements for over 17,000 transcripts, proteins, metabolites, and lipids and compiled them with clinical data into a curated relational database. This resource offers the unique opportunity to perform systems analysis and cross-ome correlations to both molecules and patient outcomes. In total 219 molecular features were mapped with high significance to COVID-19 status and severity, including those involved in processes such as complement system activation, dysregulated lipid transport, and B cell activation. In one example, we detected a trio of covarying molecules – citrate, plasmenyl-phosphatidylcholines, and gelsolin (GSN) – that offer both pathophysiological insight and potential novel therapeutic targets. Further, our data revealed in some cases, and supported in others, that several biological processes were dysregulated in COVID-19 patients including vessel damage, platelet activation and degranulation, blood coagulation, and acute phase response. For example, we observed that the coagulation-related protein, cellular fibronectin (cFN), was highly increased within COVID-19 patients and provide new evidence that the upregulated proteoform stems from endothelial cells, consistent with endothelial injury as a major activator of the coagulation cascade. The abundance of prothrombin, which is cleaved to form thrombin during clotting, was significantly reduced and correlated with severity and might help to explain the hyper coagulative environment of SARS-CoV-2 infection. From transcriptomic analysis of leukocytes, we concluded that COVID-19 patients with acute respiratory distress syndrome (ARDS) demonstrated a phenotype that overlapped with, but was distinct from, that found in patients with non-COVID-19-ARDS. To aid in the global efforts toward elucidation of disease pathophysiology and therapeutic development, we created a web-based tool with interactive visualizations allowing for easy navigation of this systems-level compendium of biomolecule abundance in relation to COVID-19 status and severity. Finally, we leveraged these multi-omic data to predict COVID-19 patient outcomes with machine learning, which highlighted the predictive power of these expansive molecular measurements beyond the standardized clinical estimate of 10-year survival Charlson score.

Dataset Information

Unbiased identification of clinical characteristics predictive of COVID-19 severity.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets