Predictive modeling of morbidity and mortality in COVID-19 hospitalized patients and its clinical implications.
Ontology highlight
ABSTRACT:
SUBMITTER: Wang JM
PROVIDER: S-EPMC7724684 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundThe COVID-19 pandemic began in early 2021 and placed significant strains on health care systems worldwide. There remains a compelling need to analyze factors that are predictive for patients at elevated risk of morbidity and mortality.ObjectiveThe goal of this retrospective study of patients who tested positive with COVID-19 and were treated at NYU (New York University) Langone Health was to identify clinical markers predictive of disease severity in order to assist in clinical decision triage and to provide additional biological insights into disease progression.MethodsThe clinical activity of 3740 patients at NYU Langone Hospital was obtained between January and August 2020; patient data were deidentified. Models were trained on clinical data during different parts of their hospital stay to predict three clinical outcomes: deceased, ventilated, or admitted to the intensive care unit (ICU).ResultsThe XGBoost (eXtreme Gradient Boosting) model that was trained on clinical data from the final 24 hours excelled at predicting mortality (area under the curve [AUC]=0.92; specificity=86%; and sensitivity=85%). Respiration rate was the most important feature, followed by SpO2 (peripheral oxygen saturation) and being aged 75 years and over. Performance of this model to predict the deceased outcome extended 5 days prior, with AUC=0.81, specificity=70%, and sensitivity=75%. When only using clinical data from the first 24 hours, AUCs of 0.79, 0.80, and 0.77 were obtained for deceased, ventilated, or ICU-admitted outcomes, respectively. Although respiration rate and SpO2 levels offered the highest feature importance, other canonical markers, including diabetic history, age, and temperature, offered minimal gain. When lab values were incorporated, prediction of mortality benefited the most from blood urea nitrogen and lactate dehydrogenase (LDH). Features that were predictive of morbidity included LDH, calcium, glucose, and C-reactive protein.ConclusionsTogether, this work summarizes efforts to systematically examine the importance of a wide range of features across different endpoint outcomes and at different hospitalization time points.
| S-EPMC8274681 | biostudies-literature
Project description:BackgroundThe novel coronavirus SARS-CoV-2 and its associated disease, COVID-19, have caused worldwide disruption, leading countries to take drastic measures to address the progression of the disease. As SARS-CoV-2 continues to spread, hospitals are struggling to allocate resources to patients who are most at risk. In this context, it has become important to develop models that can accurately predict the severity of infection of hospitalized patients to help guide triage, planning, and resource allocation.ObjectiveThe aim of this study was to develop accurate models to predict the mortality of hospitalized patients with COVID-19 using basic demographics and easily obtainable laboratory data.MethodsWe performed a retrospective study of 375 hospitalized patients with COVID-19 in Wuhan, China. The patients were randomly split into derivation and validation cohorts. Regularized logistic regression and support vector machine classifiers were trained on the derivation cohort, and accuracy metrics (F1 scores) were computed on the validation cohort. Two types of models were developed: the first type used laboratory findings from the entire length of the patient's hospital stay, and the second type used laboratory findings that were obtained no later than 12 hours after admission. The models were further validated on a multicenter external cohort of 542 patients.ResultsOf the 375 patients with COVID-19, 174 (46.4%) died of the infection. The study cohort was composed of 224/375 men (59.7%) and 151/375 women (40.3%), with a mean age of 58.83 years (SD 16.46). The models developed using data from throughout the patients' length of stay demonstrated accuracies as high as 97%, whereas the models with admission laboratory variables possessed accuracies of up to 93%. The latter models predicted patient outcomes an average of 11.5 days in advance. Key variables such as lactate dehydrogenase, high-sensitivity C-reactive protein, and percentage of lymphocytes in the blood were indicated by the models. In line with previous studies, age was also found to be an important variable in predicting mortality. In particular, the mean age of patients who survived COVID-19 infection (50.23 years, SD 15.02) was significantly lower than the mean age of patients who died of the infection (68.75 years, SD 11.83; P<.001).ConclusionsMachine learning models can be successfully employed to accurately predict outcomes of patients with COVID-19. Our models achieved high accuracies and could predict outcomes more than one week in advance; this promising result suggests that these models can be highly useful for resource allocation in hospitals.
| S-EPMC7572117 | biostudies-literature
Project description: Not available
| S-BSST1269 | biostudies-other
Project description:This SuperSeries is composed of the SubSeries listed below.
2021-07-24 | GSE180743 | GEO
Project description:Coronavirus disease 2019 (COVID-19) can be asymptomatic or lead to a wide spectrum of symptoms, ranging from mild upper respiratory system involvement to acute respiratory distress syndrome, multi-organ damage and death. In this study, we explored the potential of microRNAs (miRNA) in delineating patient condition and in predicting clinical outcome. Analysis of the circulating miRNA profile of COVID-19 patients, sampled at different hospitalization intervals after admission, allowed to identify miR-144-3p as a dynamically regulated miRNA in response to COVID-19.
2023-01-27 | GSE195898 | GEO
Project description:BackgroundInflammatory markers are often elevated in patients with COVID-19. The objective of this study is to assess the prognostic capability of these tests in predicting clinical outcomes.MethodsThis was a retrospective cohort study including all patients at least 16 years old with COVID-19 who were admitted from one of five Emergency Departments between March 6th and April 4th, 2020. We included 1123 laboratory-confirmed cases of COVID-19. We analyzed white blood cell count (WBC), absolute lymphocyte count (ALC), lactate dehydrogenase (LDH), C-reactive protein (CRP), procalcitonin (PCT), D-dimer, ferritin, and erythrocyte sedimentation rate (ESR). We looked at clinical outcomes including death, the need for endotracheal intubation (ETT), the need for renal replacement therapy (RRT), and ICU admission. We report Spearman's ρ2 and statistical significance for each correlation with outcomes. We also report positive predictive value, negative predictive value, sensitivity, specificity, positive likelihood ratios, and negative likelihood ratios.ResultsThe mean age of our patient population was 62 (SD 16). Thirty-seven percent of patients self-reported Spanish/Hispanic/Latino ethnicity, 47% reported their race as Black or African-American, and 10% reported their race as non-Hispanic white. Inter-rater reliability was 96%. There was no laboratory value that had both sensitivity and specificity of at least 0.90, or that had a positive predictive value and negative predictive value of at least 0.90, or that had likelihood ratios that could reliably predict a severe course of disease.ConclusionInflammatory markers drawn within 48 h of arrival, though often correlated with clinical outcomes, are not individually highly predictive of which patients in a predominantly older and minority population will die or require intubation, RRT, or ICU admission.
| S-EPMC7685065 | biostudies-literature
Project description:Background The scope of pericardial involvement in COVID-19 infection is unknown. We aimed to evaluate the prevalence, associates, and clinical impact of pericardial involvement in hospitalized patients with COVID-19. Methods and Results Consecutive patients with COVID-19 underwent clinical and echocardiographic examination, irrespective of clinical indication, within 48 hours as part of a prospective predefined protocol. Protocol included clinical symptoms and signs suggestive of pericarditis, calculation of modified early warning score, ECG and echocardiographic assessment for pericardial effusion, left and right ventricular systolic and diastolic function, and hemodynamics. We identified predictors of mortality and assessed the adjunctive value of pericardial effusion on top of clinical and echocardiographic parameters. The study included 530 patients. Pericardial effusion was found in 75 (14%), but only 17 patients (3.2%) fulfilled the criteria for acute pericarditis. Pericardial effusion was independently associated with modified early warning score, brain natriuretic peptide, and right ventricular function. It was associated with excess mortality (hazard ratio [HR], 2.44; P=0.0005) in nonadjusted analysis. In multivariate analysis adjusted for modified early warning score and echocardiographic and hemodynamic parameters, it was marginally associated with mortality (HR, 1.86; P=0.06) and improvement in the model fit (P=0.07). Combined assessment for pericardial effusion with modified early warning score, left ventricular ejection fraction, and tricuspid annular plane systolic excursion was an independent predictor of outcome (HR, 1.86; P=0.02) and improved model fit (P=0.02). Conclusions In hospitalized patients with COVID-19, pericardial effusion is prevalent, but rarely attributable to acute pericarditis. It is associated with myocardial dysfunction and mortality. A limited echocardiographic examination, including left ventricular ejection fraction, tricuspid annular plane systolic excursion, and assessment for pericardial effusion, can contribute to outcome prediction.
| S-EPMC9075494 | biostudies-literature
Project description:BackgroundAlthough there have been several studies investigating prognostic factors for mortality in COVID-19, there have been lack of studies in low- and middle-income countries, including Indonesia. To date, the country has the highest mortality rate among Asian countries.ObjectiveWe sought to identify the prognostic factors of mortality in hospitalized patients with COVID-19 in Jakarta.MethodsIn this retrospective cohort study, we included all adult inpatients (≥18 years old) with confirmed COVID-19 from Koja General Hospital (North Jakarta, Indonesia) who had been hospitalized between March 20th and July 31st, 2020. Demographic, clinical, laboratory, and radiology data were extracted from the medical records and compared between survivors and non-survivors. Univariate and multivariate logistic regression analysis were used to explore the prognostic factors associated with in-hospital death.ResultsTwo hundred forty-three patients were included in the study, of whom 32 died. Comorbid of hypertension (OR 3.59; 95% CI 1.12-11.48; p = 0.031), obesity (OR 6.34; 95% CI 1.68-23.98; p = 0.007), immediate need of HFNC and/or IMV (OR 64.93; 95% CI 11.08-380.61; p < 0.001), abnormal RDW (OR 3.68; 95% CI 1.09-12.34; p = 0.035), ALC < 1,000/µL (OR 3.51; 95% CI 1.08-11.44; p = 0.038), D-dimer > 500 ng/mL (OR 9.36; 95% CI 1.53-57.12; p = 0.015) on admission, as well as chloroquine treatment (OR 3.61; 95% CI 1.09-11.99; p = 0.036) were associated with greater risk of overall mortality in COVID-19 patients. The likelihood of mortality increased with increasing number of prognostic factors.ConclusionThe potential prognostic factors of hypertension, obesity, immediate need of HFNC and/or IMV, abnormal RDW, ALC < 1,000/µL, D-dimer > 500 ng/mL, and chloroquine treatment could help clinicians to identify COVID-19 patients with poor prognosis at an early stage.
| S-EPMC8855735 | biostudies-literature
Project description:BackgroundStudying the characteristics of hospitalized Coronavirus Disease 2019 (COVID-19) patients is vital for understanding the disease and preparing for future outbreaks. The aim of this study was to analyze and describe the clinical profiles and factors associated with mortality among COVID-19 patients admitted to Jimma Medical Center COVID-19 Treatment Center (JMC CTC) in Ethiopia.MethodsAll confirmed COVID-19 patients admitted to JMC CTC between 17 April 2020 and 05 March 2022 were included in this study. Socio-demographic data, clinical information, and outcome variables were collected retrospectively from medical records and COVID-19 database at the hospital. Bivariable and multivariable analyses were performed to determine factors associated with COVID-19 severity and mortality. A P-value < 0.05 was considered statistically significant.ResultsA total of 542 confirmed COVID-19 patients were admitted to JMC CTC, of which 322 (59.4%) were male. Their median age was 48 years (IQR 32-64). About 51% (n = 277) of them had severe COVID-19 upon admission. Patients with hypertension [AOR: 2.8 (95% CI: 1.02-7.7, p = 0.046)], diabetes [AOR: 8.8 (95% CI: 1.2-17.3, p = 0.039)], and underlying respiratory diseases [AOR: 18.8 (95% CI: 2.06-71.51, p = 0.009)] were more likely to present with severe COVID-19 cases. Overall, 129 (23.8%) died in the hospital. Death rate was higher among patients admitted with severe disease [AHR = 5.5 (3.07-9.9) p < 0.001)] and those with comorbidities such as hypertension [AHR = 3.5 (2.28-5.41), p < 0.001], underlying respiratory disease [AHR = 3.4 (1.97-5.94), p < 0.001], cardiovascular disease (CVDs) [AHR = 2.8 (1.73-4.55), p < 0.001], and kidney diseases [AHR = 3.7 (2.3-5.96), p < 0.001].ConclusionAbout half of COVID-19 cases admitted to the hospital had severe disease upon admission. Comorbidities such as hypertension, diabetes, and respiratory diseases were linked to severe illness. COVID-19 admissions were associated with high inpatient mortality, particularly among those with severe disease and comorbidities.
| S-EPMC11370003 | biostudies-literature
Project description:BackgroundTwelve-lead electrocardiogram (ECG) represents the first-line approach for cardiovascular assessment in patients with Covid-19.ObjectivesWe sought to describe and compare admission ECG findings in 3 different hospital settings: intensive-care unit (ICU) (invasive ventilatory support), respiratory care unit (RCU) (non-invasive ventilatory support) and Covid-19 dedicated internal-medicine unit (IMU) (oxygen supplement with or without high flow). We also aimed to assess the prognostic impact of admission ECG variables in Covid-19 patients.MethodsWe retrospectively analyzed the admission 12-lead ECGs of 1124 consecutive patients hospitalized for respiratory distress and Covid-19 in a single III-level hospital. Age, gender, main clinical data and in-hospital survival were recorded.Results548 patients were hospitalized in IMU, 361 in RCU, 215 in ICU. Arrhythmias in general were less frequently found in RCU (16% vs 26%, p<0.001). Deaths occurred more frequently in ICU patients (43% vs 20-21%, p<0.001). After pooling predictors of mortality (age, intensity of care setting, heart rate, ST-elevation, QTc prolongation, Q-waves, right bundle branch block, and atrial fibrillation), the risk of in-hospital death can be estimated by using a derived score. Three zones of mortality risk can be identified: <5%, score <5 points; 5-50%, score 5-10, and >50%, score >10 points. The accuracy of the score assessed at ROC curve analysis was 0.791.ConclusionsECG differences at admission can be found in Covid-19 patients according to different clinical settings and intensity of care. A simplified score derived from few clinical and ECG variables may be helpful in stratifying the risk of in-hospital mortality.
| S-EPMC8872825 | biostudies-literature