CVD22: Explainable artificial intelligence determination of the relationship of troponin to D-Dimer, mortality, and CK-MB in COVID-19 patients.

BACKGROUND AND PURPOSE: COVID-19, which emerged in Wuhan (China), is one of the deadliest and fastest-spreading pandemics as of the end of 2019. According to the World Health Organization (WHO), there are more than 100 million infectious cases worldwide. Therefore, research models are crucial for managing the pandemic scenario. However, because the behavior of this epidemic is so complex and difficult to understand, an effective model must not only produce accurate predictive results but must also have a clear explanation that enables human experts to act proactively. For this reason, an innovative study has been planned to diagnose Troponin levels in the COVID-19 process with explainable white box algorithms to reach a clear explanation. METHODS: Using the pandemic data provided by Erzurum Training and Research Hospital (decision number: 2022/13-145), an interpretable explanation of Troponin data was provided in the COVID-19 process with SHApley Additive exPlanations (SHAP) algorithms. Five machine learning (ML) algorithms were developed. Model performances were determined based on training, test accuracies, precision, F1-score, recall, and AUC (Area Under the Curve) values. Feature importance was estimated according to Shapley values by applying the SHApley Additive exPlanations (SHAP) method to the model with high accuracy. The model created with Streamlit v.3.9 was integrated into the interface with the name CVD22. RESULTS: Among the five-machine learning (ML) models created with pandemic data, the best model was selected with the values of 1.0, 0.83, 0.86, 0.83, 0.80, and 0.91 in train and test accuracy, precision, F1-score, recall, and AUC values, respectively. As a result of feature selection and SHApley Additive exPlanations (SHAP) algorithms applied to the XGBoost model, it was determined that DDimer mean, mortality, CKMB (creatine kinase myocardial band), and Glucose were the features with the highest importance over the model estimation. CONCLUSIONS: Recent advances in new explainable artificial intelligence (XAI) models have successfully made it possible to predict the future using large historical datasets. Therefore, throughout the ongoing pandemic, CVD22 (https://cvd22covid.streamlitapp.com/) can be used as a guide to help authorities or medical professionals make the best decisions quickly.

A significant predictor of in-hospital and long-term mortality and progression in COVID-19 patients: The end-stage liver disease (MELD) score model.

The aim of this study is to investigate the relationship between the model for end-stage liver disease (MELD) score and disease progression and mortality in COVID-19 patients. The files of 4213 patients over the age of 18 who were hospitalized with the diagnosis of COVID-19 between March 20, 2020 and May 1, 2021 were retrospectively scanned. Sociodemographic characteristics, chronic diseases, hemogram and biochemical parameters at the time they were diagnosed with COVID-19 of the patients, duration of hospitalization, duration of intensive care unit (ICU), duration of intubation, in-hospital mortality from COVID-19 and outside-hospital mortality for another reason (within the last 1 year) and recurrent hospitalization (within the last 1 year) were recorded. The MELD scores of the patients were calculated. Two groups were formed as MELD score < 10 and MELD score ≥ 10. The rate of ICU, in-hospital mortality from COVID-19 and outside-hospital mortality from other causes, intubation rate, and recurrent hospitalization were significantly higher in the MELD ≥ 10 group. The duration of ICU, hospitalization, intubation were significantly higher in the MELD ≥ 10 group (p < 0.001). As a result of Univariate and Multivariate analysis, MELD score was found to be the independent predictors of ICU, in-hospital mortality, intubation, and recurrent hospitalization (p < 0.001). MELD score 18.5 predicted ICU with 99% sensitivity and 100% specificity (area under curve [AUC]: 0.740, 95% confidence interval [CI]: 0.717-0.763, p < 0.001) also MELD score 18.5 predicted in-hospital mortality with 99% sensitivity and 100% specificity (AUC: 0.797, 95% CI: 0.775-0.818, p < 0.001). The MELD score was found to be the independent predictors of in-hospital mortality, ICU admission, and intubation in COVID-19 patients.

The relationship between pericardial effusion and pulmonary involvement, prognosis, mortality in COVID-19 patients.

Aim    Comprehensive studies on the coexistence of COVID-19 and pericardial effusion (PEff) are limited. In this study, we investigated the relationship between pneumonia severity and PEff, predisposing factors, and the effect of PEff on clinical prognosis and mortality in COVID-19 patients.Material and methods    Between March and November 2020, 5 575 patients were followed up in our pandemic hospital due to COVID-19. 3 794 patients with positive polymerase chain reaction (PCR) test results and thoraxcomputerized tomography (CT) imaging at admission were included in the study. The clinical and demographic characteristics, CT images, hematological and biochemical parameters of these patients were retrospectively examined. Pulmonary involvement of 3794 patients was divided into three groups and its relationship with PEff was investigated retrospectively.Results    There were 560 patients who did not have pulmonary involvement, 2 639 patients with pulmonary involvement below 50 %, and 595 patients with 50 % or more pulmonary involvement. As pulmonaryinvolvement or the severity of the disease increased, male gender and advanced age become statistically significant. The mean age of patients with PEff was higher, and PEff was more common in males. Patients with PEff had more comorbid diseases and significantly elevated serum cardiac and inflammatory biomarkers. The need for intensive care and mortality rates were higher in these patients. While the in-hospital mortality rate was 56.9 % in patients with PEff and pulmonary involvement above 50 %, in-hospital mortality rate was 34.4 % in patients with pulmonary involvement above 50 % and without PEff (p<0.001).The presence of PEff during admission for COVID-19 disease, the appearance of PEff or increase in the degree of PEff during follow-up were closely related to mortality and prognosis.Conclusion    As the severity of pulmonary involvement or the clinical severity of the disease increased, PEff occurred in patients or the degree of PEff increased. The clinical prognosis of patients presenting with PEff was quite poor, and the frequency of intensive care admissions and mortality were significantly higher. PEff was an important finding in the follow-up and management of patients with COVID-19, and it reflected the clinical prognosis.

Does new onset and pre-existing atrial fibrillation predict mortality in COVID-19 patients?

BACKGROUND: Coronavirus Disease-2019 (COVID-19), caused by Severe Acute Respiratory Syndrome-Coronavirus-2, still remains prevalent and severe. We aimed to evaluate the effects of pre-existing atrial fibrillation and new-onset atrial fibrillation (NOAF) on the clinical severity and mortality of COVID-19. RESULTS: Between April and December 2020, 5577 patients with positive PCR and/or COVID-19 compatible findings in computed tomography hospitalized were enrolled retrospectively. Total and in-hospital mortality, need for intensive care unit (ICU), need for mechanical ventilation, and recurrent hospitalization results of 286 patients with pre-existing AF before hospitalization and 82 patients with NOAF during hospitalization were evaluated. Preexisting AF was associated with a 2-fold increase in total and in-hospital mortality [OR (2.16 (1.62-2.89), 2.02 (1.48-2.76), P < 0.001, respectively]. NOAF was associated with a 14-fold increase in total mortality and a 12-fold increase in in-hospital mortality [OR(14.72 (9.22-23.5), 12.56 (8.02-19.68), P < 0.001], respectively]. However, pre-existing AF and NOAF resulted in increased ICU admission, mechanical ventilation, and recurrent hospitalization. In the Cox regression analysis, NOAF was observed as an independent risk factor for mortality. CONCLUSIONS: Pre-existing AF and in-hospital NOAF were associated with increased mortality and severity in hospitalized COVID-19 patients. In addition, NOAF was observed as an independent prognostic indicator in terms of total mortality.

Relationship Between Blood Lipid Levels and Mortality in Hospitalized COVID-19 Patients.

People with comorbid conditions are at increased risk of developing severe/fatal coronavirus disease 2019 (COVID-19). We aimed to investigate the relationship between lipid levels and mortality in patients hospitalized for COVID-19 infection. In this retrospective study, we collected the details of 5274 COVID-19 patients who were diagnosed using the polymerase chain reaction and/or computed tomography and were hospitalized between March and November 2020. Patients (n = 4118) whose blood lipid levels were checked within the first 24 h after hospitalization were included in the study. Multivariable cox proportional hazards regression was used to assess the relationship between lipid variables such as low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG) and death. There was a statistically significant association between LDL-C, HDL-C, and TG levels and the risk of death (P =.002, <.001, and .035, respectively). Low and high LDL-C, low HDL-C, and high TG levels were negatively associated with COVID-19-related mortality. Blood lipid levels may be useful predictors of mortality in COVID-19 patients.

Significance of MPV, RDW and PDW with the Severity and Mortality of COVID-19 and Effects of Acetylsalicylic Acid Use.

We aimed to investigate association between mean platelet volume (MVP), platelet distribution width (PDW) and red cell distribution width (RDW) and mortality in patients with COVID-19 and find out in which patients the use of acetylsalicylic acid (ASA) affects the prognosis due to the effect of MPV on thromboxan A2. A total of 5142 patients were divided into those followed in the intensive care unit (ICU) and those followed in the ward. Patient medical records were examined retrospectively. ROC analysis showed that the area under curve (AUC) values were 0.714, 0.750, 0.843 for MPV, RDW and D-Dimer, the cutoff value was 10.45fl, 43.65fl, 500.2 ng/mL respectively. (all P < .001). Survival analysis showed that patients with MPV >10.45 f/l and D-Dimer >500.2 ng/mL, treatment with ASA had lower in-hospital and 180-day mortality than patients without ASA in ICU patients (HR = 0.773; 95% CI = 0.595-0.992; P = .048, HR = 0.763; 95% CI = 0.590-0.987; P = .036). Administration of low-dose ASA in addition to anti-coagulant according to MPV and D-dimer levels reduces mortality.

Association between previous anticoagulant use and mortality among hospitalized patients with COVID-19.

Although COVID-19 disease primarily affects the respiratory system, it has been seen in many studies that it causes thromboembolic (TE) events in many tissues and organs. So that, to prevent TE can reduce mortality and morbidity. In this context, this study aimed to investigate the relationship between the previous use of warfarin or other new direct oral anticoagulants (OAC) and mortality in patients hospitalized with a diagnosis of COVID-19 before hospitalization. A total of 5575 patients who were diagnosed with COVID-19 were hospitalized and started treatment between March 21 and November 30, 2020 were included in the study. The primary outcome was in-hospital all-cause mortality. A retrospective cohort study design was planned. Patients were followed up until death or censoring on November 30, 2020. The candidate predictors for primary outcome should be clinically and biologically plausible, and their relationships with all-cause death should be demonstrated in previous studies. We considered all candidate predictors included in the model in accordance with these principles. The main candidate predictor was previous OAC use. The primary analysis method was to compare the time to deaths of patients using and not using previous OAC by a multivariable Cox proportional hazard model (CPHM). In the CPHM, previous OAC use was found to be associated with a significantly lower mortality risk (adjusted hazard ratio 0.62, 95% CI 0.42-0.92, p = 0.030). In hospitalized COVID-19 patients, in patients who previously used anticoagulantswas associated with lower risk of in-hospital death than in those who did not.