scholarly journals A Machine-learning Parsimonious Multivariable Predictive Model of Mortality Risk in Patients With Covid-19

2021 ◽  
Author(s):  
Rita Murri ◽  
Jacopo Lenkowicz ◽  
Carlotta Masciocchi ◽  
Chiara Iacomini ◽  
Massimo Fantoni ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Risk ◽  
Mortality Rate ◽  
Risk Score ◽  
Predictive Value ◽  
Cross Validation ◽  
C Reactive Protein ◽  
Training Cohort ◽  
Reactive Protein ◽  
Fold Cross Validation

Abstract BackgroundThe COVID-19 pandemic is impressively challenging the healthcare system. Several prognostic models have been validated but few of them are implemented in daily practice. The objective of the study was to validate a machine-learning risk prediction model using easy-to-obtain parameters, potentially available at home, to help identifying patients with COVID-19 who are at higher risk of death.MethodsThe training cohort included all patients admitted to Fondazione Policlinico Gemelli with COVID-19 from March 5, 2020 to November 5, 2020. Afterwards, the model was tested on all patients admitted to the same hospital with COVID-19 from November 6, 2020 to February 5 2021. The primary outcome was in-hospital mortality.The out-of-sample performance of the model was estimated from the training set in terms of Area under the Receiving Operator Curve (AUROC) and classification matrix statistics by averaging the results of 5-fold cross validation repeated 3-times and comparing the results with those obtained on the test set. An explanation analysis of the model, based on the SHapley Additive exPlanations (SHAP), is also presented. To assess the subsequent time evolution, the change in paO2/FiO2 (P/F) at 48 hours after the baseline measurement was plotted against its baseline value.ResultsAmong the 921 patients included in the training cohort, 120 died (13%). Variables selected for the model were age, platelet count, SpO2, blood urea nitrogen (BUN), hemoglobin, C-reactive protein, neutrophil count, and sodium. The results of the 5-fold cross-validation repeated 3-times gave AUROC of 0.87, and statistics of the classification matrix to the Youden index as follows: sensitivity 0.840, specificity 0.774, negative predictive value 0.971. Then, the model was tested on a new population (n=1463) in which the mortality rate was 22.6 %. The test model showed AUROC 0.818, sensitivity 0.813, specificity 0.650, negative predictive value 0.922. Considering the first quartile of the predicted risk score (low-risk score group), the mortality rate was 1.6%, 17.8% in the second and third quartile (high-risk score group) and 53.5% in the fourth quartile (very high-risk score group). The three risk score groups showed good discrimination for the P/F value at admission, and a positive correlation was found for the low-risk class to P/F at 48 hours after admission (adjusted R-squared= 0.48).ConclusionsWe developed a predictive model of death for people with SARS-CoV-2 infection by including only easy-to-obtain variables (abnormal blood count, BUN, C-reactive protein, sodium and lower SpO2). It demonstrated good accuracy and high power of discrimination. The simplicity of the model makes the risk prediction applicable for patients at home, in the Emergency Department, or during hospitalization.

scholarly journals A machine-learning parsimonious multivariable predictive model of mortality risk in patients with Covid-19

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Rita Murri ◽  
Jacopo Lenkowicz ◽  
Carlotta Masciocchi ◽  
Chiara Iacomini ◽  
Massimo Fantoni ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Risk ◽  
Risk Score ◽  
Case Fatality Rate ◽  
Cross Validation ◽  
Case Fatality ◽  
Fatality Rate ◽  
C Reactive Protein ◽  
Training Cohort ◽  
Reactive Protein

AbstractThe COVID-19 pandemic is impressively challenging the healthcare system. Several prognostic models have been validated but few of them are implemented in daily practice. The objective of the study was to validate a machine-learning risk prediction model using easy-to-obtain parameters to help to identify patients with COVID-19 who are at higher risk of death. The training cohort included all patients admitted to Fondazione Policlinico Gemelli with COVID-19 from March 5, 2020, to November 5, 2020. Afterward, the model was tested on all patients admitted to the same hospital with COVID-19 from November 6, 2020, to February 5, 2021. The primary outcome was in-hospital case-fatality risk. The out-of-sample performance of the model was estimated from the training set in terms of Area under the Receiving Operator Curve (AUROC) and classification matrix statistics by averaging the results of fivefold cross validation repeated 3-times and comparing the results with those obtained on the test set. An explanation analysis of the model, based on the SHapley Additive exPlanations (SHAP), is also presented. To assess the subsequent time evolution, the change in paO2/FiO2 (P/F) at 48 h after the baseline measurement was plotted against its baseline value. Among the 921 patients included in the training cohort, 120 died (13%). Variables selected for the model were age, platelet count, SpO2, blood urea nitrogen (BUN), hemoglobin, C-reactive protein, neutrophil count, and sodium. The results of the fivefold cross-validation repeated 3-times gave AUROC of 0.87, and statistics of the classification matrix to the Youden index as follows: sensitivity 0.840, specificity 0.774, negative predictive value 0.971. Then, the model was tested on a new population (n = 1463) in which the case-fatality rate was 22.6%. The test model showed AUROC 0.818, sensitivity 0.813, specificity 0.650, negative predictive value 0.922. Considering the first quartile of the predicted risk score (low-risk score group), the case-fatality rate was 1.6%, 17.8% in the second and third quartile (high-risk score group) and 53.5% in the fourth quartile (very high-risk score group). The three risk score groups showed good discrimination for the P/F value at admission, and a positive correlation was found for the low-risk class to P/F at 48 h after admission (adjusted R-squared = 0.48). We developed a predictive model of death for people with SARS-CoV-2 infection by including only easy-to-obtain variables (abnormal blood count, BUN, C-reactive protein, sodium and lower SpO2). It demonstrated good accuracy and high power of discrimination. The simplicity of the model makes the risk prediction applicable for patients in the Emergency Department, or during hospitalization. Although it is reasonable to assume that the model is also applicable in not-hospitalized persons, only appropriate studies can assess the accuracy of the model also for persons at home.

scholarly journals Establishment and Validation of Nomogram Based on Combination of Pretreatment C-Reactive Protein/Albumin Ratio–EBV DNA Grade in Nasopharyngeal Carcinoma Patients Who Received Concurrent Chemoradiotherapy

2021 ◽  
Vol 11 ◽  
Author(s):  
Zhang-Zan Huang ◽  
Wen Wen ◽  
Xin Hua ◽  
Chen-Ge Song ◽  
Xi-Wen Bi ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Cross Validation ◽  
Concurrent Chemoradiotherapy ◽  
Multivariate Analyses ◽  
C Reactive Protein ◽  
Training Set ◽  
Albumin Ratio ◽  
Reactive Protein ◽  
Ebv Dna ◽  
Fold Cross Validation

BackgroundA higher ratio of pretreatment C-reactive protein/albumin ratio (CAR) is associated with poor prognosis in nasopharyngeal carcinoma (NPC), and Epstein–Barr virus (EBV) DNA level is known to not only participate in the occurrence of nasopharyngeal carcinoma but also affect the development and prognosis of the disease. Herein, we proposed that a combination of both these markers could improve the predictive prognostic ability.MethodsIn all, 842 NPC patients who received concurrent chemoradiotherapy (CCRT) were entered in this study. We collected all patients’ blood samples and EBV DNA copy numbers within one week before any treatment. Receiver operating characteristic (ROC) curve was used to determine the optimal cut-off. We employed the Kaplan–Meier method for survival analyses and the univariate and multivariate analyses (Cox proportional hazards regression model) for statistical analysis. A nomogram was constructed based on multivariate analyses results of the validation set. The model was internally validated using 1000 bootstrap samples to avoid overfitting. Another validation of 10-fold cross-validation was also applied. Calibration curves and concordance index (C-index) were calculated to determine predictive and discriminatory capacity.ResultsIn the whole cohort, we observed that higher CAR, EBV DNA level, and CAR-EBV DNA (C-E) grade were associated with shorter overall survival (OS) and distant metastasis-free survival (DMFS) (all P<0.05). In univariate and multivariate analyses, C-E grade was an independent prognostic factor (all P<0.05). In the training set, we gained the similar results with the whole set. According to multivariate analyses of the training set, we constructed a nomogram. The results of bootstrap samples and 10-fold cross-validation showed favorable predictive efficacy. And calibration curves of the model provided credibility to its predictive capability.ConclusionC-E grade was confirmed as an independent prognostic predictor in patients with NPC who received CCRT. Higher level of pretreatment C-E grade could signify a higher risk of metastasis and shorter OS. The prognostic nomogram based on C-E grade was dependable in nasopharyngeal carcinoma patients.

scholarly journals S64. COGNITIVE IMPAIRMENTS AND PREDICTION OF FUNCTIONAL OUTCOME IN INDIVIDUALS AT CLINICAL HIGH-RISK FOR PSYCHOSIS

2020 ◽  
Vol 46 (Supplement_1) ◽  
pp. S57-S58
Author(s):  
Kate Haining ◽  
Gina Brunner ◽  
Ruchika Gajwani ◽  
Joachim Gross ◽  
Andrew Gumley ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Risk ◽  
Functional Outcome ◽  
Processing Speed ◽  
Cross Validation ◽  
Clinical High Risk ◽  
Role Functioning ◽  
Gaf Scores ◽  
Fold Cross Validation

Abstract Background Research in individuals at clinical-high risk for psychosis (CHR-P) has focused on developing algorithms to predict transition to psychosis. However, it is becoming increasingly important to address other outcomes, such as the level of functioning of CHR-P participants. To address this important question, this study investigated the relationship between baseline cognitive performance and functional outcome between 6–12 months in a sample of CHR-P individuals using a machine-learning approach to identify features that are predictive of long-term functional impairments. Methods Data was available for 111 CHR-P individuals at 6–12 months follow-up. In addition, 47 CHR-negative (CHR-N) participants who did not meet CHR criteria and 55 healthy controls (HCs) were recruited. CHR-P status was assessed using the Comprehensive Assessment of At-Risk Mental States (CAARMS) and the Schizophrenia Proneness Instrument, Adult version (SPI-A). Cognitive assessments included the Brief Assessment of Cognition in Schizophrenia (BACS) and the Penn Computerized Neurocognitive Battery (CNB). Global, social and role functioning scales were used to measure functional status. CHR-P individuals were divided into good functional outcome (GFO, GAF ≥ 65) and poor functional outcome groups (PFO, GAF < 65). Feature selection was performed using LASSO regression with the LARS algorithm and 10-fold cross validation with GAF scores at baseline as the outcome variable. The following features were identified as predictors of GAF scores at baseline: verbal memory, verbal fluency, attention, emotion recognition, social and role functioning and SPI-A distress. This model explained 47% of the variance in baseline GAF scores. In the next step, Support Vector Machines (SVM), Linear Discriminant Analysis (LDA), Logistic Regression (LR), Gaussian Naïve Bayes (GNB), and Random Forest (RF) classifiers with 10-fold cross validation were then trained on those features with GAF category at follow-up used as the binary label column. Models were compared using a calculated score incorporating area under the curve (AUC), accuracy, and AUC consistency across runs, whereby AUC was given a higher weighting than accuracy due to class imbalance. Results CHR-P individuals had slower motor speed, reduced attention and processing speed and increased emotion recognition reaction times (RTs) compared to HCs and reduced attention and processing speed compared to CHR-Ns. At follow-up, 66% of CHR-P individuals had PFO. LDA emerged as the strongest classifier, showing a mean AUC of 0.75 (SD = 0.15), indicating acceptable classification performance for GAF category at follow-up. PFO was detected with a sensitivity of 75% and specificity of 58%, with a total mean weighted accuracy of 68%. Discussion The CHR-P state was associated with significant impairments in cognition, highlighting the importance of interventions such as cognitive remediation in this population. Our data suggest that the development of features using machine learning approaches is effective in predicting functional outcomes in CHR-P individuals. Greater levels of accuracy, sensitivity and specificity might be achieved by increasing training sets and validating the classifier with external data sets. Indeed, machine learning methods have potential given that trained classifiers can easily be shared online, thus enabling clinical professionals to make individualised predictions.

Prediction of K562 Cells Functional Inhibitors Based on Machine Learning Approaches

2020 ◽  
Vol 25 (40) ◽  
pp. 4296-4302 ◽  
Author(s):  
Yuan Zhang ◽  
Zhenyan Han ◽  
Qian Gao ◽  
Xiaoyi Bai ◽  
Chi Zhang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Inclusion Bodies ◽  
Cross Validation ◽  
Independent Set ◽  
K562 Cells ◽  
Machine Learning Algorithms ◽  
Learning Approaches ◽  
Validation Test ◽  
Excess Number ◽  
Fold Cross Validation

Background: β thalassemia is a common monogenic genetic disease that is very harmful to human health. The disease arises is due to the deletion of or defects in β-globin, which reduces synthesis of the β-globin chain, resulting in a relatively excess number of α-chains. The formation of inclusion bodies deposited on the cell membrane causes a decrease in the ability of red blood cells to deform and a group of hereditary haemolytic diseases caused by massive destruction in the spleen. Methods: In this work, machine learning algorithms were employed to build a prediction model for inhibitors against K562 based on 117 inhibitors and 190 non-inhibitors. Results: The overall accuracy (ACC) of a 10-fold cross-validation test and an independent set test using Adaboost were 83.1% and 78.0%, respectively, surpassing Bayes Net, Random Forest, Random Tree, C4.5, SVM, KNN and Bagging. Conclusion: This study indicated that Adaboost could be applied to build a learning model in the prediction of inhibitors against K526 cells.

scholarly journals Predictive value of C-reactive protein for radiographic spinal progression in axial spondyloarthritis in dependence on genetic determinants of fibrin clot formation and fibrinolysis

RMD Open ◽  
2021 ◽  
Vol 7 (2) ◽  
pp. e001751
Author(s):  
Berthold Hoppe ◽  
Christian Schwedler ◽  
Hildrun Haibel ◽  
Maryna Verba ◽  
Fabian Proft ◽  
...  
Keyword(s):  
Predictive Value ◽  
Axial Spondyloarthritis ◽  
Fibrin Clot ◽  
Clot Formation ◽  
Genetic Determinants ◽  
C Reactive Protein ◽  
Inception Cohort ◽  
Reactive Protein ◽  
A Subunit ◽  
Fibrin Clots

ObjectiveGenetic determinants of fibrin clot formation and fibrinolysis have an impact on local and systemic inflammatory response. The aim of the present study was to assess whether coagulation-related genotypes affect the predictive value of C-reactive protein (CRP) in regards of radiographic spinal progression in axial spondyloarthritis (axSpA).MethodsTwo hundred and eight patients with axSpA from the German Spondyloarthritis Inception Cohort were characterised for genotypes of α-fibrinogen, β-fibrinogen (FGB) and γ-fibrinogen, factor XIII A-subunit (F13A) and α2-antiplasmin (A2AP). The relation between CRP levels and radiographic spinal progression defined as worsening of the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) by ≥2 points over 2 years was assessed in dependence on the respective genetic background in logistic regression analyses.ResultsOverall, CRP was associated with mSASSS progression ≥2 points: time-averaged CRP ≥10 mg/L, OR: 3.32, 95% CI 1.35 to 8.13. After stratification for coagulation-related genotypes, CRP was strongly associated with mSASSS progression in individuals predisposed to form loose, fibrinolysis-susceptible fibrin clots (FGB rs1800790GG, OR: 6.86, 95% CI 2.08 to 22.6; A2AP 6Trp, OR: 5.86, 95% CI 1.63 to 21.0; F13A 34Leu, OR: 8.72, 95% CI 1.69 to 45.1), while in genotypes predisposing to stable fibrin clots, the association was absent or weak (FGB rs1800790A, OR: 0.83, 95% CI 0.14 to 4.84; A2AP 6Arg/Arg, OR: 1.47, 95% CI 0.35 to 6.19; F13A 34Val/Val, OR: 1.72, 95% CI 0.52 to 5.71).ConclusionsElevated CRP levels seem to be clearly associated with radiographic spinal progression only if patients are predisposed for loose fibrin clots with high susceptibility to fibrinolysis.

Predictive value of C-reactive protein at different cutoff levels in acute appendicitis

2005 ◽  
Vol 23 (4) ◽  
pp. 449-453 ◽  
Author(s):  
Han-Ping Wu ◽  
Ching-Yuang Lin ◽  
Chin-Fu Chang ◽  
Yu-Jun Chang ◽  
Chin-Yi Huang
Keyword(s):  
Acute Appendicitis ◽  
Predictive Value ◽  
C Reactive Protein ◽  
Reactive Protein ◽  
Cutoff Levels

C-Reactive Protein as A Fungal Infection Marker in Acute Leukemia Patients

2021 ◽  
Vol 27 (2) ◽  
pp. 212
Author(s):  
Brigitte Rina Aninda Sidharta ◽  
JB. Suparyatmo ◽  
Avanti Fitri Astuti
Keyword(s):  
Acute Leukemia ◽  
Predictive Value ◽  
Bacterial Infections ◽  
Fungal Infections ◽  
Lymphoblastic Leukemia ◽  
Negative Likelihood Ratio ◽  
Receiver Operating Curve ◽  
High Morbidity ◽  
C Reactive Protein ◽  
Reactive Protein

Invasive Fungal Infections (IFIs) can cause serious problems in cancer patients and may result in high morbidity andmortality. C-reactive protein levels increase in response to injury, infection, and inflammation. C-reactive protein increasesin bacterial infections (mean of 32 mg/L) and in fungal infections (mean of 9 mg/L). This study aimed to determineC-Reactive Protein (CRP) as a marker of fungal infections in patients with acute leukemia by establishing cut-off values ofCRP. This study was an observational analytical study with a cross-sectional approach and was carried out at the Departmentof Clinical Pathology and Microbiology of Dr. Moewardi Hospital in Surakarta from May until August 2019. The inclusioncriteria were patients with acute leukemia who were willing to participate in this study, while exclusion criteria were patientswith liver disease. There were 61 samples consisting of 30 male and 31 female patients with ages ranging from 1 to 70 years.Fifty-four patients (88.5%) were diagnosed with Acute Lymphoblastic Leukemia (ALL) and 30 (49.18%) were in themaintenance phase. The risk factors found in those patients were neutropenia 50-1500 μL (23.8%), use of intravenous line(22%), and corticosteroid therapy for more than one week (20.9%). The median of CRP in the group of patients with positiveculture results was 11.20 mg/L (11.20-26.23 mg/L) and negative culture results in 0.38 mg/L (0.01-18.63 mg/L). The cut-offvalue of CRP using the Receiver Operating Curve (ROC) was 9.54 mg/L (area under curve 0.996 and p. 0.026), with a sensitivityof 100%, specificity of 93.2%, Positive Predictive Value (PPV) of 33.3%, Negative Predictive Value (PPV) of 100%, PositiveLikelihood Ratio (PLR) of 1.08, Negative Likelihood Ratio (NLR) of 0 and accuracy of 93.4%. C-reactive protein can be used asa screening marker for fungal infections in patients with acute leukemia.

Sign in / Sign up

Export Citation Format

Share Document