Risk Prediction Model For Mortality Research Articles

A novel risk score incorporating frailty or congestion status predicted mortality in patients with heart failure with preserved ejection fraction.

Abstract Introduction In patients with heart failure (HF), frailty or congestion is strongly associated with clinical outcomes. However, existing risk prediction models for HF prognosis do not incorporate these important clinical factors. The estimated plasma volume status (ePVS) is an easy-to-calculate index to estimate plasma volume. which is associated with the prognosis in HF patients. Purpose This study intends to construct an inclusive risk prediction model for all-cause mortality in patients with HF with preserved ejection fraction (HFpEF), incorporating evaluations of frailty and congestion alongside traditional risk factors. Methods We analyzed data from 1212 patients (81±9 years, 55% female) registered in the PURSUIT-HFpEF (Prospective, Multicenter, Observational Study of Patients with Heart Failure with Preserved Ejection Fraction) study, a multicenter registry of patients hospitalized for acute decompensated HFpEF. We evaluated clinical factors including Clinical Frailty Scale (CFS) and blood test results collected before discharge. We calculated ePVS at discharge using the Duarte formula as ePVS (mL/g)=100×(1-hematocrit)/hemoglobin (g/dL). Patients were followed for median of 713 days (IQR 369-1088 days) to observe all-cause death. Patients were randomly assigned to the derivation and validation cohorts (2:1 ratio). Using the Bayesian Information Criterion (BIC), a stepwise model selection was executed on 51 clinical parameters, including CFS and ePVS, to identify independent predictors of mortality. A risk model for mortality was constructed based on these predictors and their coefficients. The accuracy of the risk score was validated using time-dependent ROC curves both in the derivation cohort (n=801) and in the validation cohort (n=411). Results During the follow-up period, all-cause death was observed in 315 patients (25.9%). Age, sex, CFS, NT-proBNP and ePVS were selected as independent predictors of all-cause death. The score shown in the Table was applied to each factor, and the sum of the scores was calculated as the risk score. The AUC of the risk score for predicting 1-year mortality in the derivation cohort was 0.79 (95% CI 0.74, 0.85). The optimal cut-off value of 6 provided a sensitivity and specificity of 74.9% and 74.4%, respectively. The AUC for 1-year mortality in the validation cohort was 0.70 (95% CI 0.63, 0.77). Conclusions The novel risk prediction model incorporating CFS and ePVS effectively predicted mortality in HFpEF. This model could be a valuable tool for risk stratification.Table

Development and validation of a predicting model for risk of mortality in heart failure patients with mildly reduced ejection fraction post discharge

Abstract Background Prognostic prediction for heart failure with mildly reduced ejection fraction (HFmrEF) patients remains challenging. We aimed to create and validate a mortality risk prediction model for HFmrEF patients at 6-month, 1-year, and 3-year post discharge. Methods Clinical data of 1691 HFmrEF patients registered in the heart failure registry from 2015 to 2020 at the Heart Center of our hospital were analyzed. Patients were assigned into a training (1183 patients) and validation cohort (508 patients) at 7:3 ratio. Predictive variables for mortality at various period post discharge were identified by the Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis. The predicting model was then established based on these variables. Model performance was evaluated with ROC curve and Decision Curve Analysis (DCA). Results Eight predictors were identified including age, NT-proBNP, hemoglobin levels, use of beta-blockers, ACEI/ARB, invasive ventilation, PCI procedures, and pulmonary artery systolic pressure values. The model achieved a C-index of 0.748 (training set) and 0.755 (validation set). Area Under the Curve (AUC) for training set at 6 months, 1 year, and 3 years were 0.813, 0.784, and 0.774, and AUC for validation set were 0.775, 0.744, and 0.770, respectively. The DCA analysis confirmed the favorable net benefit of the established nomogram model. Conclusion This predicting model might be used for post-discharge risk stratification and individual decision-making of post-discharge management in HFmrEF patients.

A machine learning algorithm-based risk prediction score for in-hospital/30-day mortality after adult cardiac surgery.

A study of the performance of in-hospital/30-day mortality risk prediction models using an alternative machine learning algorithm (XGBoost) in adults undergoing cardiac surgery. Retrospective analyses of prospectively routinely collected data on adult patients undergoing cardiac surgery in the UK from January 2012 to March 2019. Data were temporally split 70:30 into training and validation subsets. Independent mortality prediction models were created using sequential backward floating selection starting with 61 variables. Assessments of discrimination, calibration, and clinical utility of the resultant XGBoost model with 23 variables were then conducted. A total of 224,318 adults underwent cardiac surgery during the study period with a 2.76% (N = 6,100) mortality. In the testing cohort, there was good discrimination (area under the receiver operator curve 0.846, F1 0.277) and calibration (especially in high-risk patients). Decision curve analysis showed XGBoost-23 had a net benefit till a threshold probability of 60%. The most important variables were the type of operation, age, creatinine clearance, urgency of the procedure and the New York Heart Association score. Feature-selected XGBoost showed good discrimination, calibration and clinical benefit when predicting mortality post-cardiac surgery. Prospective external validation of a XGBoost-derived model performance is warranted.

S1122 Development and Validation of a Machine Learning Risk-Prediction Model for In-Hospital Mortality from Clostridiodes Difficile in Patients with Inflammatory Bowel Disease: National Inpatient Sample-Based Study

S1122 Development and Validation of a Machine Learning Risk-Prediction Model for In-Hospital Mortality from Clostridiodes Difficile in Patients with Inflammatory Bowel Disease: National Inpatient Sample-Based Study

S2225 Development and Validation of a Machine Learning Risk-Prediction Model for In-Hospital Mortality from C. difficile: National Inpatient Sample Study

S2225 Development and Validation of a Machine Learning Risk-Prediction Model for In-Hospital Mortality from C. difficile: National Inpatient Sample Study

A comparison of some existing and novel methods for integrating historical models to improve estimation of coefficients in logistic regression

Abstract Model integration refers to the process of incorporating a fitted historical model into the estimation of a current study to increase statistical efficiency. Integration can be challenging when the current model includes new covariates, leading to potential model misspecification. We present and evaluate seven existing and novel model integration techniques, which employ both likelihood constraints and Bayesian informative priors. Using a simulation study of logistic regression, we quantify how efficiency—assessed by bias and variance—changes with the sample sizes of both historical and current studies and in response to violations to transportability assumptions. We also apply these methods to a case study in which the goal is to use novel predictors to update a risk prediction model for in-hospital mortality among pediatric extracorporeal membrane oxygenation patients. Our simulation study and case study suggest that (i) when historical sample size is small, accounting for this statistical uncertainty is more efficient; (ii) all methods lose efficiency when there exist differences between the historical and current data-generating mechanisms; (iii) additional shrinkage to zero can improve efficiency in higher-dimensional settings but at the cost of bias in estimation.

Image-based ECG analyzing deep-learning algorithm to predict biological age and mortality risks: interethnic validation.

Cardiovascular risk assessment is a critical component of healthcare, guiding preventive and therapeutic strategies. In this study, we developed and evaluated an image-based electrocardiogram (ECG) analyzing an artificial intelligence (AI) model that estimates biological age and mortality risk. Using a dataset of 978 319 ECGs from 250 145 patients at Seoul National University Bundang Hospital, we developed a deep-learning model utilizing printed 12-lead ECG images to estimate patients' age (ECG-Age) and 1- and 5-year mortality risks. The model was validated externally using the CODE-15% dataset from Brazil. The ECG-Age showed a high correlation with chronological age in both the internal and external validation datasets (Pearson's R = 0.888 and 0.852, respectively). In the internal validation, the direct mortality risk prediction models showed area under the curves (AUCs) of 0.843 and 0.867 for 5- and 1-year all-cause mortality, respectively. For 5- and 1-year cardiovascular mortality, the AUCs were 0.920 and 0.916, respectively. In the CODE-15%, the mortality risk predictions showed AUCs of 0.818 and 0.836 for the prediction of 5- and 1-year all-cause mortality, respectively. Compared to the neutral Delta-Age (ECG-Age - chronological age) group, hazard ratios for deaths were 1.88 [95% confidence interval (CI): 1.14-3.92], 2.12 (95% CI: 1.15-3.92), 4.46 (95% CI: 2.22-8.96) and 7.68 (95% CI: 3.32-17.76) for positive Delta-Age groups (5-10, 10-15, 15-20, >20), respectively. An image-based AI-ECG model is a feasible tool for estimating biological age and assessing all-cause and cardiovascular mortality risks, providing a practical approach for utilizing standardized ECG images in predicting long-term health outcomes.

Enhancing Preoperative Outcome Prediction: A Comparative Retrospective Case-Control Study on Machine Learning versus the International Esodata Study Group Risk Model for Predicting 90-Day Mortality in Oncologic Esophagectomy.

Risk prediction prior to oncologic esophagectomy is crucial for assisting surgeons and patients in their joint informed decision making. Recently, a new risk prediction model for 90-day mortality after esophagectomy using the International Esodata Study Group (IESG) database was proposed, allowing for the preoperative assignment of patients into different risk categories. However, given the non-linear dependencies between patient- and tumor-related risk factors contributing to cumulative surgical risk, machine learning (ML) may evolve as a novel and more integrated approach for mortality prediction. We evaluated the IESG risk model and compared its performance to ML models. Multiple classifiers were trained and validated on 552 patients from two independent centers undergoing oncologic esophagectomies. The discrimination performance of each model was assessed utilizing the area under the receiver operating characteristics curve (AUROC), the area under the precision-recall curve (AUPRC), and the Matthews correlation coefficient (MCC). The 90-day mortality rate was 5.8%. We found that IESG categorization allowed for adequate group-based risk prediction. However, ML models provided better discrimination performance, reaching superior AUROCs (0.64 [0.63-0.65] vs. 0.44 [0.32-0.56]), AUPRCs (0.25 [0.24-0.27] vs. 0.11 [0.05-0.21]), and MCCs (0.27 ([0.25-0.28] vs. 0.15 [0.03-0.27]). Conclusively, ML shows promising potential to identify patients at risk prior to surgery, surpassing conventional statistics. Still, larger datasets are needed to achieve higher discrimination performances for large-scale clinical implementation in the future.

Development and validation of a mortality risk prediction model for chronic obstructive pulmonary disease: a cross-sectional study using probabilistic graphical modelling

Chronic Obstructive Pulmonary Disease (COPD) is a leading cause of mortality. Predicting mortality risk in patients with COPD can be important for disease management strategies. Although all-cause mortality predictors have been developed previously, limited research exists on factors directly affecting COPD-specific mortality. In a retrospective study, we used probabilistic graphs to analyse clinical cross-sectional data (COPDGene cohort), including demographics, spirometry, quantitative chest imaging, and symptom features, as well as gene expression data. COPDGene recruited current and former smokers, aged 45-80 years with >10 pack-years smoking history, from across the USA (Phase 1, 11/2007-4/2011) and invited them for a follow-up visit (Phase 2, 7/2013-7/2017). ECLIPSE cohort recruited current and former smokers (COPD patients and controls from USA and Europe), aged 45-80 with smoking history >10 pack-years (12/2005-11/2007). We applied graphical models on multi-modal data COPDGene Phase 1 participants to identify factors directly affecting all-cause and COPD-specific mortality (primary outcomes); and on Phase 2 follow-up cohort to identify additional molecular and social factors affecting mortality. We used penalized Cox regression with features selected by the causal graph to build VAPORED, a mortality risk prediction model. VAPORED was compared to existing scores (BODE: BMI, airflow obstruction, dyspnoea, exercise capacity; ADO: age, dyspnoea, airflow obstruction) on the ability to rank individuals by mortality risk, using four evaluation metrics (concordance, concordance probability estimate (CPE), cumulative/dynamic (C/D) area under the receiver operating characteristic curve (AUC), and integrated C/D AUC). The results were validated in ECLIPSE. Graphical models, applied on the COPDGene Phase 1 samples (n=8610), identified 11 and 7 variables directly linked to all-cause and COPD-specific mortality, respectively. Although many appear in both models, non-lung comorbidities appear only in the all-cause model, while forced vital capacity (FVC %predicted) appears in COPD-specific mortality model only. Additionally, the graph model of Phase 2 data (n=3182) identified internet access, CD4 T cells and platelets to be linked to lower mortality risk. Furthermore, using the 7 variables linked to COPD-specific mortality (forced expiratory volume in 1s/forced vital capacity (FEV1/FVC) ration, FVC %predicted, age, history of pneumonia, oxygen saturation, 6-min walk distance, dyspnoea) we developed VAPORED mortality risk score, which we validated on the ECLIPSE cohort (3-yr all-cause mortality data, n=2312). VAPORED performed significantly better than ADO, BODE, and updated BODE indices in predicting all-cause mortality in ECLIPSE in terms of concordance (VAPORED [0.719] vs ADO [0.693; FDR p-value 0.014], BODE [0.695; FDR p-value 0.020], and updated BODE [0.694; FDR p-value 0.021]); CPE (VAPORED [0.714] vs ADO [0.673; FDR p-value <0.0001], BODE [0.662; FDR p-value <0.0001], and updated BODE [0.646; FDR p-value <0.0001]); 3-year C/D AUC (VAPORED [0.728] vs ADO [0.702; FDR p-value 0.017], BODE [0.704; FDR p-value 0.021], and updated BODE [0.703; FDR p-value 0.024]); integrated C/D AUC (VAPORED [0.723] vs ADO [0.698; FDR p-value 0.047], BODE [0.695; FDR p-value 0.024], and updated BODE [0.690; FDR p-value 0.021]). Finally, we developed a web tool to help clinicians calculate VAPORED mortality risk and compare it to ADO and BODE predictions. Our work is an important step towards improving our identification of high-risk patients and generating hypotheses of potential biological mechanisms and social factors driving mortality in patients with COPD at the population level. The main limitation of our study is the fact that the analysed datasets consist of older people with extensive smoking history and limited racial diversity. Thus, the results are relevant to high-risk individuals or those diagnosed with COPD and the VAPORED score is validated for them. This research was supported by NIH [NHLBI, NLM]. The COPDGene study is supported by the COPD Foundation, through grants from AstraZeneca, Bayer Pharmaceuticals, Boehringer Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer and Sunovion.

Application of machine learning to predict in-hospital mortality after transcatheter mitral valve repair

IntroductionTranscatheter mitral valve repair offers a minimally invasive treatment option for patients at high risk for traditional open repair. We sought to develop dynamic machine-learning risk prediction models for in-hospital mortality after transcatheter mitral valve repair using a national cohort. MethodsAll adult hospitalization records involving transcatheter mitral valve repair were identified in the 2016–2020 Nationwide Readmissions Database. As a result of initial class imbalance, undersampling of the majority class and subsequent oversampling of the minority class using Synthetic Minority Oversampling TEchnique were employed in each cross-validation training fold. Machine-learning models were trained to predict patient mortality after transcatheter mitral valve repair and compared with traditional logistic regression. Shapley additive explanations plots were also developed to understand the relative impact of each feature used for training. ResultsAmong 2,450 patients included for analysis, the in-hospital mortality rate was 1.8%. Naïve Bayes and random forest models were the best at predicting transcatheter mitral valve repair postoperative mortality, with an area under the receiver operating characteristic curve of 0.83 ± 0.05 and 0.82 ± 0.04, respectively. Both models demonstrated superior ability to predict mortality relative to logistic regression (P < .001 for both). Medicare insurance coverage, comorbid liver disease, congestive heart failure, renal failure, and previous coronary artery bypass grafting were associated with greater predicted likelihood of in-hospital mortality, whereas elective surgery and private insurance coverage were linked with lower odds of mortality. ConclusionMachine-learning models significantly outperformed traditional regression methods in predicting in-hospital mortality after transcatheter mitral valve repair. Furthermore, we identified key patient factors and comorbidities linked with greater postoperative mortality. Future work and clinical validation are warranted to continue improving risk assessment in transcatheter mitral valve repair .

Multiple biomarkers are equivalent to clinical pulmonary arterial hypertension survival risk models

Risk assessment in pulmonary arterial hypertension (PAH) is fundamental to guiding treatment and improved outcomes. Clinical models are excellent at identifying high-risk patients but leave uncertainty amongst moderate risk patients. Can a multiple blood biomarker model of PAH, using previously described biomarkers, improve risk discrimination over current models? Using multiplex ELISA, we measured NT-proBNP, ST2, IL-6, Endostatin, Galectin-3, HDGF, and IGF binding proteins (IGFBP1-7) in train (n=1623), test (n=696) and validation (n=237) cohorts. Clinical variables, biomarkers were evaluated by principal component analysis. NT-proBNP was not included to develop an NT-proBNP independent model. Unsupervised k-means clustering classified subjects into clusters. Transplant-free survival by cluster was examined using Kaplan-Meier and Cox proportional hazard regressions. Hazard by cluster was compared to NT-proBNP, REVEAL, and ESC/ERS Risk models alone, and combined clinical and biomarker models. The algorithm generated 5 clusters with good risk discrimination using 6 biomarkers, weight, height, and age at PAH diagnosis. In the test and validation cohorts the biomarker model alone performed equivalent to REVEAL (AUC 0.74). Adding the biomarker model to the ESC/ERS, and REVEAL scores improved the ESC/ERS and REVEAL scores. The best overall model was the biomarker model adjusted for NT-proBNP with the best C-statistic, AIC, and calibration for the adjusted model compared to either the biomarker or NT-proBNP model alone. A multi-biomarker model alone was equivalent to current PAH clinical mortality risk prediction models and improved performance when combined, and added to NT-proBNP. Clinical risk scores offer excellent predictive models but require multiple tests; adding blood biomarkers to models can improve prediction or enable more frequent, non-invasive monitoring of risk in PAH to support therapeutic decision making.

Quality Improvement Study Using a Machine Learning Mortality Risk Prediction Model Notification System on Advance Care Planning in High-Risk Patients

Background: Advance care planning (ACP) is an important aspect of patient care that is underutilized. Machine learning (ML) models can help identify patients appropriate for ACP. The objective was to evaluate the impact of using provider notifications based on an ML model on the rate of ACP documentation and patient outcomes. Methods: This was a pre-post QI intervention study at a tertiary academic hospital. Adult patients admitted to general medicine teams identified to be at elevated risk of mortality using an ML model were included in the study. The intervention consisted of notifying a provider by email and page for a patient identified by the ML model. Results: A total of 479 encounters were analyzed of which 282 encounters occurred post-intervention. The covariate-adjusted proportion of higher-risk patients with documented ACP rose from 6.0% at baseline to 56.5% (Risk Ratio (RR)= 9.42, 95% CI: 4.90 - 18.11). Patients with ACP were more than twice as likely to have code status reduced when ACP was documented (29.0% vs. 10.8% RR=2.69, 95% CI: 1.64 – 4.27). Additionally, patients with ACP had twice the odds of hospice referral (22.2% vs. 12.6% Odds Ratio=2.16, 95% CI: 1.16 – 4.01). However, patients with ACP documented had a longer mean LOS (9.7 vs. 7.6 days, Event time ratio = 1.29, 95% CI: 1.10 - 1.53). Conclusion: Provider notifications using an ML model can lead to an increase in completion of ACP documentation by frontline clinicians in the inpatient setting.

Hacia una mejor predicción de la mortalidad en pacientes oncológicos en UCI: análisis comparativo de escalas pronósticas: revisión sistemática de la literatura

ObjectiveTo evaluate the predictive ability of mortality prediction scales in cancer patients admitted to intensive care units. DesignA systematic review of the literature was conducted using a search algorithm in October 2022. The following databases were searched: PubMed, Scopus, Virtual Health Library (BVS), and Medrxiv. The risk of bias was assessed using the QUADAS-2 scale. SettingIntensive care units admitting cancer patients. ParticipantsStudies that included adult patients with an active cancer diagnosis who were admitted to the intensive care unit. InterventionsIntegrative study without interventions. Main variables of interestMortality prediction, standardized mortality, discrimination, and calibration. ResultsSeven mortality risk prediction models were analyzed in cancer patients in the ICU. Most models (APACHE II, APACHE IV, SOFA, SAPS-II, SAPS-III, and MPM II) underestimated mortality, while the ICMM overestimated it. The APACHE II had the SMR (Standardized Mortality Ratio) value closest to 1, suggesting a better prognostic ability compared to the other models. ConclusionsPredicting mortality in intensive care unit cancer patients remains an intricate challenge due to the lack of a definitive superior model and the inherent limitations of available prediction tools. For evidence-based informed clinical decision-making, it is crucial to consider the healthcare team's familiarity with each tool and its inherent limitations. Developing novel instruments or conducting large-scale validation studies is essential to enhance prediction accuracy and optimize patient care in this population.

Effect of COVID-19 on Thrombosis Incidence and Patient Prognosis in Kidney Transplant Recipients.

BACKGROUND Thrombosis poses a grave threat to patients undergoing kidney transplants, with a heightened risk of mortality. While previous studies have established a link between COVID-19 and thrombosis, the specific association between COVID-19 and thrombosis in this patient population remains unexplored. MATERIAL AND METHODS We conducted a retrospective analysis utilizing data from 394 individuals who underwent kidney transplantation within the period of September 1, 2015, to April 1, 2023. To evaluate overall survival, we employed Kaplan-Meier analysis and utilized a logistic regression model for risk analysis. Furthermore, we developed a prediction model and assessed its accuracy through calibration curves. RESULTS Out of the 394 patients included in our study, a total of 51 individuals experienced thrombosis, resulting in 2 deaths. Our analysis revealed that COVID-19 infection significantly increased the risk of thrombosis (odds ratio [OR] 8.60, 95% confidence interval 3.13-24.74, P<0.01). Additionally, the use of cyclosporine was found to elevate the risk of death (OR 20.86, 95% CI 7.93-59.24, P<0.01) according to multifactorial analysis. Logistic models were employed to screen variables, and predictive models were constructed based on the presence of COVID-19 infection and the usage of cyclosporine. A nomogram was developed, demonstrating promising accuracy in estimating the risk of thrombosis during internal validation, with a corrected C-index of 0.869. CONCLUSIONS Our study suggests that both COVID-19 infection and the use of cyclosporine can serve as reliable predictors of thrombosis risk in patients undergoing renal transplantation. Furthermore, we developed a mortality risk prediction model based on COVID-19 in assessing thrombosis.

A single-center, retrospective study of hospitalized patients with lower respiratory tract infections: clinical assessment of metagenomic next-generation sequencing and identification of risk factors in patients

IntroductionLower respiratory tract infections(LRTIs) in adults are complicated by diverse pathogens that challenge traditional detection methods, which are often slow and insensitive. Metagenomic next-generation sequencing (mNGS) offers a comprehensive, high-throughput, and unbiased approach to pathogen identification. This retrospective study evaluates the diagnostic efficacy of mNGS compared to conventional microbiological testing (CMT) in LRTIs, aiming to enhance detection accuracy and enable early clinical prediction.MethodsIn our retrospective single-center analysis, 451 patients with suspected LRTIs underwent mNGS testing from July 2020 to July 2023. We assessed the pathogen spectrum and compared the diagnostic efficacy of mNGS to CMT, with clinical comprehensive diagnosis serving as the reference standard. The study analyzed mNGS performance in lung tissue biopsies and bronchoalveolar lavage fluid (BALF) from cases suspected of lung infection. Patients were stratified into two groups based on clinical outcomes (improvement or mortality), and we compared clinical data and conventional laboratory indices between groups. A predictive model and nomogram for the prognosis of LRTIs were constructed using univariate followed by multivariate logistic regression, with model predictive accuracy evaluated by the area under the ROC curve (AUC).Results(1) Comparative Analysis of mNGS versus CMT: In a comprehensive analysis of 510 specimens, where 59 cases were concurrently collected from lung tissue biopsies and BALF, the study highlights the diagnostic superiority of mNGS over CMT. Specifically, mNGS demonstrated significantly higher sensitivity and specificity in BALF samples (82.86% vs. 44.42% and 52.00% vs. 21.05%, respectively, p < 0.001) alongside greater positive and negative predictive values (96.71% vs. 79.55% and 15.12% vs. 5.19%, respectively, p < 0.01). Additionally, when comparing simultaneous testing of lung tissue biopsies and BALF, mNGS showed enhanced sensitivity in BALF (84.21% vs. 57.41%), whereas lung tissues offered higher specificity (80.00% vs. 50.00%). (2) Analysis of Infectious Species in Patients from This Study: The study also notes a concerning incidence of lung abscesses and identifies Epstein-Barr virus (EBV), Fusobacterium nucleatum, Mycoplasma pneumoniae, Chlamydia psittaci, and Haemophilus influenzae as the most common pathogens, with Klebsiella pneumoniae emerging as the predominant bacterial culprit. Among herpes viruses, EBV and herpes virus 7 (HHV-7) were most frequently detected, with HHV-7 more prevalent in immunocompromised individuals. (3) Risk Factors for Adverse Prognosis and a Mortality Risk Prediction Model in Patients with LRTIs: We identified key risk factors for poor prognosis in lower respiratory tract infection patients, with significant findings including delayed time to mNGS testing, low lymphocyte percentage, presence of chronic lung disease, multiple comorbidities, false-negative CMT results, and positive herpesvirus affecting patient outcomes. We also developed a nomogram model with good consistency and high accuracy (AUC of 0.825) for predicting mortality risk in these patients, offering a valuable clinical tool for assessing prognosis.ConclusionThe study underscores mNGS as a superior tool for lower respiratory tract infection diagnosis, exhibiting higher sensitivity and specificity than traditional methods.

Performance Drift in Machine Learning Models for Cardiac Surgery Risk Prediction: Retrospective Analysis.

The Society of Thoracic Surgeons and European System for Cardiac Operative Risk Evaluation (EuroSCORE) II risk scores are the most commonly used risk prediction models for in-hospital mortality after adult cardiac surgery. However, they are prone to miscalibration over time and poor generalization across data sets; thus, their use remains controversial. Despite increased interest, a gap in understanding the effect of data set drift on the performance of machine learning (ML) over time remains a barrier to its wider use in clinical practice. Data set drift occurs when an ML system underperforms because of a mismatch between the data it was developed from and the data on which it is deployed. In this study, we analyzed the extent of performance drift using models built on a large UK cardiac surgery database. The objectives were to (1) rank and assess the extent of performance drift in cardiac surgery risk ML models over time and (2) investigate any potential influence of data set drift and variable importance drift on performance drift. We conducted a retrospective analysis of prospectively, routinely gathered data on adult patients undergoing cardiac surgery in the United Kingdom between 2012 and 2019. We temporally split the data 70:30 into a training and validation set and a holdout set. Five novel ML mortality prediction models were developed and assessed, along with EuroSCORE II, for relationships between and within variable importance drift, performance drift, and actual data set drift. Performance was assessed using a consensus metric. A total of 227,087 adults underwent cardiac surgery during the study period, with a mortality rate of 2.76% (n=6258). There was strong evidence of a decrease in overall performance across all models (P<.0001). Extreme gradient boosting (clinical effectiveness metric [CEM] 0.728, 95% CI 0.728-0.729) and random forest (CEM 0.727, 95% CI 0.727-0.728) were the overall best-performing models, both temporally and nontemporally. EuroSCORE II performed the worst across all comparisons. Sharp changes in variable importance and data set drift from October to December 2017, from June to July 2018, and from December 2018 to February 2019 mirrored the effects of performance decrease across models. All models show a decrease in at least 3 of the 5 individual metrics. CEM and variable importance drift detection demonstrate the limitation of logistic regression methods used for cardiac surgery risk prediction and the effects of data set drift. Future work will be required to determine the interplay between ML models and whether ensemble models could improve on their respective performance advantages.

Interpretability-based machine learning for predicting the risk of death from pulmonary inflammation in Chinese intensive care unit patients.

The objective of this research was to create a machine learning predictive model that could be easily interpreted in order to precisely determine the risk of premature death in patients receiving intensive care after pulmonary inflammation. In this study, information from the China intensive care units (ICU) Open Source database was used to examine data from 2790 patients who had infections between January 2019 and December 2020. A 7:3 ratio was used to randomly assign the whole patient population to training and validation groups. This study used six machine learning techniques: logistic regression, random forest, gradient boosting tree, extreme gradient boosting tree (XGBoost), multilayer perceptron, and K-nearest neighbor. A cross-validation grid search method was used to search the parameters in each model. Eight metrics were used to assess the models' performance: accuracy, precision, recall, F1 score, area under the curve (AUC) value, Brier score, Jordon's index, and calibration slope. The machine methods were ranked based on how well they performed in each of these metrics. The best-performing models were selected for interpretation using both the Shapley Additive exPlanations (SHAP) and Local interpretable model-agnostic explanations (LIME) interpretable techniques. A subset of the study cohort's patients (120/1668, or 7.19%) died in the hospital following screening for inclusion and exclusion criteria. Using a cross-validated grid search to evaluate the six machine learning techniques, XGBoost showed good discriminative ability, achieving an accuracy score of 0.889 (0.874-0.904), precision score of 0.871 (0.849-0.893), recall score of 0.913 (0.890-0.936), F1 score of 0.891 (0.876-0.906), and AUC of 0.956 (0.939-0.973). Additionally, XGBoost exhibited excellent performance with a Brier score of 0.050, Jordon index of 0.947, and calibration slope of 1.074. It was also possible to create an interactive internet page using the XGBoost model. By identifying patients at higher risk of early mortality, machine learning-based mortality risk prediction models have the potential to significantly improve patient care by directing clinical decision making and enabling early detection of survival and mortality issues in patients with pulmonary inflammation disease.

Critical appraisal and assessment of bias among studies evaluating risk prediction models for in-hospital and 30-day mortality after percutaneous coronary intervention: a systematic review

ObjectiveWe systematically assessed prediction models for the risk of in-hospital and 30-day mortality in post-percutaneous coronary intervention (PCI) patients.DesignSystematic review and narrative synthesis.Data sourcesSearched PubMed, Web of Science, Embase, Cochrane...

Risk factors and prediction model for mortality in HIV/Talaromyces marneffei co-infection: A retrospective cohort study

BackgroundThis study aimed to identify the risk factors associated with mortality among patients co-infected with human immunodeficiency virus (HIV) and Talaromyces marneffei (TM) in China, and develop a risk prediction model. MethodsIn this retrospective cohort analysis conducted from 2013 to 2024, comprehensive clinical data from 160 patients were analyzed using a logistic regression model to identify mortality predictors and construct a predictive model. An additional 36 patients constituted the validation cohort, which was specifically designed to evaluate the predictive value of the model. Model performance was assessed using the area under the curve (AUC). ResultsThe overall mortality rate for hospitalized patients with HIV/TM co-infection was 17.35 %. The median age was 35.0 years, and 89.30 % were male. Additionally, 89.80 % of the patients reported fever and 87.76 % presented with lymphadenopathy. Key independent risk factors associated with mortality included age (odds ratio (OR): 1.103, 95 % confidence interval (CI) = 1.033–1.178, P = 0.003), procalcitonin (PCT) levels (OR: 1.270, 95 % CI = 1.052–1.534, P = 0.013), and urea to albumin ratio (UAR) (OR: 1.491, 95 % CI = 1.175–1.892, P < 0.001). Advanced age, elevated PCT levels, and increased UAR were identified as independent risk factors of mortality. Furthermore, the mortality prediction probability combining age, PCT, and UAR exhibited a high predictive value in patients with HIV/TM co-infection. Additionally, the AUC showed a good discrimination ability in the validation group (AUC, 0.898). ConclusionsAdvanced age, elevated PCT levels, and increased UAR significantly determine mortality in patients with HIV/TM co-infection. These findings underscore the potential of using laboratory parameters as predictive indicators of mortality, facilitating the early identification of HIV/TM co-infection cases in clinical practice.

Construction of a mortality risk prediction model for patients with acute diquat poisoning based on clinically accessible data

BackgroundTo examine the risk factors associated with mortality in individuals suffering from acute diquat poisoning and to develop an effective prediction model using clinical data.MethodsA retrospective review was conducted on the clinical records of 107 individuals who were hospitalized for acute diquat poisoning at a tertiary hospital in Sichuan Province between January 2017 and September 30, 2023, and further categorized into survivor and nonsurvivor groups based on their mortality status within 30 days of poisoning. The patient’s demographic information, symptoms within 24 h of admission, and details of the initial clinical ancillary examination, as well as the APACHE II score, were documented. The model was developed using backward stepwise logistic regression, and its performance was assessed using receiver operating characteristic curves, calibration curves, Brier scores, decision curve analysis curves, and bootstrap replicates for internal validation.ResultsMultifactorial logistic regression analysis revealed that blood pressure (hypertension, OR 19.73, 95% CI 5.71–68.16; hypotension, OR 61.38, 95% CI 7.40–509.51), white blood count (OR 1.35, 95% CI 1.20–1.52), red cell distribution width-standard deviation (OR 1.22, 95% CI 1.08–1.38), and glomerular filtration rate (OR 0.96, 95% CI 0.94–0.97) were identified as independent risk factors for mortality in patients with diquat. Subsequently, a nomogram with an area under the curve of 0.97 (95% CI: 0.93–1) was developed. Internal bootstrap resampling (1000 repetitions) confirmed the model’s adequate discriminatory power, with an area under the curve of 0.97. Decision curve analysis demonstrated greater net gains for the nomogram, while the clinical impact curves indicated greater predictive validity.ConclusionThe nomogram model developed in this study using available clinical data enhances the prediction of risk for DQ patients and has the potential to provide valuable clinical insights to guide patient treatment decisions.