Predicted Risk Of Mortality Research Articles

Central vs. Brachial Blood Pressure and Pulse Pressure Amplification for Mortality Risk Prediction in Patients Undergoing Coronary Angiography.

Arterial hypertension is a significant risk factor for cardiovascular (CV) morbidity and mortality. Although central blood pressure (BP) evaluation is considered the gold standard, the reliability of non-invasive measurements remains unclear. Therefore, we compared the predictive value of invasively measured central BP with non-invasively measured brachial BP and analyzed pulse pressure (PP) amplification (delta-PP; difference between central and peripheral PP) as an independent predictor of mortality. We analyzed systolic (SBP), diastolic (DBP), mean arterial BP (MAP), PP and delta-PP as predictors of CV and all-cause mortality in the Ludwigshafen Risk and Cardiovascular Health (LURIC) study, involving 3316 patients referred for coronary angiography. All brachial BP parameters, except DBP, were significantly linked to all-cause and CV mortality in a univariate analysis. A 10 mmHg increase in SBP, MAP, and PP corresponded to increased risks of all-cause (11%, 10%, and 19%) and CV mortality (11%, 11%, and 18%). Central SBP and PP showed similar, but numerically weaker, associations with increased risks of all-cause (5% and 10%) and CV mortality (4% and 8%).After adjusting for age, sex, BMI, diabetes mellitus, and eGFR, only delta-PP independently predicted mortality with a 10 mmHg increase linked to a 4% reduction in all-cause and 6% reduction in CV mortality. Neither brachial nor centrally measured BP parameters were independent mortality predictors in contrast to PP amplification, which remained an independent predictor of mortality in multivariate analysis, in a cohort with a medium to high CV risk profile. As PP amplification decreased, mortality increased.

The association between the metabolic score for insulin resistance and mortality in patients with cardiovascular disease: a national cohort study

BackgroundThe metabolic score for insulin resistance (METS-IR) is a novel index for evaluating insulin resistance and identifying high-risk cardiovascular disease (CVD) patients. This study aims to assess the prognostic value of METS-IR in predicting mortality risk in CVD patients.MethodsWe analyzed data from 2,515 CVD patients in the National Health and Nutrition Examination Survey (NHANES). Associations between METS-IR and all-cause mortality and cardiovascular mortality were evaluated using multivariable Cox proportional hazards models and restricted cubic splines (RCS). Threshold effects and sensitivity analyses were conducted to ensure robustness.ResultsOver a median follow-up of 91.4 months, 1,090 patients died, including 447 from cardiovascular causes. A U-shaped relationship was identified between lnMETS-IR and all-cause and cardiovascular mortality, with thresholds at 3.70 and 3.67. Below thresholds, an increase of lnMETS-IR was associated with a 75% reduction in the risk of all-cause mortality (HR: 0.25, 95% CI: 0.14–0.46) and a 79% reduction in the risk of cardiovascular mortality (HR: 0.21, 95% CI: 0.07–0.56). While above thresholds, an increase of lnMETS-IR was associated with a 180% increase in the risk of all-cause mortality (HR: 2.80, 95% CI: 1.61–4.88) and a 233% increase in the risk of cardiovascular mortality (HR: 3.33, 95% CI: 1.43–7.75).ConclusionsThis study identified a U-shaped association between lnMETS-IR and mortality among CVD patients, underscoring the potential of METS-IR as a valuable prognostic marker for mortality risk in patients with CVD.

Elevated concentrations of cardiac troponin T are associated with thoracic aortic calcification in non-dialysis chronic kidney disease patients of stage G3 to G5

Background Vascular calcification (VC), especially coronary artery calcification (CAC), serves as a robust predictor of cardiovascular mortality in chronic kidney disease (CKD) patients. Recent studies have revealed that the presence of extra-coronary calcifications (ECCs) contributes to cardiovascular disease (CVD). Elevated myocardial injury markers predict mortality risk in CKD patients and are associated with CVD. Nevertheless, the relationship between VC, including CAC and ECCs, and myocardial injury markers remain unexplored in non-dialysis CKD patients. Methods In 278 non-dialysis CKD patients of stage G3 to G5, we assessed calcified scores in CAC (Agatston score) and ECCs including thoracic aortic calcification (TAC), abdominal aortic calcification (AAC), carotid artery calcification, and valvular calcification. We analyzed the relationships between VC and myocardial injury markers of cardiac troponin T (cTnT) and creatine kinase-MB (CK-MB). Results A total of 278 non-dialysis CKD patients (median age 52.4 ± 13.2; male 65.1%; diabetes 33.5%) were enrolled. A total of 71.8% (227) of patients had cTnT levels above the upper limit of normal (> 0.014 ng/mL). Moderate to severe (calcified score ≥100 vs. <100), CAC (OR 6.39; 95% CI 1.03–39.61) and TAC (OR 6.16; 95% CI 1.76–21.55) were significantly associated with higher cTnT concentrations after adjustment for confounders. Additionally, male sex and a lower eGFR were also associated with cTnT elevation. However, when we included CAC and TAC in one model, only moderate to severe TAC (OR 4.85; 95% CI 1.38–16.96) was a risk factor for cTnT elevation, but not CAC. Furthermore, patients with severer TAC presented lower diastolic blood pressure (DBP), wider pulse pressure (p < 0.001) and higher prevalence of left ventricular hypertrophy (LVH). Conclusion Moderate to severe thoracic aortic calcification (TAC score ≥ 100) is significantly associated with elevated cTnT concentrations in non-dialysis CKD patients of stage G3 to G5. The linkage may result from decreased coronary perfusion and relative myocardial ischemia.

Comparing machine learning techniques for neonatal mortality prediction: insights from a modeling competition.

Predicting mortality risk in neonatal intensive care units (NICUs) is challenging due to complex, variable clinical and physiological data. Machine learning (ML) offers potential for more accurate risk stratification. To compare the performance of various ML models in predicting NICU mortality using a team-based modeling competition. We conducted a modeling competition with five neonatologist-led teams applying ML techniques-logistic regression, CatBoost, neural networks, random forest, and XGBoost-to a shared dataset from over 6,000 NICU admissions. The dataset included static demographic and clinical variables, alongside daily samples of heart rate and oxygen saturation. Each team developed models to predict mortality risk at baseline and within 7 days. Models were evaluated using the area under the receiver operator characteristic curve (AUC). Results were presented at a national meeting, where an audience poll ranked models before AUC results were revealed. The audience favored the most complex model (CNN) for real-world application, though logistic regression achieved the highest AUC on test data. Teams employed varied feature selection, tuning, and evaluation strategies. Logistic regression outperformed more complex models, highlighting the importance of selecting modeling methods based on data characteristics, interpretability, and expertise rather than model complexity alone. By demonstrating that model complexity does not necessarily equate to better predictive performance, this research encourages the careful selection of modeling approaches.

Association of cardiovascular-kidney-metabolic index with all-cause mortality during hospitalization in critically ill patients: a retrospective cohort study from MIMIC IV2.2

BackgroundThe cardiovascular-kidney-metabolic index (CKMI), a novel functional indicator proposed in this study, aims to accurately reflect the functional status of the heart, kidneys, and metabolism. However, its ability to predict mortality risk in critically ill patients during their stay in the intensive care unit (ICU) remains uncertain. Therefore, this study aims to validate the correlation between the CKMI during hospitalization and all-cause mortality.MethodsThe study utilized the Medical Information Mart for Intensive Care IV 2.2 (MIMIC-IV) dataset for a retrospective analysis of cohorts. The cohorts were divided into quartiles based on CKMI index levels. The primary endpoint was all-cause mortality during ICU and hospital stay, while secondary endpoints included the duration of ICU stay and overall hospitalization period. We established Cox proportional hazards models and employed multivariable Cox regression analysis and restricted cubic spline (RCS) regression analysis to explore the relationship between CKMI index and all-cause mortality during hospitalization in critically ill patients. Additionally, subgroup analyses were conducted based on different subgroups.ResultsThe study enrolled 1,576 patients (male 60.79%). In-patient and ICU mortality was 11.55% and 6.73%. Multivariate COX regression analysis demonstrated a significant negative correlation between CKMI index and the risk of hospital death [HR, 0.26 (95% CI 0.07–0.93), P = 0.038] and ICU mortality [HR, 0.13 (95% CI 0.03–0.67), P = 0.014].RCS regression model revealed that in-hospital mortality (P-value =0.015, P-Nonlinear =0.459) and ICU mortality (P-value =0.029, P-Nonlinear =0.432) increased linearly with increasing CKMI index. Subgroup analysis confirmed consistent effect size and direction across different subgroups, ensuring stable results.ConclusionOur research findings suggest that a higher CKMI index is associated with a significant reduction in both in-hospital and ICU mortality among critically ill patients. Therefore, CKMI index emerges as a highly valuable prognostic indicator for predicting the risk of in-hospital death in this population. However, to strengthen the validity of these results, further validation through larger-scale prospective studies is imperative.

Association of red blood cell distribution width-albumin ratio with in-hospital mortality in abdominal aortic aneurysm patients.

To explore whether red blood cell distribution width-albumin ratio (RAR) is relevant to in-hospital mortality among abdominal aortic aneurysm (AAA). This is a retrospective study retrieving data from the MIMIC-IV database. Patients were divided into survivor or non-survivor groups by the in-hospital mortality. Receiver operating characteristic curve analysis, logistic regression models, subgroup analysis, interaction analysis, and restricted cubic spline analysis were conducted to analyze the correlation between RAR and in-hospital mortality. Then, we divided patients into 2 groups by an optimal cutoff value of RAR to identify the factors independently linked to RAR. Following this, the mediation analysis was conducted to reveal the potential regulatory path. Finally, we assessed the clinical value of RAR in secondary outcomes containing length of hospital stay, intensive care unit (ICU) admission, and ICU stay. Totally 770 participants with AAA were enrolled: 722 survivors and 48 non-survivors. Higher RAR was observed in the non-survivor group and its level performed satisfactorily in predicting in-hospital mortality. AAA patients were more likely to die during in-hospital with the increase of RAR (P < .05) and this linear correlation was revealed by restricted cubic spline (P non-linear > .05). Additionally, urea nitrogen and creatinine were independently related to RAR. RAR served as a mediator in the association of urea nitrogen/creatinine with in-hospital mortality. Finally, the length of hospital stay and ICU stay were longer in the RAR ≥ 4.658 group (P < .05). RAR is a potential risk predictor for in-hospital mortality in AAA patients. Further, RAR upregulation was significantly correlated with prolonged length of hospital stay and ICU stay.

Impact of Chinese visceral adiposity index on all-cause mortality risk in community-dwelling older adults: a prospective cohort study

BackgroundsWhether excess visceral fat tissue increases the risk of death in older individuals remains controversial.AimsTo investigate the association between the Chinese Visceral Adiposity Index (CVAI) and all-cause mortality risk in older Chinese individuals.MethodsThis cohort study utilized data of individuals aged ≥ 65 years in 2014 to 2018 wave from the Chinese Longitudinal Healthy Longevity Survey database. Older individuals in the 2014 wave were included and followed up in 2018. CVAI was calculated based on age, body size, and blood lipid parameters, with higher values indicating increased visceral fat. Survival status was determined from official death certificates, local primary healthcare providers, or the family members of participants. Kaplan-Meier survival curve and log-rank test were employed to analyze cumulative mortality risk through CVAI tertiles (tertile 1: CVAI index < 97.34; tertile 2: 97.43 ≤ CVAI index < 132.21; and tertile 3: CVAI index ≥ 132.21). A Cox proportional hazards regression model was used to assess the relationship between the CVAI groups and all-cause mortality risk. Additionally, a sensitivity analysis was performed by excluding participants who died within the first year of follow-up. A subgroup analysis was performed based on age and sex, and a restricted cubic spline plot was created to analyze the dose-response relationship between CVAI and mortality risk.ResultsA total of 1414 individuals were included, and the mean age of the participants was 84.6 (standard deviation: 10.9) years, of which 46.4% were women and 32.8% were died during a median follow-up time of 36.4 months. In the multivariable adjusted Cox regression model, we observed a significantly lower risk of mortality in the CVAI tertile 2 and 3 groups than in the tertile 1 group. The hazard ratios (HR) of the tertile 2 and 3 groups were 0.68 (95% CI, approximately 0.52–0.89) and 0.63 (95% CI, approximately 0.48–0.82), respectively. Subgroup analysis revealed that the protective effect of higher CVAI levels on mortality was more pronounced in participants aged 65–79 years and in women.ConclusionOur study established a linear relationship between CVAI and mortality risk among community-dwelling older adults, with higher CVAI levels associated with a lower risk of all-cause mortality. These findings highlight the potential importance of visceral adiposity in predicting mortality risk in community-dwelling older adults.

Impact of Inflammation After Cardiac Surgery on 30-day Mortality and Machine Learning Risk Prediction

Impact of Inflammation After Cardiac Surgery on 30-day Mortality and Machine Learning Risk Prediction

Association between handgrip strength and mortality risk among cardiovascular disease patients: A cohort study from the NHANES 2011-2014.

Cardiovascular disease (CVD) and handgrip strength (HGS) are recognized mortality risk factors. However, the synergistic effect of CVD and HGS on mortality remains unclear. This study investigated the relationship between HGS and mortality in CVD patients. Data from the National Health and Nutrition Examination Survey (2011-2014) were analyzed. Weighted Cox proportional hazards models and restricted cubic splines (RCS) were used to examine associations, with subgroup, sensitivity and predictive performance analyses. Among 8,262 adults (50.56% female; 851 with CVD), 695 deaths (6.74%) occurred over a median follow-up of 80 months. Compared to non‒CVD individuals with high-HGS, CVD patients with low-HGS had the highest risk of all-cause mortality [hazard ratio (HR)=8.76; 95% CI: 4.20-18.30] and CVD mortality (HR=4.83; 95% CI: 3.48-6.70), while CVD patients with high-HGS showed no significant mortality risk increase. Among CVD patients, the HRs for all-cause and CVD mortality in the low-HGS group were 3.60 (95% CI, 2.21-5.86) and 4.01 (95% CI, 1.68-9.59). RCS analyses revealed that the relationships were linear (P‒nonlinear >0.05), and subgroup analyses revealed stroke status potentially modified the association for CVD mortality (P‒interaction=0.002). The addition of HGS significantly improved the predictive performance of current models for mortality (P<0.001). Low HGS may be associated with a higher risk of all-cause and CVD mortality among CVD patients. High HGS appears to reduce mortality risk among CVD patients. These findings suggest that HGS may serve as a valuable predictor of mortality risk in CVD patients.

Cardiogenic shock mortality according to Aetiology in a Mediterranean cohort: Results from the Shock-CAT study.

Mortality in cardiogenic shock (CS) remains elevated, with the potential for CS causes to impact prognosis and risk stratification. The aim was to investigate in-hospital prognosis and mortality in CS patients according to aetiology. We also assessed the prognostic accuracy of CardShock and IABP-SHOCK II scores. Shock-CAT study was a multicentre, prospective, observational study conducted from December 2018 to November 2019 in eight university hospitals in Catalonia, including non-selected consecutive CS patients. Data on clinical presentation, management, including mechanical circulatory support (MCS) were analysed comparing acute myocardial infarction (AMI) related CS and non-AMI-CS. The accuracy of CardShock and IABP-SHOCK II scores to assess 90day mortality risk were also compared. A total of 382 CS patients were included, age 65.3 (SD 13.9) years, 75.1% men. Patients were classified as AMI-CS (n=232, 60.7%) and non-AMI-CS (n=150, 39.3%). In the AMI-CS group, 77.6% were STEMI. Main aetiologies for non-AMI-CS were heart failure (36.2%), arrhythmias (22.1%) and valve disease (8.0%). AMI-CS patients required more MCS than non-AMI-CS (43.1% vs. 16.7%, P<0.001). In-hospital mortality was higher in AMI-CS (37.1 vs. 26.7%, P=0.035), with a two-fold increased risk after multivariate adjustment (odds ratio 2.24, P=0.019). The IABP-SHOCK II had superior discrimination for predicting 90day mortality when compared with CardShock in AMI-CS patients [area under the curve (AUC) 0.74 vs. 0.66, P=0.047] although both scores performed similarly in non-AMI-CS (AUC 0.64 vs. 0.62, P=0.693). In our cohort, AMI-CS mortality was increased by two-fold when compared with non-AMI-CS. IABP-SHOCK II score provides better 90day mortality risk prediction than CardShock score in AMI-CS, but both scores performed similar in non-AMI-CS patients.

Longitudinal changes in plasma Cystatin C and all-cause mortality risk among the middle-aged and elderly Chinese population

ObjectiveElevated plasma Cystatin C levels are associated with an increased mortality risk among middle-aged and elderly Chinese individuals. This study explores whether tracking the longitudinal changes in Cystatin C can improve the prediction of mortality risk and allow better risk stratification, jointly with baseline measurements. Design & MethodsThis analysis includes 3,195 participants from the China Health and Retirement Longitudinal Study who completed plasma Cystatin C measurements in two waves (2011 and 2015) and were followed through 2020. To evaluate the association between Cystatin C levels/changes and mortality risk, multivariate Cox proportional hazard models were employed, adjusting for potential confounders. Survival probabilities were compared using Kaplan-Meier curves and log-rank tests, while restricted cubic splines were utilized to illustrate any nonlinear relationships between Cystatin C levels and hazard ratios. ResultsParticipants in the highest quartile of baseline Cystatin C show an increased risk of mortality compared to those in the lowest quartile (hazard ratio (HR): 1.51, 95 % CI: 1.02–2.24, p = 0.04). Including longitudinal changes in Cystatin C further strengthens this association (HR: 1.81, 95 % CI: 1.20–2.74, p < 0.001). Kaplan-Meier plots show that baseline levels effectively stratify both the entire cohort and gender-specific subgroups (p < 0.001). Moreover, integrating baseline levels with the longitudinal changes in Cystatin C levels provides additional stratification benefits. The predictive performance significantly improves by including longitudinal changes in Cystatin C in baseline-only models, with the concordance index increasing from 0.745 to 0.839 and the area under the receiver operator characteristic curve rising from 0.751 to 0.845. Additionally, significant nonlinear relationships between changes in Cystatin C and HR are observed in the entire population, the males and the females (p = 0.003, 0.018, 0.025). ConclusionsDynamic monitoring of changes in Cystatin C could enhance the prediction of mortality risk among middle-aged and elderly individuals.

TCKAN: a novel integrated network model for predicting mortality risk in sepsis patients.

Sepsis poses a major global health threat, accounting for millions of deaths annually and significant economic costs. Accurately predicting the risk of mortality in sepsis patients enables early identification, promotes the efficient allocation of medical resources, and facilitates timely interventions, thereby improving patient outcomes. Current methods typically utilize only one type of data-either constant, temporal, or ICD codes. This study introduces a novel approach, the Time-Constant Kolmogorov-Arnold Network (TCKAN), which uniquely integrates temporal data, constant data, and ICD codes within a single predictive model. Unlike existing methods that typically rely on one type of data, TCKAN leverages a multi-modal data integration strategy, resulting in superior predictive accuracy and robustness in identifying high-risk sepsis patients. Validated against the MIMIC-III and MIMIC-IV datasets, TCKAN surpasses existing machine learning and deep learning methods in accuracy, sensitivity, and specificity. Notably, TCKAN achieved AUCs of 87.76% and 88.07%, demonstrating superior capability in identifying high-risk patients. Additionally, TCKAN effectively combats the prevalent issue of data imbalance in clinical settings, improving the detection of patients at elevated risk of mortality and facilitating timely interventions. These results confirm the model's effectiveness and its potential to transform patient management and treatment optimization in clinical practice. Although the TCKAN model has already incorporated temporal, constant, and ICD code data, future research could include more diverse medical data types, such as imaging and laboratory test results, to achieve a more comprehensive data integration and further improve predictive accuracy.

Association between lactate/albumin ratio and 28-day mortality in ICU critical patients with coronary heart disease: a retrospective analysis of the MIMIC-IV database.

The serum lactate/albumin ratio (LAR) is commonly employed for monitoring and evaluating the prognosis of critically ill patients. Both elevated lactate levels and decreased albumin levels may reflect the body's stress response and inflammatory reaction. Coronary heart disease (CHD), with common complications including myocardial infarction, arrhythmia, heart failure, is one of the leading causes of global death. Therefore, it is crucial to explore biomarkers that can predict the prognosis and mortality of CHD patients. This is a retrospective study in which the data is from the MIMIC-IV database. Our study assessed the association between LAR value and mortality within 28 days of admission in a total of 1,902 CHD patients from the Beth Israel Deaconess Medical Center. The results demonstrated a significant increase in 28-day mortality among individuals with higher LAR values. Multivariate analysis by Cox proportional hazard model revealed an incremental rise in mortality across each quartile with the increase of LAR value. Furthermore, restricted cubic spline (RCS) Cox regression analysis further revealed that higher LAR values were associated with increased 28-day mortality in the CHD patients. And subgroup analysis confirmed that the LAR level could serve as an independent predictor of 28-day mortality with CHD patients. Our study demonstrated that the LAR value can be an important risk predictor of 28-day mortality in patients with CHD, and a higher LAR associate with increased mortality rate.

Preoperative goal directed therapy in geriatric hip fracture patients – a retrospective quality improvement study

BackgroundHip fractures in older adults are common and carry a high risk of morbidity and mortality. Preoperative dehydration, a key risk factor for adverse outcomes, is often underdiagnosed at admission. It is important to identify high risk patients and optimize modifiable risk factors to improve the postoperative outcome. The p-POSSUM Score is a useful predictor of postoperative mortality risk. Implementing a defined fluid resuscitation protocol early in high-risk groups has proven effective in reducing both postoperative morbidity and mortality.Materials and methodsWe conducted a single-center, retrospective quality improvement study at a Level 1 trauma center in Switzerland, focusing on geriatric patients over 70 with a p-POSSUM Score > 5% and a proximal femur fracture, undergoing surgery between February 2015 and September 2019. We hypothesized that our institutional goal-directed fluid resuscitation protocol (GDT) would result in lower 30- and 90-day mortality rates and fewer complications for these high-risk patients. Outcome changes over time were analyzed annually.Results312 patients were included in our study. 147 followed our institutional GDT protocol, while 165 patients received standard care and were used as a control group. Initially, the odds ratio (OR) for 30-days mortality favored the GDT group; 2015 0.30 (CI: 0.07–1.18), 2016 0.28 (CI: 0.07–1.07), 2017 0.51 (CI: 0.13–2.04). This shifted toward the Non-GDT group in the year 2018 with a OR of 2.14 (CI: 0.59–7.84) and 2019 with 1.92 (CI 0.47–7.83). The pattern for 90-day mortality results was similar and slightly more pronounced. The estimates for the odds ratios remained consistent even after adjustment for the p-POSSUM score. The relative frequencies of complications showed no remarkable differences between the groups (GDT vs. Non-GDT).ConclusionIn our study, the treatment according to our GDT protocol was associated with survival advantage in the first three years. However, this trend reversed in 2018, with Non-GDT patients faring better. Although our retrospective study does not provide enough evidence of causality between the protocol and the mortality rates, it revealed that continuous critical evaluation of internal processes is essential in healthcare for quality management. This allows timely identification and adaptation of processes to issues, especially after initially positive results.

Predictive value of qSOFA and hypothermia combined with PT for prognosis in patients with severe trauma: a single-center retrospective cohort study

BackgroundTrauma represents a significant global health challenge.The development of an effective scoring tool capable of predicting mortality risk in trauma cases is essential. This study aimed to investigate the combined effects of quick sequential organ failure assessment (qSOFA) and hypothermia (H) along with prothrombin time (PT) in predicting the prognosis of patients with severe trauma.MethodsA retrospective cohort study was conducted to analyze data from severe trauma patients in the Trauma Database of the Trauma Center at the Second Affiliated Hospital of Soochow University between January 2017 and December 2021. Patients were categorized into survival and non-survival groups based on clinical outcomes. Baseline and clinical data were compared between the groups, and prognostic factors were explored using logistic regression analysis. Receiver operating characteristic (ROC) curves generated by 10-fold cross-validation using the caret in R programming language were used to assess the predictive efficacy of Injury Severity Score (ISS) and qSOFA + H + PT score for trauma patient mortality.ResultsA total of 509 severe trauma patients (377 males and 132 females) were included, with a median age of 53 years (range: 42–65 years). The mortality rate was found to be 23.4%. Logistic regression analysis revealed that age, ISS, and qSOFA + H + PT were significant predictors of death in severe trauma patients, with odds ratios of 1.035 (95%CI:1.014–1.057), 1.052 (95%CI:1.017–1.090), and 6.124 (95%CI:3.107–12.072), respectively (P < 0.05). The predictive efficacy of ISS and qSOFA + H + PT for mortality prediction was 0.742 and 0.816, respectively.The predictive efficacy of qSOFA + H + PT for emergency blood transfusion and operation was 0.743 and 0.702.ConclusionqSOFA + H + PT are identified as significant predictors to the death of severe trauma patients. They could be utilized as early intervention indicators in emergency departments, facilitating clinical management strategies such as emergency blood transfusion, emergency operation, and prognosis prediction.

Controlling nutritional status score and geriatric nutritional risk index as a predictor of mortality and hospitalization risk in hospitalized older adults

BackgroundThe COntrolling NUTritional Status (CONUT) score and the Global Nutrition Risk Index (GNRI) are screening tools for assessing the risk of malnutrition based on widely available biochemical parameters. The primary objective of this study was to investigate the predictive value of CONUT and GNRI score on 36 months mortality and hospitalization risk in hospitalized older patients. MethodsData of 382 patients (196 women, mean age 80.9±6.8 years) were retrieved from the multicenter Italian Study conducted by the Gruppo Lavoro Italiano Sarcopenia–Trattamento e Nutrizione (GLISTEN) in 12 Acute Care Wards. Sarcopenia was defined as presence of low handgrip strength plus low skeletal mass index (EWGSOP2 criteria). CONUT score was calculated based on serum albumin, total cholesterol and total lymphocyte count, whilst the GNRI was calculated using serum albumin and present body weight/ideal body weight ratio. ResultsDuring the 36-month follow-up, 120 out of 382 participants died (31.4%). From the results of the survival analysis, and after adjustment for potential confounders, participants with CONUT-derived moderate to high risk of malnutrition had shorter survival (HR = 2.67, 95%CI 1.34–5.33 and HR = 3.98, 95% CI: 1.77–8.97, respectively), as well as shorter survival free of urgent hospitalization (HR = 1.91; 95% CI: 1.03–3.55 and HR = 1.98; 95% CI: 1.14–3.42, respectively). Conversely, only GNRI indicative of high risk of malnutrition was an independent predictor of mortality 1.96 (95% CI: 1.06–3.62), but not of hospitalization. ConclusionThe CONUT score seems a valid tool to predict long-term mortality and hospitalization risk. Conversely, the GNRI is associated with long-term mortality, but not with hospital readmissions.

Machine learning-based prediction of one-year mortality in ischemic stroke patients.

Accurate prediction of mortality following an ischemic stroke is essential for tailoring personalized treatment strategies. This study evaluates the effectiveness of machine learning models in predicting one-year mortality after an ischemic stroke. Five machine learning models were trained using data from a national stroke registry, with logistic regression demonstrating the highest performance. The SHapley Additive exPlanations (SHAP) analysis explained the model's outcomes and defined the influential predictive factors. Analyzing 8183 ischemic stroke patients, logistic regression achieved 83% accuracy, 0.89 AUC, and an F1 score of 0.83. Significant predictors included stroke severity, pre-stroke functional status, age, hospital-acquired pneumonia, ischemic stroke subtype, tobacco use, and co-existing diabetes mellitus (DM). The model highlights the importance of predicting mortality in enhancing personalized stroke care. Apart from pneumonia, all predictors can serve the early prediction of mortality risk which supports the initiation of early preventive measures and in setting realistic expectations of disease outcomes for all stakeholders. The identified tobacco paradox warrants further investigation. This study offers a promising tool for early prediction of stroke mortality and for advancing personalized stroke care. It emphasizes the need for prospective studies to validate these findings in diverse clinical settings.

QTc interval prolongation impact on in-hospital mortality in acute coronary syndromes patients using artificial intelligence and machine learning

BackgroundPrediction of mortality in hospitalized patients is a crucial and important problem. Several severity scoring systems over the past few decades and machine learning models for mortality prediction have been developed to predict in-hospital mortality. Our aim in this study was to apply machine learning (ML) algorithms using QTc interval to predict in-hospital mortality in ACS patients and compare them to the validated conventional risk scores.ResultsThis study was retrospective, using supervised learning, and data mining. Out of a cohort of 500 patients admitted to a tertiary care hospital from September 2018 to August 2020, who presented with ACS. Prediction models for in-hospital mortality in ACS patients were developed using 3 ML algorithms. We employed the ensemble learning random forest (RF) model, the Naive Bayes (NB) model and the rule-based projective adaptive resonance theory (PART) model. These models were compared to one another and to two conventional validated risk scores; the Global Registry of Acute Coronary Events (GRACE) risk score and Thrombolysis in Myocardial Infarction (TIMI) risk score. Out of the 500 patients included in our study, 164 (32.8%) patients presented with unstable angina, 148 (29.6%) patients with non-ST-elevation myocardial infarction (NSTEMI) and 188 (37.6%) patients were having ST-elevation myocardial infarction (STEMI). 64 (12.8%) patients died in-hospital and the rest survived. Performance of prediction models was measured in an area under the receiver operating characteristic curve (AUC) ranged from 0.83 to 0.93 using all available variables compared to the GRACE score (0.9 SD 0.05) and the TIMI score (0.75 SD 0.02). Using QTc as a stand-alone variable yielded (0.67 SD 0.02) with a cutoff value 450 using Bazett’s formula, whereas using QTc in addition to other variables of personal and clinical data and other ECG variables, the result was 0.8 SD 0.04. Results of RF and NB models were almost the same, but PART model yielded the least results. There was no significant difference of AUC values after replacing the missing values and applying class balancer.ConclusionsThe proposed method can effectively predict patients at high risk of in-hospital mortality early in the setting of ACS using only clinical and ECG data. Prolonged QTc interval can be used as a risk predictor of in-hospital mortality in ACS patients.

Abstract 4138355: Machine Learning for Prediction of All-Cause Mortality in Acute Coronary Syndrome

Background: Acute coronary syndrome (ACS) remains one of the leading causes of death globally. Accurate and reliable mortality risk prediction of ACS patients is essential for developing targeted treatment strategies. Existing traditional models offer moderate discriminative value, and do not incorporate contemporary predictors of ACS prognosis. Machine learning (ML) models have emerged as an alternate method that may offer improved risk assessment. This study aims to compare machine learning models with traditional risk scores for predicting all-cause mortality in patients with ACS. Methods: PubMed, EMBASE, Web of Science and Cochrane databases were searched until 1 st November 2023 for studies comparing ML models with traditional statistical methods for event prediction of ACS patients. The primary outcome was comparative discrimination measured by C-statistics with 95% confidence intervals between ML models and traditional methods in estimating the risk of all-cause mortality. Results: Ten studies were included (239627 patients). The summary C-statistic of all ML models across all endpoints was 0.89 (95% CI, 0.86-0.92), compared to traditional methods 0.82 (95% CI, 0.79-0.85). The difference in C-statistic between all ML models and traditional methods was 0.07 (p&lt;0.05). Three models undertook external validation, and calibration was inconsistently reported. Conclusion: ML models demonstrated superior discrimination of all-cause mortality for ACS patients compared to traditional risk scores. Despite outperforming well-established prognostic tools such as the GRACE and TIMI scores, current clinical applications of ML approaches remain uncertain, particularly in view of the methodological limitations of existing studies and the need for greater model validation.

Lymph Node Yield and Long-Term Mortality Risk in Patients with Colon Cancer: A 20-Year Follow-Up National Study.

Lymph node status is a well-established prognostic factor for colon cancer, but the optimal number of nodes for accurate staging remains unclear. This study explored the relationship between lymph node yield (LNY) and 5-year mortality rates in colon cancer patients in New Zealand. Data from the New Zealand Cancer Registry were retrospectively analyzed for patients with TNM stage I, II, and III colon cancer between August 2003 and December 2021, with follow-up until January 2024. The primary outcome was the 5-year all-cause mortality rate, with LNY, age, sex, ethnicity, tumor site, district health board (DHB), and the number of positive nodes as covariates. Statistical analyses included univariate analysis, Cox regression modeling, and chi-squared tests. LNY was a significant predictor of 5-year mortality risk (hazard ratio 0.985, p < 0.0001), adjusted for age, sex, ethnicity, tumor site, and DHB. The strongest association between LNY and mortality rate was observed at 12 nodes. Further increases in LNY beyond 22 nodes did not lead to statistically significant differences in mortality rates. Lymph node ratio (LNR) was strongly associated with survival in stage III colon cancer, independent of LNY and the number of positive nodes. Higher LNY is significantly associated with reduced 5-year mortality rates in stage I-III colon cancer up to the 22-node mark. The strong correlation between LNR and mortality highlights its potential value for improving treatment planning in future clinical practice.