Accurate Predictor Of Mortality Research Articles

Cardiorespiratory Mortality Prediction Based on Air Pollution Using Tree-Based Ensemble Models

Air pollution has a substantial negative impact on human wellbeing and health. Cardiorespiratory mortality is one of the primary effects of air pollution. In this study, we provide analysis of air pollution, cardiorespiratory mortality and the cardiorespiratory mortality is predicted based on air pollution using tree-based ensemble models. The tree-based ensemble models utilized in this study are Voting Regressor (VR), Random Forest (RF), Gradient Tree Boosting (GB), and Extreme Gradient Boosting (XGBoost). The used dataset contains data for five research locations: Shah Alam (SA), Klang (KLN), Putrajaya (PUJ), Cheras, Kuala Lumpur (CKL), and Petaling Jaya (PJ) from January 2006 to December 2016. The results show that XGBoost and VR models outperformed the rest of the models with the best evaluation metric scores in the Klang study area, XGBoost(MAE:0.005, RMSE:0.010, MAPE:0.70%) and VR (MAE:0.005, RMSE:0.011, MAPE:0.70%). The results reveal that the utilized models provided an excellent and accurate prediction of cardiorespiratory mortality based on air pollution and can follow the trend of cardiorespiratory mortality.

Machine learning for prediction of all-cause mortality after transcatheter aortic valve implantation.

Prediction of adverse events in mid-term follow-up after transcatheter aortic valve implantation (TAVI) is challenging. We sought to develop and validate a machine learning model for prediction of 1-year all-cause mortality in patients who underwent TAVI and were discharged following the index procedure. The model was developed on data of patients who underwent TAVI at a high-volume centre between January 2013 and March 2019. Machine learning by extreme gradient boosting was trained and tested with repeated 10-fold hold-out testing using 34 pre- and 25 peri-procedural clinical variables. External validation was performed on unseen data from two other independent high-volume TAVI centres. Six hundred four patients (43% men, 81±5 years old, EuroSCORE II 4.8 [3.0-6.3]%) in the derivation and 823 patients (46% men, 82±5 years old, EuroSCORE II 4.7 [2.9-6.0]%) in the validation cohort underwent TAVI and were discharged home following the index procedure. Over the 12 months of follow-up, 68 (11%) and 95 (12%) subjects died in the derivation and validation cohorts, respectively. In external validation, the machine learning model had an area under the receiver-operator curve of 0.82 (0.78-0.87) for prediction of 1-year all-cause mortality following hospital discharge after TAVI, which was superior to pre- and peri-procedural clinical variables including age 0.52 (0.46-0.59) and the EuroSCORE II 0.57 (0.51-0.64), P<0.001 for a difference. Machine learning based on readily available clinical data allows accurate prediction of 1-year all-cause mortality following a successful TAVI.

Risk assessment validation in patients with pulmonary arterial hypertension: Data from a Southern Brazil registry (RESPHIRAR study).

Pulmonary arterial hypertension (PAH) is a severe and progressive disease characterized by increased pulmonary vascular resistance, ultimately leading to right heart failure and death. Registries are a valuable tool in the research of rare conditions such as PAH. Moreover, the risk assessment strategy has been validated in European and North American registries and has been reported to provide an accurate prediction of mortality and the clinical advantage of reaching low-risk status. However, there is no available data from Brazil. Thus, the aim of the present study was to describe the characteristics of a sample of PAH from Southern Brazil and to retrospectively validate the risk assessment at our population. The RESPHIRAR is a retrospective and multicentric registry on pulmonary hypertension. With a join collaboration from nine centers in Southern Brazil, demographics, clinical presentation, and hemodynamics data of PAH were collected between 2007 and 2017. Moreover, the REVEAL 2.0 and REVEAL 2.0 Lite risk assessments were validated in our population. Overall, 370 PAH patients were included in the present study. Patients were predominantly female (78.5%) and had a mean age of 41.8 ± 18.8 years. Most patients (33.4%) had idiopathic PAH, 30.2% had PAH associated with congenital heart disease, and 23.5% had PAH associated with connective tissue disease. The low-risk group showed significantly lower mortality than the intermediated- or high-risk group at diagnosis (p < 0.05). In conclusion, our data suggest that REVEAL 2.0 and REVEAL 2.0 Lite risk assessments can predict mortality risk in PAH patients in Southern Brazil.

348 PROGNOSTIC EVALUATION OF HEART FAILURE IN WOMEN: INSIGHT FROM THE MECKI SCORE DATABASE

Abstract Background Heart failure (HF) in women often assumes extremely peculiar characteristics. The risk stratification and prognostic evaluation of HF in women is a challenge for clinicians. The actual prognostic scores are, in fact, lacking a specific sex-oriented assessment. In HF women show better survival despite a comparatively low peak oxygen consumption (VO2) at cardiopulmonary exercise testing: this raises doubt about the accuracy of risk assessment in women. Previous studies on MECKI score database have demonstrated that female prognostic advantage is lost when sex-specific differences are correctly considered. The present analysis has the aim to identify parameters which could be differently associated to prognosis in men and women and to re-calibrate the MECKI score according to these differences. Methods The new weights of the MECKI score variables according to gender, were calculated using the means of the 200 repetitions obtained to cross-validation procedure. The primary outcome of the study was the composite of all-cause mortality, urgent heart transplant and LVAD implant. The difference in predictive ability between the native and gender re-calibrated MECKI was calculated with the ROC curve at 2 years follow-up and calibration plot. Results We retrospectively analyzed 7900 HF patients with reduced ejection fraction included in the MECKI score registry (61±13 years, 6456 M/ 1444 F, LVEF 33±10%, VO2 /kg 14.9±4.9 mL/min Kg). Patients follow up was: 4.05 years [1.72-7.47]. Table 1 presents the weights of the individual MECKI variables for males and females. The main differences have been detected for the intercept value, hemoglobin and Na+. We used these variables to recalibrate the algorithm of MECKI score. We obtained higher AUC for MECKI re-calibrated by gender than for Native MECKI (0.7893 vs. 0.7799, p=0.0194, Fig. 1 left panel). Moreover, the gender MECKI was better calibrated (Fig. 1, right panels). Conclusions The calibration of MECKI Score according to gender specific differences improves the prognostic power and accuracy of mortality prediction at two years follow up.

Multitask Deep Learning for Cost-Effective Prediction of Patient's Length of Stay and Readmission State Using Multimodal Physical Activity Sensory Data.

In a hospital, accurate and rapid mortality prediction of Length of Stay (LOS) is essential since it is one of the essential measures in treating patients with severe diseases. When predictions of patient mortality and readmission are combined, these models gain a new level of significance. Therefore, the most expensive components of patient care are LOS and readmission rates. Several studies have assessed readmission to the hospital as a single-task issue. The performance, robustness, and stability of the model increase when many correlated tasks are optimized. This study develops multimodal multitasking Long Short-Term Memory (LSTM) Deep Learning (DL) model that can predict both LOS and readmission for patients using multi-sensory data from 47 patients. Continuous sensory data is divided into eight sections, each of which is recorded for an hour. The time steps are constructed using a dual 10-second window-based technique, resulting in six steps per hour. The 30 statistical features are computed by transforming the sensory input into the resulting vector. The proposed multitasking model predicts 30-day readmission as a binary classification problem and LOS as a regression task by constructing discrete time-step data based on the length of physical activity during a hospital stay. The proposed model is compared to a random forest for a single-task problem (classification or regression) because typical machine learning algorithms are unable to handle the multitasking challenge. In addition, sensory data combined with other cost-effective modalities such as demographics, laboratory tests, and comorbidities to construct reliable models for personalized, cost-effective, and medically acceptable prediction. With a high accuracy of 94.84%, the proposed multitask multimodal DL model classifies the patient's readmission status and determines the patient's LOS in hospital with a minimal Mean Square Error (MSE) of 0.025 and Root Mean Square Error (RMSE) of 0.077, which is promising, effective, and trustworthy.

The patient generated-subjective global assessment (PG-SGA) and ECOG performance status are associated with mortality in patients hospitalized with breast cancer

This study evaluated the association between risk of malnutrition and performance status, and mortality in hospitalized breast cancer patients. Prospective cohort study with hospitalized breast cancer patients evaluated at a referral Cancer Center. The Risk of malnutrition was assessed by the Patient-Generated Subjective Global Assessment (PG-SGA) and performance status was determined using the Eastern Cooperative Oncology Group Performance Status Scale (ECOG PS). Logistic regression was used to analyze the factors associated with death, using the odds ratio (OR) with a 95% confidence interval (CI) as an effect measure. A total of 195 woman were included, with a mean age of 56.3 (±12.6) years. Patients with an overall PG-SGA score ≥18 (OR: 2.11; 95% CI: 1.03-4.62) and ECOG PS≥3 (OR: 3.34; 95% CI: 1.48-7.52) had a higher occurrence of death during hospitalization, regardless of age or disease stage. The concomitant presence of these two factors improved the accuracy of the association (OR: 5.32; 95% CI: 3.11-9.76) and showed good predictive accuracy (C-statistics: 0.77). Nutritional risk and poor performance status were associated with a higher occurrence of death in women with breast cancer. The use of these two indicators improves their predictive accuracy for mortality.

Using the Charlson comorbidity index as a prognostic factor of lower gastrointestinal system bleeding: the experience of a tertiary center

Introduction: Lesions in the gastroinestinal (GI) tract that are distal to the Treitz ligament are what cause the lower gastrointestinal bleeding (LGB) system. The purpose of this study was to investigate and compare the Charlson Comorbidity Index (CCI), mortality rates, length of hospital stays, need for intensive care, need for blood products, and surgical rates in patients with acute LGB. Material and Method: Retrospective research was done on patients who had lower GI bleeding and had been seen in our gastroenterology clinic between 2015 and 2021. We looked into the impact of CCI on patients' follow-up after LGB. Results: The mean age of the 210 patients who had lower GI bleeding was 67.70±13.67 years. For all of the patients, the median CCI value was 4.00. (2.00-5.00). While 16 study participants (group 1) passed away, 194 participants (group 2) were released from the hospital. The variance in the median CCI values between the two groups was statistically significant (p&amp;gt;0.001). The results of a multivariate logistic regression analysis revealed that CCI was a reliable predictor of mortality (p&amp;gt;0.001). Conclusion: It was found that CCI was an accurate predictor of mortality. CCI ought to be regarded as a crucial factor in the treatment of patients who are bleeding from their lower gastrointestinal tract.

DeepMPM: a mortality risk prediction model using longitudinal EHR data

BackgroundAccurate precision approaches have far not been developed for modeling mortality risk in intensive care unit (ICU) patients. Conventional mortality risk prediction methods can hardly extract the information in longitudinal electronic medical records (EHRs) effectively, since they simply aggregate the heterogeneous variables in EHRs, ignoring the complex relationship and interactions between variables and the time dependence in longitudinal records. Recently deep learning approaches have been widely used in modeling longitudinal EHR data. However, most existing deep learning-based risk prediction approaches only use the information of a single disease, neglecting the interactions between multiple diseases and different conditions.ResultsIn this paper, we address this unmet need by leveraging disease and treatment information in EHRs to develop a mortality risk prediction model based on deep learning (DeepMPM). DeepMPM utilizes a two-level attention mechanism, i.e. visit-level and variable-level attention, to derive the representation of patient risk status from patient’s multiple longitudinal medical records. Benefiting from using EHR of patients with multiple diseases and different conditions, DeepMPM can achieve state-of-the-art performances in mortality risk prediction.ConclusionsExperiment results on MIMIC III database demonstrates that with the disease and treatment information DeepMPM can achieve a good performance in terms of Area Under ROC Curve (0.85). Moreover, DeepMPM can successfully model the complex interactions between diseases to achieve better representation learning of disease and treatment than other deep learning approaches, so as to improve the accuracy of mortality prediction. A case study also shows that DeepMPM offers the potential to provide users with insights into feature correlation in data as well as model behavior for each prediction.

Mortality prediction upon hospital admission - the value of clinical assessment: A retrospective, matched cohort study.

Accurate prediction of mortality upon hospital admission is of great value, both for the sake of patients and appropriate resources’ allocation. A myriad of assessment tools exists for this purpose. The evidence relating to the comparative value of clinical assessment versus established indexes are scarce. We analyzed the accuracy of a senior physician’s clinical assessment in a retrospective cohort of patients in a crude, general patients’ population and later on a propensity matched patients’ population. In one department of internal medicine in a tertiary hospital, of 9891 admitted patients, 973 (10%) were categorized as prone to death in a 6-months’ duration by a senior physician. The risk of death was significantly higher for these patients [73.1% vs 14.1% mortality within 180 days; hazard ratio (HR) = 7.58; confidence intervals (CI) 7.02‐8.19, P < .001]. After accounting for multiple, other patients’ variables associated with increased risk of mortality, the correlation remained significant (HR = 3.25; CI 2.85‐3.71, P < .001). We further performed a propensity matching analysis (a subgroup of 710 patients, subdivided to two groups with 355 patients each): survival rates were as low as 45% for patients categorized as prone to death compared to 78% in patients who weren’t categorized as such (P < .001). Reliance on clinical evaluation, done by an experienced senior physician, is an appropriate tool for mortality prediction upon hospital admission, achieving high accuracy rates.

Time evolution of cytokine profiles associated with mortality in COVID-19 hospitalized patients.

BackgroundHigh cytokine levels have been associated with severe COVID-19 disease. Although many cytokine studies have been performed, not many of them include combinatorial analysis of cytokine profiles through time. In this study we investigate the association of certain cytokine profiles and its evolution, and mortality in SARS-CoV2 infection in hospitalized patients.MethodsSerum concentration of 45 cytokines was determined in 28 controls at day of admission and in 108 patients with COVID-19 disease at first, third and sixth day of admission. A principal component analysis (PCA) was performed to characterize cytokine profiles through time associated with mortality and survival in hospitalized patients.ResultsAt day of admission non-survivors present significantly higher levels of IL-1α and VEGFA (PC3) but not through follow up. However, the combination of HGF, MCP-1, IL-18, eotaxine, and SCF (PC2) are significantly higher in non-survivors at all three time-points presenting an increased trend in this group through time. On the other hand, BDNF, IL-12 and IL-15 (PC1) are significantly reduced in non-survivors at all time points with a decreasing trend through time, though a protective factor. The combined mortality prediction accuracy of PC3 at day 1 and PC1 and PC2 at day 6 is 89.00% (p<0.001).ConclusionsHypercytokinemia is a hallmark of COVID-19 but relevant differences between survivors and non-survivors can be early observed. Combinatorial analysis of serum cytokines and chemokines can contribute to mortality risk assessment and optimize therapeutic strategies. Three clusters of cytokines have been identified as independent markers or risk factors of COVID mortality.

AIMS65 and PALBI Scores as Predictors of Six Months’ Mortality in Cirrhotic Patients with Acute Variceal Bleeding

Background &amp; Aims:Bleeding gastroesophageal varices are a cause of high mortality among cirrhotic patients. We aimed to investigate late mortality predictors and prognostic models using easily verified factors at admission in cirrhotic patients with acute variceal bleeding (AVB).Methods:Between January 2020 and June 2020, 142 patients with AVB from Tanta university hospital were included. Investigating multiple prognostic models was done using multiple logistic regression after identifying significant predictors of 6 months' mortality. Mortality prediction accuracy was assessed with area under the receiver operating characteristic (AUROC) curve.Results:The 6 months’ overall mortality rate was 31% (44 patients had died). AIMS56, Child-Turcotte-Pugh (CTP) grade C and MELD scores were significantly higher among non survivors (p&lt;0.001) while Platelet-albumin-bilirubin (PALBI) was significantly more negative among survivors (P=0.001). Hepatocellular carcinoma was not significantly related to the mortality (p =0.364). Univariate analysis showed that high CTP, MELD, AIMS65 and PALBI scores were predictors of mortality and associated with decreased survival with high sensitivity and low specificity; while multivariate analysis showed that only AIMS56 was independently associated with mortality (p 0.004).Conclusion:CTP, MELD, AIMS65 and PALBI scores are simple, bed side risk scores that can be used for the prediction of 6 months’ mortality after AVB in cirrhotic patients with high sensitivities and lower specificities.

An Artificial Intelligence Approach to Guiding the Management of Heart Failure Patients Using Predictive Models: A Systematic Review.

Heart failure (HF) is one of the leading causes of mortality and hospitalization worldwide. The accurate prediction of mortality and readmission risk provides crucial information for guiding decision making. Unfortunately, traditional predictive models reached modest accuracy in HF populations. We therefore aimed to present predictive models based on machine learning (ML) techniques in HF patients that were externally validated. We searched four databases and the reference lists of the included papers to identify studies in which HF patient data were used to create a predictive model. Literature screening was conducted in Academic Search Ultimate, ERIC, Health Source Nursing/Academic Edition and MEDLINE. The protocol of the current systematic review was registered in the PROSPERO database with the registration number CRD42022344855. We considered all types of outcomes: mortality, rehospitalization, response to treatment and medication adherence. The area under the receiver operating characteristic curve (AUC) was used as the comparator parameter. The literature search yielded 1649 studies, of which 9 were included in the final analysis. The AUCs for the machine learning models ranged from 0.6494 to 0.913 in independent datasets, whereas the AUCs for statistical predictive scores ranged from 0.622 to 0.806. Our study showed an increasing number of ML predictive models concerning HF populations, although external validation remains infrequent. However, our findings revealed that ML approaches can outperform conventional risk scores and may play important role in HF management.

Mortality after transjugular intrahepatic portosystemic shunt in older adult patients with cirrhosis: A validated prediction model.

Implantation of a transjugular intrahepatic portosystemic shunt (TIPS) improves survival in patients with cirrhosis with refractory ascites and portal hypertensive bleeding. However, the indication for TIPS in older adult patients (greater than or equal to 70 years) is debated, and a specific prediction model developed in this particular setting is lacking. The aim of this study was to develop and validate a multivariable model for an accurate prediction of mortality in older adults. We prospectively enrolled 411 consecutive patients observed at four referral centers with de novo TIPS implantation for refractory ascites or secondary prophylaxis of variceal bleeding (derivation cohort) and an external cohort of 415 patients with similar indications for TIPS (validation cohort). Older adult patients in the two cohorts were 99 and 76, respectively. A cause-specific Cox competing risks model was used to predict liver-related mortality, with orthotopic liver transplant and death for extrahepatic causes as competing events. Age, alcoholic etiology, creatinine levels, and international normalized ratio in the overall cohort, and creatinine and sodium levels in older adults were independent risk factors for liver-related death by multivariable analysis. After TIPS implantation, mortality is increased by aging, but TIPS placement should not be precluded in patients older than 70 years. In older adults, creatinine and sodium levels are useful predictors for decision making. Further efforts to update the prediction model with larger sample size are warranted.

Predictive Accuracy of Neutrophil-to-Lymphocyte Ratio on Severity and Outcomes in COVID-19 Patients

Background/Aims: Neutrophil-to-lymphocyte ratio (NLR) is a proven marker in coronavirus disease 2019 (COVID-19) severity and mortality. However, the utility of a sequential NLR 2 (on day 5) in comparison to baseline NLR in predicting clinical outcomes and severity remains largely unexplored. Methods: This was a hospital-based retrospective observational study. Results: Higher mortality (19.9% vs. 48%) and a more severe disease (14.8% vs. 21%) were observed with elevated NLR 1 and NLR 2, respectively. NLR 2 at a cutoff of 9.88 was a better predictor of mortality, when compared to NLR 1 at 5.67, and NLR 2 has a strong correlation with mortality rates in COVID-19. Conclusion: Our study demonstrated that NLR 1 and NLR 2 were more reliable predictors of mortality than disease severity; in comparison, NLR 2 is a more accurate predictor of mortality than NLR 1. The study unravels the potential role of a sequential NLR 2, to have a better correlation in predicting the clinical severity and outcomes. The potential role of NLR 2 in assessing the interim progression of the disease and thereby initiating specific interventions at critical junctures to influence the outcome is unveiled and merits exploration in detail by larger studies.

The lung allocation score and other available models lack predictive accuracy for post-lung transplant survival.

Improved predictive models are needed in lung transplantation in the setting of a proposed allocation system that incorporates longer-term post-transplant survival in the United States. Allocation systems require accurate mortality predictions to justly allocate organs. Utilizing the United Network for Organ Sharing database (2005-2017), we fit models to predict 1-year mortality based on the Lung Allocation Score (LAS), the Chan, etal, 2019 model, a novel "clinician" model (a priori clinician selection of pre-transplant covariates), and two machine learning models (Least Absolute Shrinkage and Selection Operator; LASSO and Random Forests) for predicting 1-year and 3-year post-transplant mortality. We compared predictive accuracy among models. We evaluated the calibration of models by comparing average predicted probability vs observed outcome per decile. We repeated analyses fit for 3-year mortality, disease category, including donor covariates, and LAS era. The area under the cure for all models was low, ranging from 0.55 to 0.62. All exhibited reasonable negative predictive values (0.87-0.90), but the positive predictive value for was poor (all <0.25). Evaluating LAS calibration found 1-year post-transplant estimates consistently overestimated risk of mortality, with greater differences in higher deciles. LASSO, Random Forests, and clinician models showed no improvement when evaluated by disease category or with the addition of donor covariates and performed worse for 3-year outcomes. The LAS overestimated patients' risk of post-transplant death, thus underestimating transplant benefit in the sickest candidates. Novel models based on pre-transplant recipient covariates failed to improve prediction. There should be wariness in post-transplant survival predictions from available models.

Multi-center validation of machine learning model for preoperative prediction of postoperative mortality

Accurate prediction of postoperative mortality is important for not only successful postoperative patient care but also for information-based shared decision-making with patients and efficient allocation of medical resources. This study aimed to create a machine-learning prediction model for 30-day mortality after a non-cardiac surgery that adapts to the manageable amount of clinical information as input features and is validated against multi-centered rather than single-centered data. Data were collected from 454,404 patients over 18 years of age who underwent non-cardiac surgeries from four independent institutions. We performed a retrospective analysis of the retrieved data. Only 12–18 clinical variables were used for model training. Logistic regression, random forest classifier, extreme gradient boosting (XGBoost), and deep neural network methods were applied to compare the prediction performances. To reduce overfitting and create a robust model, bootstrapping and grid search with tenfold cross-validation were performed. The XGBoost method in Seoul National University Hospital (SNUH) data delivers the best performance in terms of the area under receiver operating characteristic curve (AUROC) (0.9376) and the area under the precision-recall curve (0.1593). The predictive performance was the best when the SNUH model was validated with Ewha Womans University Medical Center data (AUROC, 0.941). Preoperative albumin, prothrombin time, and age were the most important features in the model for each hospital. It is possible to create a robust artificial intelligence prediction model applicable to multiple institutions through a light predictive model using only minimal preoperative information that can be automatically extracted from each hospital.

A Comparative Study between Peptic Ulcer Perforation Score, Mannheim Peritonitis Index, ASA Score, and Jabalpur Score in Predicting the Mortality in Perforated Peptic Ulcers

Introduction Peptic ulcer disease continues to be a major public health in most developing countries despite the advances in medical management. The incidence of perforations remains high and has the highest mortality rate of any complication of ulcer disease. Risk stratification of cases will lead to better preoperative management and efficient utilization of intensive care unit resources. The purpose of the present study is to compare different existing scoring systems and identify the most accurate predictor of mortality in perforated peptic ulcer (PPU) cases. Materials and Methods This is an observational study conducted in Karnataka Institute of Medical Sciences, Hubli, India. All cases of PPU disease admitted from December 2017 to August 2019 who were treated surgically were included in the study. Demographic data were collected and peptic ulcer perforation (PULP) score, Mannheim peritonitis index (MPI), American Society of Anesthesiologists (ASA) score, and Jabalpur score (JS) were calculated for individual patient and compared. The patient was followed up during the postoperative period. Observation A total of 45 patients were included in the study with a mean age of 42.5 years. Most of the patients presented with 24 hours of the onset of symptoms. Nonsteroidal anti-inflammatory drug use was noted in 8.9% patients, and steroid use was present in 2.2% patients. Of the 45 patients, 7 deaths were reported. Between the various scoring systems, the MPI and JS were better predictors of mortality with a p -value of <0.001 and 0.007, respectively. In contrast, the PULP and ASA scores had p -value not statistically significant. However, the PULP score was a better predictor of postoperative complication with a p -value of 0.047. Conclusion Of the four scoring systems validated, the MPI and JS were better predictors of mortality in the given population. PULP score is a better predictor of postoperative complications in the present study.

Evaluating High-Dimensional Machine Learning Models to Predict Hospital Mortality Among Older Patients With Cancer.

Older hospitalized cancer patients face high risks of hospital mortality. Improved risk stratification could help identify high-risk patients who may benefit from future interventions, although we lack validated tools to predict in-hospital mortality for patients with cancer. We evaluated the ability of a high-dimensional machine learning prediction model to predict inpatient mortality and compared the performance of this model to existing prediction indices. We identified patients with cancer older than 75 years from the National Emergency Department Sample between 2016 and 2018. We constructed a high-dimensional predictive model called Cancer Frailty Assessment Tool (cFAST), which used an extreme gradient boosting algorithm to predict in-hospital mortality. cFAST model inputs included patient demographic, hospital variables, and diagnosis codes. Model performance was assessed with an area under the curve (AUC) from receiver operating characteristic curves, with an AUC of 1.0 indicating perfect prediction. We compared model performance to existing indices including the Modified 5-Item Frailty Index, Charlson comorbidity index, and Hospital Frailty Risk Score. We identified 2,723,330 weighted emergency department visits among older patients with cancer, of whom 144,653 (5.3%) died in the hospital. Our cFAST model included 240 features and demonstrated an AUC of 0.92. Comparator models including the Modified 5-Item Frailty Index, Charlson comorbidity index, and Hospital Frailty Risk Score achieved AUCs of 0.58, 0.62, and 0.71, respectively. Predictive features of the cFAST model included acute conditions (respiratory failure and shock), chronic conditions (lipidemia and hypertension), patient demographics (age and sex), and cancer and treatment characteristics (metastasis and palliative care). High-dimensional machine learning models enabled accurate prediction of in-hospital mortality among older patients with cancer, outperforming existing prediction indices. These models show promise in identifying patients at risk of severe adverse outcomes, although additional validation and research studying clinical implementation of these tools is needed.

Highly sensitive C-reactive protein, pentraxin-3 and mannose-binding lectin in patients with hemodialysis and their association with cardiovascular complications and mortality

Introduction: Cardiovascular disorder(CVD) is one of the most important comorbidities in patients with hemodialysis(HD). Several markers for kidney disease and kidney transplantation have been investigated. Objectives: This study aimed to evaluate the association between three inflammatory biomarkers including C-reactive protein (CRP), pentraxin 3 (PTX-3) and mannose-binding lectin (MBL) with CVD events and mortality in HD patients. Patients and Methods: This study was carried on patients with immunosuppressive therapy aged between 18-80 years with chronic renal failure undergoing HD. Patients were visited at baseline and every six month for 12 months. The levels of biomarkers and complications were recorded before, one week later and 2 months after transplantation. Echocardiography and calculation of ejection fraction(EF) were performed to determine the occurrence of CVD events. Results: Ninety patients with the mean age of 54.0±3.24 years were included. All biomarkers had significant area under the curve for accurate diagnosis between the patients with low EF (with a decreasing measure) and normal EF patients (p&lt;0.05); however, none of them had a significant accurate diagnosis between patients with low EF (with a constant measure) and normal EF patients (P&gt;0.05). PTX-3 and MBL showed a significant correlation to detect CVD events (r=0.987; p&lt;0.0001). Also, PTX-3 and MBL had significant correlations with hs-CRP (P&lt;0.05). None of these markers had significant accuracy for prediction of mortality. Conclusion: Serum level of PTX-3 and MBL not only increase reactive protein, but also may be considered as diagnostic biomarkers for early detection of CVD events in patients with end stage renal disorders.

Mortality Prediction in Cardiac Intensive Care Unit Patients: A Systematic Review of Existing and Artificial Intelligence Augmented Approaches.

The medical complexity and high acuity of patients in the cardiac intensive care unit make for a unique patient population with high morbidity and mortality. While there are many tools for predictions of mortality in other settings, there is a lack of robust mortality prediction tools for cardiac intensive care unit patients. The ongoing advances in artificial intelligence and machine learning also pose a potential asset to the advancement of mortality prediction. Artificial intelligence algorithms have been developed for application of electrocardiogram interpretation with promising accuracy and clinical application. Additionally, artificial intelligence algorithms applied to electrocardiogram interpretation have been developed to predict various variables such as structural heart disease, left ventricular systolic dysfunction, and atrial fibrillation. These variables can be used and applied to new mortality prediction models that are dynamic with the changes in the patient's clinical course and may lead to more accurate and reliable mortality prediction. The application of artificial intelligence to mortality prediction will fill the gaps left by current mortality prediction tools.