Current Risk Scores Research Articles

Decision Trees for Predicting Mortality in Transcatheter Aortic Valve Implantation.

Current prognostic risk scores in cardiac surgery do not benefit yet from machine learning (ML). This research aims to create a machine learning model to predict one-year mortality of a patient after transcatheter aortic valve implantation (TAVI). We adopt a modern gradient boosting on decision trees classifier (GBDTs), specifically designed for categorical features. In combination with a recent technique for model interpretations, we developed a feature analysis and selection stage, enabling the identification of the most important features for the prediction. We base our prediction model on the most relevant features, after interpreting and discussing the feature analysis results with clinical experts. We validated our model on 270 consecutive TAVI cases, reaching a C-statistic of 0.83 with CI [0.82, 0.84]. The model has achieved a positive predictive value ranging from 57% to 64%, suggesting that the patient selection made by the heart team of professionals can be further improved by taking into consideration the clinical data we identified as important and by exploiting ML approaches in the development of clinical risk scores. Our approach has shown promising predictive potential also with respect to widespread prognostic risk scores, such as logistic European system for cardiac operative risk evaluation (EuroSCORE II) and the society of thoracic surgeons (STS) risk score, which are broadly adopted by cardiologists worldwide.

The Growing Role of Echocardiography in Pulmonary Arterial Hypertension Risk Stratification: The Missing Piece.

Pulmonary arterial hypertension (PAH) is a rare, progressive disease with a poor prognosis. The pathophysiologic model is mainly characterized by an afterload mismatch in which an increased right ventricle afterload, driven by increased pulmonary vascular resistance (PVR), leads to right heart failure. International guidelines recommend optimization of treatment based on regular risk assessments to achieve or maintain a low-risk status. Current risk scores are based on a multi-modality approach, including demographic, clinical, functional, exercise, laboratory, and hemodynamic parameters, which lack significant echocardiographic parameters. The originality of echocardiography relies on the opportunity to assess in a non-invasive way a physiologically meaningful combination of easy to measure variables tightly related to right ventricle adaptation/maladaptation to increased afterload, the main determinant of a patient’s prognosis. Echo-derived morphological and functional parameters have been investigated in PAH, proving to have prognostic relevance. Different therapeutic strategies proved to have different effects in reducing PVR. An upfront combination of drugs, including a parenteral prostacyclin, has shown to be associated with right heart reverse remodeling in a greater proportion of patients than other treatment strategies as a function of PVR reduction. Adding echocardiographic data to current risk scores would allow better identification of right ventricle (RV) adaptation in PAH patients’ follow-up. This additional information would allow better stratification of the patient, leading to optimized and personalized therapeutic management.

A Systematic Review on Bleeding Risk Scores' Accuracy after Percutaneous Coronary Interventions in Acute and Elective Settings.

Dual antiplatelet therapy (DAT) is recommended for all patients undergoing percutaneous coronary intervention (PCI), as it significantly reduces the ischemic risk at the cost of increasing the incidence of bleeding events. Several clinical predictive models were developed to better stratify the bleeding risk associated with DAT. This systematic review aims to perform a literature survey of both standard and emerging bleeding risk scores and report their performance on predicting hemorrhagic events, especially in the era of second-generation drug-eluting stents and more potent P2Y12 inhibitors. We searched PubMed, ScienceDirect, and Cochrane databases for full-text studies that developed or validated bleeding risk scores in adult patients undergoing PCI with subsequent DAT. The risk of bias for each study was assessed using the prediction model risk of bias assessment tool (PROBAST). Eighteen studies were included in the present systematic review. Bleeding risk scores showed a modest to good discriminatory power with c-statistic ranging from 0.49 (95% CI, 0.45–0.53) to 0.82 (95% CI, 0.80–0.85). Clinical models that predict in-hospital bleeding events had a relatively good predictive performance, with c-statistic ranging from 0.70 (95% CI, 0.67–0.72) to 0.80 (95% CI, 0.73–0.87), depending on the risk scores and major hemorrhagic event definition used. The knowledge and utilization of the current bleeding risk scores in appropriate clinical contexts could improve the prediction of bleeding events.

Association of gut-related metabolites with outcome in acute heart failure

Trimethylamine N-oxide (TMAO), a gut-related metabolite, is associated with heart failure (HF) outcomes. However, TMAO is the final product of a complex metabolic pathway (ie, choline/carnitine) that has never been entirely investigated in HF. The present study investigates a panel of metabolites involved in the TMAO-choline/carnitine metabolic pathway for their associations with outcome in acute HF patients. In total, 806 plasma samples from acute HF patients were analyzed for TMAO, trimethyllysine, L-carnitine, acetyl-L-carnitine, γ-butyrobetaine, crotonobetaine, trimethylamine, betaine aldehyde, choline, and betaine using a developed liquid chromatography-tandem mass spectrometry method. Associations with outcome of all-cause mortality (death) and a composite of all-cause mortality and/or rehospitalization caused by HF (death/HF) at 30 days and 1 year were investigated. TMAO, trimethyllysine, L-carnitine, acetyl-L-carnitine, and γ-butyrobetaine were associated with death and death/HF at 30 days (short term; hazard ratio 1.30-1.49, P≤ .021) and at 1 year (long term; hazard ratio 1.15-1.25, P≤ .026) when adjusted for cardiac risk factors. L-carnitine and acetyl-L-carnitine were superior for short-term outcomes whereas TMAO was the superior metabolite for association with long-term outcomes. Furthermore, acetyl-L-carnitine and L-carnitine were superior for in-hospital mortality and improved risk stratification when combined with current clinical risk scores (ie, Acute Decompensated HEart Failure National REgistry, Organized Program To Initiate Lifesaving Treatment In Hospitalized Patients With Heart Failure, and Get With The Guidelines-Heart Failure; odds ratio (OR) ≥ 1.52, P≤ .020). Carnitine-related metabolites show associations with adverse outcomes in acute HF, in particular L-carnitine and acetyl-L-carnitine for short-term outcomes, and TMAO for long-term outcomes. Further studies are warranted to investigate the role and implications of carnitine metabolites including intervention in the pathogenesis of HF.

Novel mammogram‐based measures improve breast cancer risk prediction beyond an established mammographic density measure

Mammograms contain information that predicts breast cancer risk. We developed two novel mammogram-based breast cancer risk measures based on image brightness (Cirrocumulus) and texture (Cirrus). Their risk prediction when fitted together, and with an established measure of conventional mammographic density (Cumulus), is not known. We used three studies consisting of: 168 interval cases and 498 matched controls; 422 screen-detected cases and 1197 matched controls; and 354 younger-diagnosis cases and 944 controls frequency-matched for age at mammogram. We conducted conditional and unconditional logistic regression analyses of individually- and frequency-matched studies, respectively. We estimated measure-specific risk gradients as the change in odds per standard deviation of controls after adjusting for age and body mass index (OPERA) and calculated the area under the receiver operating characteristic curve (AUC). For interval, screen-detected and younger-diagnosis cancer risks, the best fitting models (OPERAs [95% confidence intervals]) involved: Cumulus (1.81 [1.41-2.31]) and Cirrus (1.72 [1.38-2.14]); Cirrus (1.49 [1.32-1.67]) and Cirrocumulus (1.16 [1.03 to 1.31]); and Cirrus (1.70 [1.48 to 1.94]) and Cirrocumulus (1.46 [1.27-1.68]), respectively. The AUCs were: 0.73 [0.68-0.77], 0.63 [0.60-0.66], and 0.72 [0.69-0.75], respectively. Combined, our new mammogram-based measures have twice the risk gradient for screen-detected and younger-diagnosis breast cancer (P ≤ 10-12 ), have at least the same discriminatory power as the current polygenic risk score, and are more correlated with causal factors than conventional mammographic density. Discovering more information about breast cancer risk from mammograms could help enable risk-based personalised breast screening.

Current Surgical Risk Scores Overestimate Risk in Minimally Invasive Aortic Valve Replacement.

Risk-scoring systems for surgical aortic valve replacement (AVR) were largely derived from sternotomy cases. We evaluated the accuracy of current risk scores in predicting outcomes after minimally invasive AVR (mini-AVR). Because transcatheter AVR (TAVR) is being considered for use in low-risk patients with aortic stenosis, accurate mini-AVR risk assessment is necessary. We reviewed 1,018 consecutive isolated mini-AVR cases (2009 to 2015). After excluding patients with Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) scores ≥4, we calculated each patient's European System for Cardiac Operative Risk Evaluation (EuroSCORE) II, TAVR Risk Score (TAVR-RS), and age, creatinine, and ejection fraction score (ACEF). We compared all 4 scores' accuracy in predicting mini-AVR 30-day mortality by computing each score's observed-to-expected mortality ratio (O:E). Area under the receiver operating characteristic (ROC) curves tested discrimination, and the Hosmer-Lemeshow goodness-of-fit tested calibration. Among 941 patients (mean age, 72 ± 12 years), 6 deaths occurred within 30 days (actual mortality rate, 0.6%). All 4 scoring systems overpredicted expected mortality after mini-AVR: ACEF (1.4%), EuroSCORE II (1.9%), STS-PROM (2.0%), and TAVR-RS (2.1%). STS-PROM best estimated risk for patients with STS-PROM scores 0 to <1 (0.6 O:E), ACEF for patients with STS-PROM scores 2 to <3 (0.6 O:E), and TAVR-RS for patients with STS-PROM scores 3 to <4 (0.7 O:E). ROC curves showed only fair discrimination and calibration across all risk scores. In low-risk patients who underwent mini-AVR, current surgical scoring systems overpredicted mortality 2-to-3-fold. Alternative dedicated scoring systems for mini-AVR are needed for more accurate outcomes assessment.

Development and validation of a tool for predicting type 2 diabetes in Mexican women of reproductive age

IntroductionDiabetes is a worldwide problem with a greater impact in developing countries, where many people are unaware of their risk. In Mexico, women show the greatest risk for T2D. Current risk scores have been developed and validated in predominantly older European cohorts. They are not the best option in Mexican women. The development of a risk model/score in this population would be useful. ObjectiveTo develop and validate a risk model and score that incorporates the most relevant risk factors for T2D in Mexican women of reproductive age. MethodsThe study was carried out in two phases, with the first phase being the development of the predictive model and the second phase the validation of the model in a separate independent population. A cohort of Mexican patients of reproductive age (“Derivation Cohort”) was used to create the predictive model. It included data on 3161 women. Risk factors for identification were assessed using Cox proportional hazards regression. Finally a score with a range of 0 to 19 points was developed to identify the 2.4 year probability of developing DM2 in Mexican women of reproductive age. Results147 new cases of T2D (4.6%) were identified in the Derivation Cohort model, 97 of 925 participants (10.48%) in the validation cohort. The risk factor predictors of T2D were: history of gestational diabetes (HR 2.69, 95% CI 1.10–6.58), BMI (HR 1.03, 95% CI 1.01–1.06), hypertriglyceridemia (HR 1.54, 95% CI 1.11–2.14) and fasting blood glucose (HR 1.06, 95% CI 1.05–1.08), with an AUC of 0.75. The AUC in the validation cohort was 0.91 (95% CI 0.87–0.94). The score had a sensitivity of 73% and specificity of 67% at a cutoff of ≥15. ConclusionsA predictive model and risk score was developed to detect cases at risk for incident T2D. It was generated using the characteristics of Mexican women of reproductive age. This risk score is a step forward in attempting to address the generational legacy that diabetes in pregnancy could have on women and their children.

ESC guidelines on stable chest pain: a comparison of the previous and current risk scores for the assessment of pre-test probability of coronary artery disease

Abstract Introduction The European Society of Cardiology (ESC) guidelines on stable chest pain recommend the use of a pre-test probability (PTP) risk score (RS) which predicts the likelihood of coronary artery disease (CAD) to guide investigation and modality. The 2019 guidelines provide an updated PTPRS compared with 2013 guidelines, adjusted for the lower prevalence of coronary artery disease in contemporary populations. We assessed the performance of the two PTPRS in a cohort of patients with stable chest pain who underwent CT coronary angiography (CTCA) as the first line investigation. Methods We retrospectively searched a radiology database from January 2017 to June 2018. CTCA reported CAD degree of stenosis as normal/minimal stenosis, mild (30–50%), moderate (50–70%), or severe (&amp;gt;70%) and retrospectively calculated ESC PTP scores for 2013 and 2019 guidelines. Results In total 652 patients underwent CTCA (mean age 55 yrs; IQR 48–63; 330 male). For ESC 2019 PTPRS there were no patients with PTP &amp;gt;85%. 2 patients had PTP 50–85%; 1 patient had moderate stenosis and 1 mild stenosis on CTCA. 267 patients had PTP 15–50%; 23 (9%) had severe stenosis and 35 (13%) moderate stenosis. Finally, 379 patients had PTP &amp;lt;15%; 11 (3%) had severe stenosis and 18 (5%) moderate CTCA stenosis. In comparison, ESC 2013 PTPRS had 2 patients with PTP &amp;gt;85%; 1 had moderate stenosis and 1 had mild stenosis on CTCA. 149 patients had PTP 50–85%; 17 (11%) had severe stenosis and 23 (15%) moderate stenosis. A further 427 patients had a PTP 15–50%; 17 (4%) had severe stenosis and 32 (8%) had moderate stenosis. Lastly, 70 patients had a PTP &amp;lt;15% and two (3%) were found to have a moderate stenosis on CTCA. Conclusions The updated ESC 2019 PTPRS appears to underestimate the presence of CAD given 11 (3%) patients with severe CTCA stenosis would have been missed. Although the 2013 PTPRS was thought to overestimate the prevalence of CAD, it did not miss anyone found to have severe CTCA stenosis. Furthermore, patients with evidence of mild or moderate CAD on CTCA may not have been investigated due to PTP &amp;lt;15% and therefore may not be commenced on medical therapy, to derive a mortality benefit as demonstrated in SCOT-Heart trial. Funding Acknowledgement Type of funding source: None

Prognostic Machine Learning Models for First-Year Mortality in Incident Hemodialysis Patients: Development and Validation Study.

BackgroundThe first-year survival rate among patients undergoing hemodialysis remains poor. Current mortality risk scores for patients undergoing hemodialysis employ regression techniques and have limited applicability and robustness.ObjectiveWe aimed to develop a machine learning model utilizing clinical factors to predict first-year mortality in patients undergoing hemodialysis that could assist physicians in classifying high-risk patients.MethodsTraining and testing cohorts consisted of 5351 patients from a single center and 5828 patients from 97 renal centers undergoing hemodialysis (incident only). The outcome was all-cause mortality during the first year of dialysis. Extreme gradient boosting was used for algorithm training and validation. Two models were established based on the data obtained at dialysis initiation (model 1) and data 0-3 months after dialysis initiation (model 2), and 10-fold cross-validation was applied to each model. The area under the curve (AUC), sensitivity (recall), specificity, precision, balanced accuracy, and F1 score were used to assess the predictive ability of the models.ResultsIn the training and testing cohorts, 585 (10.93%) and 764 (13.11%) patients, respectively, died during the first-year follow-up. Of 42 candidate features, the 15 most important features were selected. The performance of model 1 (AUC 0.83, 95% CI 0.78-0.84) was similar to that of model 2 (AUC 0.85, 95% CI 0.81-0.86).ConclusionsWe developed and validated 2 machine learning models to predict first-year mortality in patients undergoing hemodialysis. Both models could be used to stratify high-risk patients at the early stages of dialysis.

Current symptom-based risk scores for stable coronary artery disease evaluation are not applicable in COPD patients.

BackgroundCardiovascular diseases are arguably the most important comorbidity in patients with COPD. Despite an increased prevalence of coronary artery disease (CAD) in COPD patients, there are no dedicated diagnostic recommendations.ObjectivesWe investigated whether COPD patients receive adequate primary evaluation of CAD despite overlapping symptoms.MethodsIn total, 302 patients with COPD who underwent invasive coronary angiography (ICA) were retrospectively matched (for age, body mass index and cardiovascular risk factors) with 302 patients without functional lung diseases. Quality and onset of symptoms prior to ICA were documented, and individual patients’ pretest probabilities according to European Society of Cardiology (ESC) guidelines were calculated. Endpoints were delay of ICA referral after symptom onset and clinical outcome, defined as subsequent revascularisation.ResultsMean delay between symptom onset and ICA was 19.9±22.0 months in COPD patients compared to 8.3±12.7 months in the control group (p<0.0001). COPD patients had a lower rate of typical chest pain (25.2% versus 38.1%, p=0.0009), and dyspnoea only (18.2% versus 26.8%, p=0.015). Sub-analysis of Global Initiative for Chronic Obstructive Lung Disease (GOLD) grades revealed an incremental delay with increasing COPD severity: GOLD 1: 16.1±17.3 months; GOLD 2: 17.6±22.1 months; GOLD 3: 20.1±21.3 months; and GOLD 4: 24.2±23.4 months. Overall significant CAD prevalence (>70% stenosis) was 35.3%; the revascularisation rate increased with higher pretest probability for the control group but decreased for patients with COPD GOLD 1–4.ConclusionPatients with COPD are insufficiently evaluated for CAD due to overlapping symptoms. Current CAD risk scores for stable chest pain appear inappropriate for patients with COPD.

Influence of electrolyte balance on the prognosis of long-term cardiovascular events after acute coronary syndrome

Aim. To assess the effect of electrolyte changes on the prognosis of long-term cardiovascular events after acute coronary syndrome (ACS).Material and methods. The study included 105 patients with ACS who underwent coronary angiography (CA) with coronary stenting. At the study inclusion (before CA with coronary stenting), we collected data on traditional risk factors, analyzed levels of urinary sodium and potassium, kaliuresis and natriuresis. Free water clearance (FWC) and electrolyte free water clearance (EFWC), as well as fluid balance using bioelectrical impedance analysis were determined. Study endpoints (fatal and nonfatal cardiovascular events) were determined 6,2±0,2 months after CA with coronary stenting.Results. It was found that a decrease in urinary sodium (χ2=5,64, p=0,02, Constanta B0 =-0,62, Estimate =-16,5) and natriuresis (χ2=4,1, р=0,044, Constanta B0 =-1,38, Estimate =-5,2) increase the death risk. Urinary sodium of 0,2 mol/L and natriuresis of 0,5 mol are threshold levels of increased risk of death. Urinary potassium decrease was associated with an increase in death risk (threshold level — 0,5 mol/L, χ2=4,99, р=0,025, Constanta B0 =-0,63, Estimate =-70,4) and acute myocardial infarction (threshold level — 0,06 mol/L, χ2=3,93, р=0,04, Constanta B0 =-0,99, Estimate =-58,0) in the long-term period. Increase in EFWC increased the likelihood of long-term transient ischemic attack after ACS (χ2=4,61, р=0,03, Constanta B0 =-2,95, Estimate =-1,0). There were no significant relationships in the matter of FWC (p&gt;0,05). However, with a decrease in intracellular fluid volume compared to normal values and a decrease in FWC or an increase in EFWC, the likelihood of longterm composite endpoints after ACS increases.Conclusion. As a result of the study, risk markers for long-term fatal and non-fatal cardiovascular events after ACS were established: decrease in urinary sodium &lt;0,2 mol/l and potassium &lt;0,5 mol/l; decrease in FWC and increase in EFWC with or without cellular dehydration. The established markers can complement the current cardiovascular risk score methods in patients with ACS.

Machine learning-based risk prediction of intrahospital clinical outcomes in patients undergoing TAVI.

Currently, patient selection in TAVI is based upon a multidisciplinary heart team assessment of patient comorbidities and surgical risk stratification. In an era of increasing need for precision medicine and quickly expanding TAVI indications, machine learning has shown promise in making accurate predictions of clinical outcomes. This study aims to predict different intrahospital clinical outcomes in patients undergoing TAVI using a machine learning-based approach. The main clinical outcomes include all-cause mortality, stroke, major vascular complications, paravalvular leakage, and new pacemaker implantations. The dataset consists of 451 consecutive patients undergoing elective TAVI between February 2014 and June 2016. The applied machine learning methods were neural networks, support vector machines, and random forests. Their performance was evaluated using five-fold nested cross-validation. Considering all 83 features, the performance of all machine learning models in predicting all-cause intrahospital mortality (AUC 0.94-0.97) was significantly higher than both the STS risk score (AUC 0.64), the STS/ACC TAVR score (AUC 0.65), and all machine learning models using baseline characteristics only (AUC 0.72-0.82). Using an extreme boosting gradient, baseline troponin T was found to be the most important feature among all input variables. Overall, after feature selection, there was a slightly inferior performance. Stroke, major vascular complications, paravalvular leakage, and new pacemaker implantations could not be accurately predicted. Machine learning has the potential to improve patient selection and risk management of interventional cardiovascular procedures, as it is capable of making superior predictions compared to current logistic risk scores.

Assessment of HCC Risk in Patients with Chronic HBV (REACH, PAGE-B, and Beyond)

To provide an updated overview on the current status of risk prediction scores for hepatocellular carcinoma (HCC) in patients with chronic hepatitis B virus infection. More HCC risk scores have been developed, validated, and optimized (e.g., with liver stiffness measurement) in different patient and ethnic groups. Risk scores for treated patients with high negative predictive values would be able to identify patients who may not need HCC surveillance anymore as their HCC risk has been reduced by antiviral therapy. Current HCC risk scores can accurately predict HCC in specific populations, in both treatment-naive patients and those receiving antiviral therapy. Different levels of care and different intensities of HCC surveillance should be offered according to the risk profile of patients.

Development and validation of a tool for predicting type 2 diabetes in Mexican women of reproductive age

IntroductionDiabetes is a worldwide problem with a greater impact in developing countries, where many people are unaware of their risk. In Mexico, women show the greatest risk for T2D. Current risk scores have been developed and validated in predominantly older European cohorts. They are not the best option in Mexican women. The development of a risk model/score in this population would be useful. ObjectiveTo develop and validate a risk model and score that incorporates the most relevant risk factors for T2D in Mexican women of reproductive age. MethodsThe study was carried out in two phases, with the first phase being the development of the predictive model and the second phase the validation of the model in a separate independent population. A cohort of Mexican patients of reproductive age (“Derivation Cohort”) was used to create the predictive model. It included data on 3161 women. Risk factors for identification were assessed using Cox proportional hazards regression. Finally a score with a range of 0 to 19 points was developed to identify the 2.4 year probability of developing DM2 in Mexican women of reproductive age. Results147 new cases of T2D (4.6%) were identified in the Derivation Cohort model, 97 of 925 participants (10.48%) in the validation cohort. The risk factor predictors of T2D were: history of gestational diabetes (HR 2.69, 95% CI 1.10–6.58), BMI (HR 1.03, 95% CI 1.01–1.06), hypertriglyceridemia (HR 1.54, 95% CI 1.11–2.14) and fasting blood glucose (HR 1.06, 95% CI 1.05–1.08), with an AUC of 0.75. The AUC in the validation cohort was 0.91 (95% CI 0.87–0.94). The score had a sensitivity of 73% and specificity of 67% at a cutoff of ≥15. ConclusionsA predictive model and risk score was developed to detect cases at risk for incident T2D. It was generated using the characteristics of Mexican women of reproductive age. This risk score is a step forward in attempting to address the generational legacy that diabetes in pregnancy could have on women and their children.

Delayed prefrontal hemodynamic response associated with suicide risk in autism spectrum disorder

Adults diagnosed with Autism spectrum disorder (ASD) are at high risk of experiencing suicidality compared with other clinical groups. Recently, near-infrared spectroscopy (NIRS) studies have investigated the association between frontotemporal functional abnormalities and suicidality in patients with mood disorders. However, whether these prefrontal hemodynamic responses are associated with suicide vulnerability in individuals with ASD remains unclear. Here, we used 24-channel NIRS to examine the characteristics of prefrontal hemodynamic responses during a verbal fluency task in 20 adults with ASD and in age-, sex-, and intelligence quotient-matched healthy controls. In addition, we used Spearman's correlation analysis to identify the relationship between the time-course of prefrontal hemodynamic activation and the current suicide risk in patients with ASD. We found no significant differences between the verbal fluency task-induced prefrontal hemodynamic responses in the ASD vs. control group. However, we found a significant positive correlation between the current suicide risk score and the time-course of prefrontal hemodynamic activation in the ASD group. Thus, the 24-channel NIRS system appears to be useful in assessing suicide risk in individuals with ASD.

Development and validation of a deep neural network model to predict postoperative mortality, acute kidney injury, and reintubation using a single feature set

During the perioperative period patients often suffer complications, including acute kidney injury (AKI), reintubation, and mortality. In order to effectively prevent these complications, high-risk patients must be readily identified. However, most current risk scores are designed to predict a single postoperative complication and often lack specificity on the patient level. In other fields, machine learning (ML) has been shown to successfully create models to predict multiple end points using a single input feature set. We hypothesized that ML can be used to create models to predict postoperative mortality, AKI, reintubation, and a combined outcome using a single set of features available at the end of surgery. A set of 46 features available at the end of surgery, including drug dosing, blood loss, vital signs, and others were extracted. Additionally, six additional features accounting for total intraoperative hypotension were extracted and trialed for different models. A total of 59,981 surgical procedures met inclusion criteria and the deep neural networks (DNN) were trained on 80% of the data, with 20% reserved for testing. The network performances were then compared to ASA Physical Status. In addition to creating separate models for each outcome, a multitask learning model was trialed that used information on all outcomes to predict the likelihood of each outcome individually. The overall rate of the examined complications in this data set was 0.79% for mortality, 22.3% (of 21,676 patients with creatinine values) for AKI, and 1.1% for reintubation. Overall, there was significant overlap between the various model types for each outcome, with no one modeling technique consistently performing the best. However, the best DNN models did beat the ASA score for all outcomes other than mortality. The highest area under the receiver operating characteristic curve (AUC) models were 0.792 (0.775–0.808) for AKI, 0.879 (0.851–0.905) for reintubation, 0.907 (0.872–0.938) for mortality, and 0.874 (0.864–0.866) for any outcome. The ASA score alone achieved AUCs of 0.652 (0.636–0.669) for AKI, 0.787 (0.757–0.818) for reintubation, 0.839 (0.804–0.875) for mortality, and 0.76 (0.748–0.773) for any outcome. Overall, the DNN architecture was able to create models that outperformed the ASA physical status to predict all outcomes based on a single feature set, consisting of objective data available at the end of surgery. No one model architecture consistently performed the best.

Hemodynamics of Cerebral Aneurysms: Connecting Medical Imaging and Biomechanical Analysis

In the last two decades, numerous studies have conducted patient-specific computations of blood flow dynamics in cerebral aneurysms and reported correlations between various hemodynamic metrics and aneurysmal disease progression or treatment outcomes. Nevertheless, intra-aneurysmal flow analysis has not been adopted in current clinical practice, and hemodynamic factors usually are not considered in clinical decision making. This review presents the state of the art in cerebral aneurysm imaging and image-based modeling, discussing the advantages and limitations of each approach and focusing on the translational value of hemodynamic analysis. Combining imaging and modeling data obtained from different flow modalities can improve the accuracy and fidelity of resulting velocity fields and flow-derived factors that are thought to affect aneurysmal disease progression. It is expected that predictive models utilizing hemodynamic factors in combination with patient medical history and morphological data will outperform current risk scores and treatment guidelines. Possible future directions include novel approaches enabling data assimilation and multimodality analysis of cerebral aneurysm hemodynamics.

Association of NPAC score with survival after acute myocardial infarction

Background and aimsRisk stratification in acute myocardial infarction (AMI) is important for guiding clinical management. Current risk scores are mostly derived from clinical trials with stringent patient selection. We aimed to establish and evaluate a composite scoring system to improve short-term mortality classification after index episodes of AMI, independent of electrocardiography (ECG) pattern, in a large real-world cohort. MethodsUsing electronic health records, patients admitted to our regional teaching hospital (derivation cohort, n = 1747) and an independent tertiary care center (validation cohort, n = 1276), with index acute myocardial infarction between January 2013 and December 2017, as confirmed by principal diagnosis and laboratory findings, were identified retrospectively. ResultsUnivariate logistic regression was used as the primary model to identify potential contributors to mortality. Stepwise forward likelihood ratio logistic regression revealed that neutrophil-to-lymphocyte ratio, peripheral vascular disease, age, and serum creatinine (NPAC) were significant for 90-day mortality (Hosmer- Lemeshow test, p = 0.21). Each component of the NPAC score was weighted by beta-coefficients in multivariate analysis. The C-statistic of the NPAC score was 0.75, which was higher than the conventional Charlson's score (C-statistic = 0.63). Judicious application of a deep learning model to our dataset improved the accuracy of classification with a C-statistic of 0.81. ConclusionsThe NPAC score comprises four items from routine laboratory parameters to basic clinical information and can facilitate early identification of cases at risk of short-term mortality following index myocardial infarction. Deep learning model can serve as a gatekeeper to facilitate clinical decision-making.

CORONARY ARTERY CALCIUM SCORE AND CARDIOVASCULAR OUTCOMES IN PATIENTS WITH MILD TO MODERATE CHRONIC KIDNEY DISEASE INSIGHTS FROM CLARIFY

Chronic kidney disease (CKD) is associated with increased risk of cardiovascular disease. Current cardiovascular risk scores are not well validated in patients with CKD. We sought to study the utility of coronary artery calcium (CAC) scoring in risk stratification of patients with CKD. Community

Efficacy of bleeding risk scores in elderly Algerian patients with acute coronary syndromes

The incidence of acute coronary syndromes is high in the elderly population. Bleeding is associated with a poorer prognosis in this clinical setting. The available bleeding risk scores have not been validated specifically in the elderly. Our aim was to assess predictive ability of the most important bleeding risk scores in Algerian patients with acute coronary syndrome aged > 75 years. We prospectively included consecutive acute coronary syndromes patients. Baseline characteristics, laboratory findings, and hemodynamic data were collected. In-hospital bleeding was defined according to CRUSADE, Mehran, ACTION, and BARC definitions. CRUSADE, Mehran, and ACTION bleeding risk scores were calculated for each patient. The ability of these scores to predict major bleeding was assessed by binary logistic regression, receiver operating characteristic curves, and area under the curves. We included 1145 patients, with mean age of 61.2 years; 217 patients (19%) were > 75 years. Older patients had higher bleeding risk (CRUSADE, 42 vs. 22; Mehran, 25 vs. 15; ACTION, 36 vs. 28; P < 0.001) and a slightly higher incidence of major bleeding events (CRUSADE bleeding, 5.1% vs. 3.8%; P = 0.250). The predictive ability of these 3 scores was lower in the elderly (area under the curve, CRUSADE: 0.63 in older patients, 0.81 in young patients; P = 0.027; Mehran: 0.67 in older patients, 0.73 in younger patients; P = 0.340; ACTION: 0.58 in older patients, 0.75 in younger patients; P = 0.041). Current bleeding risk scores showed poorer predictive performance in elderly patients with acute coronary syndromes than in younger patients.