Natasha A. Loghmanpour

A new Bayesian network-based risk stratification model for prediction of short-term and long-term LVAD mortality.

Existing risk assessment tools for patient selection for left ventricular assist devices (LVADs) such as the Destination Therapy Risk Score and HeartMate II Risk Score (HMRS) have limited predictive ability. This study aims to overcome the limitations of traditional statistical methods by performing the first application of Bayesian analysis to the comprehensive Interagency Registry for Mechanically Assisted Circulatory Support dataset and comparing it to HMRS. We retrospectively analyzed 8,050 continuous flow LVAD patients and 226 preimplant variables. We then derived Bayesian models for mortality at each of five time end-points postimplant (30 days, 90 days, 6 month, 1 year, and 2 years), achieving accuracies of 95%, 90%, 90%, 83%, and 78%, Kappa values of 0.43, 0.37, 0.37, 0.45, and 0.43, and area under the receiver operator characteristic (ROC) of 91%, 82%, 82%, 80%, and 81%, respectively. This was in comparison to the HMRS with an ROC of 57% and 60% at 90 days and 1 year, respectively. Preimplant interventions, such as dialysis, ECMO, and ventilators were major contributing risk markers. Bayesian models have the ability to reliably represent the complex causal relations of multiple variables on clinical outcomes. Their potential to develop a reliable risk stratification tool for use in clinical decision making on LVAD patients encourages further investigation.

Cardiac Health Risk Stratification System (CHRiSS): A Bayesian-Based Decision Support System for Left Ventricular Assist Device (LVAD) Therapy

This study investigated the use of Bayesian Networks (BNs) for left ventricular assist device (LVAD) therapy; a treatment for end-stage heart failure that has been steadily growing in popularity over the past decade. Despite this growth, the number of LVAD implants performed annually remains a small fraction of the estimated population of patients who might benefit from this treatment. We believe that this demonstrates a need for an accurate stratification tool that can help identify LVAD candidates at the most appropriate point in the course of their disease. We derived BNs to predict mortality at five endpoints utilizing the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) database: containing over 12,000 total enrolled patients from 153 hospital sites, collected since 2006 to the present day, and consisting of approximately 230 pre-implant clinical variables. Synthetic minority oversampling technique (SMOTE) was employed to address the uneven proportion of patients with negative outcomes and to improve the performance of the models. The resulting accuracy and area under the ROC curve (%) for predicted mortality were 30 day: 94.9 and 92.5; 90 day: 84.2 and 73.9; 6 month: 78.2 and 70.6; 1 year: 73.1 and 70.6; and 2 years: 71.4 and 70.8. To foster the translation of these models to clinical practice, they have been incorporated into a web-based application, the Cardiac Health Risk Stratification System (CHRiSS). As clinical experience with LVAD therapy continues to grow, and additional data is collected, we aim to continually update these BN models to improve their accuracy and maintain their relevance. Ongoing work also aims to extend the BN models to predict the risk of adverse events post-LVAD implant as additional factors for consideration in decision making.

Simulation of dilated heart failure with continuous flow circulatory support.

Lumped parameter models have been employed for decades to simulate important hemodynamic couplings between a left ventricular assist device (LVAD) and the native circulation. However, these studies seldom consider the pathological descending limb of the Frank-Starling response of the overloaded ventricle. This study introduces a dilated heart failure model featuring a unimodal end systolic pressure-volume relationship (ESPVR) to address this critical shortcoming. The resulting hemodynamic response to mechanical circulatory support are illustrated through numerical simulations of a rotodynamic, continuous flow ventricular assist device (cfVAD) coupled to systemic and pulmonary circulations with baroreflex control. The model further incorporated septal interaction to capture the influence of left ventricular (LV) unloading on right ventricular function. Four heart failure conditions were simulated (LV and bi-ventricular failure with/without pulmonary hypertension) in addition to normal baseline. Several metrics of LV function, including cardiac output and stroke work, exhibited a unimodal response whereby initial unloading improved function, and further unloading depleted preload reserve thereby reducing ventricular output. The concept of extremal loading was introduced to reflect the loading condition in which the intrinsic LV stroke work is maximized. Simulation of bi-ventricular failure with pulmonary hypertension revealed inadequacy of LV support alone. These simulations motivate the implementation of an extremum tracking feedback controller to potentially optimize ventricular recovery.

Low Accuracy of the HeartMate Risk Score for Predicting Mortality using the INTERMACS Registry Data.

Selection is a key determinant of clinical outcomes after left ventricular assist device (LVAD) placement in patients with end-stage heart failure. The HeartMate II risk score (HMRS) has been proposed to facilitate risk stratification and patient selection for continuous flow pumps. This study retrospectively assessed the performance of HMRS in predicting 90 day and 1 year mortality in patients within the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS). A total of 11,523 INTERMACS patients who received a continuous flow LVAD between 2010 and 2015 were retrospectively categorized per their calculated HMRS to predict their 90 day and 1 year risk of mortality. The performance of the score was evaluated by the area under curve (AUC) of the receiver operator characteristic. We also performed multiple regression analysis using variables from the HMRS calculation on the INTERMACS data. The HMRS model showed moderate discrimination for both 90 day and 1 year mortality prediction with AUCs of 61% and 59%, respectively. The predictions had similar accuracy irrespective of whether the pump was axial or centrifugal flow. Multivariable analysis using independent variables used in the original HMRS analysis revealed different set of variables to be predictive of 90 day mortality than those used to calculate HMRS. HMRS predicts both 90 day and 1 year mortality with poor discrimination when applied to a large cohort of LVAD patients. Newer risk prediction models are therefore needed to optimize the therapeutic application of LVAD therapy. Patient selection for appropriate use of LVADs is critical. Currently available risk stratification tools (HMRS) continue to be limited in their ability to accurately predict mortality after LVAD. This study highlights these limitations when applied to a large, comprehensive, multicenter database. HMRS predicts mortality with only modest discrimination when applied to a large cohort of LVAD patients. Better risk stratification tools are needed to optimize outcomes.

Reproducibility of IVUS border detection for carotid atherosclerotic plaque assessment

Plaque composition is a potentially important diagnostic feature for carotid artery stenting (CAS). The purpose of this investigation is to evaluate the reproducibility of manual border correction in intravascular ultrasound with virtual histology (VH IVUS) images. Three images each were obtained from 51 CAS datasets on which automatic border detection was corrected manually by two trained observers. Plaque was classified using the definitions from the CAPITAL (Carotid Artery Plaque Virtual Histology Evaluation) study, listed in order from least to most pathological: no plaque, pathological intimal thickening, fibroatheroma, fibrocalcific, calcified fibroatheroma, thin-cap fibroatheroma, and calcified thin-cap fibroatheroma. Inter-observer variability was quantified using both weighted and unweighted Kappa statistics. Bland-Altman analysis was used to compare the cross-sectional areas of the vessel and lumen. Agreement using necrotic core percentage as the criterion was evaluated using the unweighted Kappa statistic. Agreement between classifications of plaque type was evaluated using the weighted Kappa statistic. There was substantial agreement between the observers based on necrotic core percentage (κ=0.63), while the agreement was moderate (κ(quadratic)=0.60) based on plaque classification. Due to the time-consuming nature of manual border detection, an improved automatic border detection algorithm is necessary for using VH IVUS as a diagnostic tool for assessing the suitability of patients with carotid artery occlusive disease for CAS.

Development of Predictive Models for Continuous Flow Left Ventricular Assist Device Patients using Bayesian Networks

Background : Existing prognostic tools for patient selection for ventricular assist devices (VADs) such as the Destination Therapy Risk Score (DTRS) and newly published HeartMate II Risk Score (HMRS) have limited predictive ability, especially with the current generation of continuous flow VADs (cfVADs). This study aims to use a modern machine learning approach, employing Bayesian Networks (BNs), which overcomes some of the limitations of traditional statistical methods. Methods : Retrospective data from 144 patients at Allegheny General Hospital and Integris Health System from 2007 to 2011 were analyzed. 43 data elements were grouped into four sets: demographics, laboratory tests, hemodynamics, and medications. Patients were stratified by survival at 90 days post LVAD. Results : The independent variables were ranked based on their predictive power and reduced to an optimal set of 10: hematocrit, aspartate aminotransferase, age, heart rate, transpulmonary gradient, mean pulmonary artery pressure, use of diuretics, platelet count, blood urea nitrogen and hemoglobin. Two BNs, NaA¯ve Bayes (NB) and Tree-Augmented NaA¯ve Bayes (TAN) outperformed the DTRS in identifying low risk patients (specificity: 91% and 93% vs. 78%) and outperformed HMRS predictions of high risk patients (sensitivity: 80% and 60% vs. 25%). Both models were more accurate than DTRS and HMRS (90% vs. 73% and 84%), Kappa (NB: 0.56 TAN: 0.48, DTRS: 0.14, HMRS: 0.22), and AUC (NB: 80%, TAN: 84%, DTRS: 59%, HMRS: 59%). Conclusion : The Bayesian Network models developed in this study consistently outperformed the DTRS and HMRS on all metrics. An added advantage is their intuitive graphical structure that closely mimics natural reasoning patterns. This warrants further investigation with an expanded patient cohort, and inclusion of adverse event outcomes.

Prediction Model For Cardiac Allograft Vasculopathy: Comparison of Three Multivariable Methods

Cardiac allograft vasculopathy (CAV) remains an important cause of graft failure after heart transplantation (HT). Although many risk factors for CAV have been identified, there are no clinical prediction models that enable clinicians to determine each recipients risk of CAV.