Limits...
Cardiac Health Risk Stratification System (CHRiSS): a Bayesian-based decision support system for left ventricular assist device (LVAD) therapy.

Loghmanpour NA, Druzdzel MJ, Antaki JF - PLoS ONE (2014)

Bottom Line: 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.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.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States of America.

ABSTRACT
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.

Show MeSH

Related in: MedlinePlus

2 Year Bayesian Model.Node colors: red =  class, blue =  parent, purple =  child, yellow =  spouse. Question marks identify nodes that do not have specific evidence set and use the population distribution as the prior distribution. LVEDD =  left ventricle end diastolic diameter, BNP =  B-type natriuretic peptide, LVEF =  left ventricle ejection fraction, GI =  gastrointestinal, IV =  intravenous, ICD =  implantable cardioverter defibrillator, PVD =  peripheral vascular disease, hx HIV =  history of human immunodeficiency virus, BMI =  body mass index, MCS =  mechanical circulatory support, CV =  cardiovascular.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4232308&req=5

pone-0111264-g005: 2 Year Bayesian Model.Node colors: red =  class, blue =  parent, purple =  child, yellow =  spouse. Question marks identify nodes that do not have specific evidence set and use the population distribution as the prior distribution. LVEDD =  left ventricle end diastolic diameter, BNP =  B-type natriuretic peptide, LVEF =  left ventricle ejection fraction, GI =  gastrointestinal, IV =  intravenous, ICD =  implantable cardioverter defibrillator, PVD =  peripheral vascular disease, hx HIV =  history of human immunodeficiency virus, BMI =  body mass index, MCS =  mechanical circulatory support, CV =  cardiovascular.

Mentions: The optimized BNs can be seen in Figures 1–5, and a summary of their performance can be found in Table 2. Accuracy was greatest for the 30 day model with 95% and lowest for the 2 year model with 71%. The True Positive (proportion of patients who were correctly predicted to not survive) was greatest for the 2 year at 65% and lowest for the 90 day model at 23%. The True Negative (proportion of patients who were correctly predicted to survive past the endpoint) was greatest for the 30 day model at nearly 100% and lowest for the 2 year model at 76%. The ROC % was greatest for the 30 day model at 93% and lowest for the 6 month, 1 year and 2 year models (all 71%).


Cardiac Health Risk Stratification System (CHRiSS): a Bayesian-based decision support system for left ventricular assist device (LVAD) therapy.

Loghmanpour NA, Druzdzel MJ, Antaki JF - PLoS ONE (2014)

2 Year Bayesian Model.Node colors: red =  class, blue =  parent, purple =  child, yellow =  spouse. Question marks identify nodes that do not have specific evidence set and use the population distribution as the prior distribution. LVEDD =  left ventricle end diastolic diameter, BNP =  B-type natriuretic peptide, LVEF =  left ventricle ejection fraction, GI =  gastrointestinal, IV =  intravenous, ICD =  implantable cardioverter defibrillator, PVD =  peripheral vascular disease, hx HIV =  history of human immunodeficiency virus, BMI =  body mass index, MCS =  mechanical circulatory support, CV =  cardiovascular.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4232308&req=5

pone-0111264-g005: 2 Year Bayesian Model.Node colors: red =  class, blue =  parent, purple =  child, yellow =  spouse. Question marks identify nodes that do not have specific evidence set and use the population distribution as the prior distribution. LVEDD =  left ventricle end diastolic diameter, BNP =  B-type natriuretic peptide, LVEF =  left ventricle ejection fraction, GI =  gastrointestinal, IV =  intravenous, ICD =  implantable cardioverter defibrillator, PVD =  peripheral vascular disease, hx HIV =  history of human immunodeficiency virus, BMI =  body mass index, MCS =  mechanical circulatory support, CV =  cardiovascular.
Mentions: The optimized BNs can be seen in Figures 1–5, and a summary of their performance can be found in Table 2. Accuracy was greatest for the 30 day model with 95% and lowest for the 2 year model with 71%. The True Positive (proportion of patients who were correctly predicted to not survive) was greatest for the 2 year at 65% and lowest for the 90 day model at 23%. The True Negative (proportion of patients who were correctly predicted to survive past the endpoint) was greatest for the 30 day model at nearly 100% and lowest for the 2 year model at 76%. The ROC % was greatest for the 30 day model at 93% and lowest for the 6 month, 1 year and 2 year models (all 71%).

Bottom Line: 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.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.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States of America.

ABSTRACT
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.

Show MeSH
Related in: MedlinePlus