Risk-prediction tool for early TAVR mortality

Experts in the Society for Thoracic Surgeons and the American College of Cardiology used data from more than 13,000 consecutive transcatheter aortic valve replacement procedures to develop a new tool for predicting the risk of in-hospital mortality in patients undergoing TAVR, according to a report published online March 9 in JAMA Cardiology.

Their risk-prediction model was only “modestly” accurate but performed better than any existing methods for assessing risk in this patient population. It should be considered the first iteration of this tool and will be modified as the procedure itself evolves and as more data concerning TAVR are collected and analyzed. Ongoing analysis “may well define clinical subsets of patients who accrue particular benefit from the procedure or, conversely, reveal subsets not well served by TAVR,” said Dr. Fred H. Edwards and his associates on the steering committee of the STS/ACC Transcatheter Valve Therapy Registry.

They noted that more models soon will be developed to predict 30-day and 1-year mortality after TAVR. Models to predict the risk of neurologic deficit following TAVR are currently being developed, and models for other nonfatal outcomes will be developed soon.

This tool predicting in-hospital mortality is expected to become “a valuable adjunct for patient counseling, performance assessment, local quality improvement, and national monitoring of the appropriateness of patient selection for TAVR,” said Dr. Edwards, who is also in the department of surgery, University of Florida, Jacksonville, and his associates.

They began by analyzing the registry data for virtually every commercial TAVR performed at 265 participating sites in the United States during a 27-month period. In general, patients were selected for TAVR because they were considered unsuitable candidates for surgical aortic valve replacement. The mean patient age was 82.1 years. A total of 730 patients died before leaving the hospital, for an in-hospital mortality of 5.3%.

Working from an initial list of 39 possible patient variables to include in their statistical prediction model, the researchers narrowed it down to the 7 most predictive factors available in the registry data: older age, poorer glomerular filtration rate, the need for hemodialysis, NYHA class IV status, the presence of severe chronic lung disease, a category 2 or 4 critical hemodynamic state (i.e., preprocedural acuity status), and need for a nonfemoral approach during the procedure.

The model was then tested in a separate validation cohort of 6,868 patients (52% men) treated at 314 sites during a 7-month period. It performed better at predicting in-hospital mortality than did either the EuroSCORE (European System for Cardiac Operative Risk Evaluation) or the FRANCE 2 (French Aortic National Corevalve and Edwards 2) models.

This STS/ACC model should assist clinicians in patient selection for TAVR, not by dictating which patients are candidates for TAVR, but by being used as “one element in the selection process, to be considered in concert with history, physical examination, laboratory information, and clinical judgment. The model may also provide useful information for patient counseling,” the investigators said (JAMA Cardiol. 2016 Mar 9. doi: 10.1001/jamacardiol.2015.0326). One factor that is generally recognized as an important risk predictor but isn’t yet incorporated into this tool is a measure of patient frailty. Data on frailty are not yet collected consistently in the STS/ACC registry. As more complete data become available, frailty likely will be included as a predictive factor in this tool.

Another important issue that eventually should be considered alongside survival prediction is the effect TAVR has on quality of life. The STS/ACC registry “is one of the few clinical registries to collect quality-of-life data,” and it could prove to be a critical adjunct to patient selection. A given patient might have a favorable outlook regarding mortality after the procedure, but would still be a poor candidate if he or she wouldn’t derive significant benefit from it, Dr. Edwards and his associates said.

Experts in the Society for Thoracic Surgeons and the American College of Cardiology used data from more than 13,000 consecutive transcatheter aortic valve replacement procedures to develop a new tool for predicting the risk of in-hospital mortality in patients undergoing TAVR, according to a report published online March 9 in JAMA Cardiology.

Their risk-prediction model was only “modestly” accurate but performed better than any existing methods for assessing risk in this patient population. It should be considered the first iteration of this tool and will be modified as the procedure itself evolves and as more data concerning TAVR are collected and analyzed. Ongoing analysis “may well define clinical subsets of patients who accrue particular benefit from the procedure or, conversely, reveal subsets not well served by TAVR,” said Dr. Fred H. Edwards and his associates on the steering committee of the STS/ACC Transcatheter Valve Therapy Registry.

They noted that more models soon will be developed to predict 30-day and 1-year mortality after TAVR. Models to predict the risk of neurologic deficit following TAVR are currently being developed, and models for other nonfatal outcomes will be developed soon.

This tool predicting in-hospital mortality is expected to become “a valuable adjunct for patient counseling, performance assessment, local quality improvement, and national monitoring of the appropriateness of patient selection for TAVR,” said Dr. Edwards, who is also in the department of surgery, University of Florida, Jacksonville, and his associates.

They began by analyzing the registry data for virtually every commercial TAVR performed at 265 participating sites in the United States during a 27-month period. In general, patients were selected for TAVR because they were considered unsuitable candidates for surgical aortic valve replacement. The mean patient age was 82.1 years. A total of 730 patients died before leaving the hospital, for an in-hospital mortality of 5.3%.

Working from an initial list of 39 possible patient variables to include in their statistical prediction model, the researchers narrowed it down to the 7 most predictive factors available in the registry data: older age, poorer glomerular filtration rate, the need for hemodialysis, NYHA class IV status, the presence of severe chronic lung disease, a category 2 or 4 critical hemodynamic state (i.e., preprocedural acuity status), and need for a nonfemoral approach during the procedure.

The model was then tested in a separate validation cohort of 6,868 patients (52% men) treated at 314 sites during a 7-month period. It performed better at predicting in-hospital mortality than did either the EuroSCORE (European System for Cardiac Operative Risk Evaluation) or the FRANCE 2 (French Aortic National Corevalve and Edwards 2) models.

This STS/ACC model should assist clinicians in patient selection for TAVR, not by dictating which patients are candidates for TAVR, but by being used as “one element in the selection process, to be considered in concert with history, physical examination, laboratory information, and clinical judgment. The model may also provide useful information for patient counseling,” the investigators said (JAMA Cardiol. 2016 Mar 9. doi: 10.1001/jamacardiol.2015.0326). One factor that is generally recognized as an important risk predictor but isn’t yet incorporated into this tool is a measure of patient frailty. Data on frailty are not yet collected consistently in the STS/ACC registry. As more complete data become available, frailty likely will be included as a predictive factor in this tool.

Another important issue that eventually should be considered alongside survival prediction is the effect TAVR has on quality of life. The STS/ACC registry “is one of the few clinical registries to collect quality-of-life data,” and it could prove to be a critical adjunct to patient selection. A given patient might have a favorable outlook regarding mortality after the procedure, but would still be a poor candidate if he or she wouldn’t derive significant benefit from it, Dr. Edwards and his associates said.

Experts in the Society for Thoracic Surgeons and the American College of Cardiology used data from more than 13,000 consecutive transcatheter aortic valve replacement procedures to develop a new tool for predicting the risk of in-hospital mortality in patients undergoing TAVR, according to a report published online March 9 in JAMA Cardiology.

Their risk-prediction model was only “modestly” accurate but performed better than any existing methods for assessing risk in this patient population. It should be considered the first iteration of this tool and will be modified as the procedure itself evolves and as more data concerning TAVR are collected and analyzed. Ongoing analysis “may well define clinical subsets of patients who accrue particular benefit from the procedure or, conversely, reveal subsets not well served by TAVR,” said Dr. Fred H. Edwards and his associates on the steering committee of the STS/ACC Transcatheter Valve Therapy Registry.

They noted that more models soon will be developed to predict 30-day and 1-year mortality after TAVR. Models to predict the risk of neurologic deficit following TAVR are currently being developed, and models for other nonfatal outcomes will be developed soon.

This tool predicting in-hospital mortality is expected to become “a valuable adjunct for patient counseling, performance assessment, local quality improvement, and national monitoring of the appropriateness of patient selection for TAVR,” said Dr. Edwards, who is also in the department of surgery, University of Florida, Jacksonville, and his associates.

They began by analyzing the registry data for virtually every commercial TAVR performed at 265 participating sites in the United States during a 27-month period. In general, patients were selected for TAVR because they were considered unsuitable candidates for surgical aortic valve replacement. The mean patient age was 82.1 years. A total of 730 patients died before leaving the hospital, for an in-hospital mortality of 5.3%.

Working from an initial list of 39 possible patient variables to include in their statistical prediction model, the researchers narrowed it down to the 7 most predictive factors available in the registry data: older age, poorer glomerular filtration rate, the need for hemodialysis, NYHA class IV status, the presence of severe chronic lung disease, a category 2 or 4 critical hemodynamic state (i.e., preprocedural acuity status), and need for a nonfemoral approach during the procedure.

The model was then tested in a separate validation cohort of 6,868 patients (52% men) treated at 314 sites during a 7-month period. It performed better at predicting in-hospital mortality than did either the EuroSCORE (European System for Cardiac Operative Risk Evaluation) or the FRANCE 2 (French Aortic National Corevalve and Edwards 2) models.

This STS/ACC model should assist clinicians in patient selection for TAVR, not by dictating which patients are candidates for TAVR, but by being used as “one element in the selection process, to be considered in concert with history, physical examination, laboratory information, and clinical judgment. The model may also provide useful information for patient counseling,” the investigators said (JAMA Cardiol. 2016 Mar 9. doi: 10.1001/jamacardiol.2015.0326). One factor that is generally recognized as an important risk predictor but isn’t yet incorporated into this tool is a measure of patient frailty. Data on frailty are not yet collected consistently in the STS/ACC registry. As more complete data become available, frailty likely will be included as a predictive factor in this tool.

Another important issue that eventually should be considered alongside survival prediction is the effect TAVR has on quality of life. The STS/ACC registry “is one of the few clinical registries to collect quality-of-life data,” and it could prove to be a critical adjunct to patient selection. A given patient might have a favorable outlook regarding mortality after the procedure, but would still be a poor candidate if he or she wouldn’t derive significant benefit from it, Dr. Edwards and his associates said.