Predictors Of One-year Mortality After Transcatheter Aortic Valve Replacement: Role Of Albumin And Meld Score
Kendra J. Grubb, Ross Micciche, Michael P. Flaherty, Jaimin R. Trivedi.
University of Louisville, Louisville, KY, USA.
OBJECTIVE: Transcatheter Aortic Valve Replacement (TAVR) outcomes have improved and as we look to the low-risk patient population we must be critical of patient selection. It is well known that existing metrics such as Society of Thoracic Surgeons (STS) risk score and EURO-Score do not accurately predict outcomes after TAVR and new risk models have been proposed.
METHODS: A single center Transcatheter Valve Therapy database was queried between January 2012 and December 2015 to identify patients undergoing TAVR procedures. Creatinine, bilirubin, and INR were used to generate a Model for End-Stage Liver Disease (MELD) score. A multivariate logistic regression model was generated to identify pre-operative factors associated with 1-year mortality. The multivariate model included factors such as: age, serum albumin, MELD score, Pre-procedure hemoglobin, sex, body mass index (BMI), chronic use of home oxygen (home O2), previous myocardial infarction (MI), diabetes, and prior aortic valve replacement.
RESULTS: There were a total of 231 non-research patients included in the study, of which 21 (9%) had died at 1 year. The median age was 81 years and 44% were females. The median STS 30-day mortality risk was 9.2% and median KCCQ score was 40. The multivariate logistic regression model (C-statistics -0.71) showed lower (<3.2) serum albumin (OR = 0.36, CI = 0.14-0.94, p=0.03) and higher (>14) MELD score (OR = 1.09, CI = 1.01-1.19, p=0.03) at time of TAVR were independent predictors of 1-year mortality. Age, Pre-procedure hemoglobin, sex, BMI, home O2, previous MI, diabetes, and prior aortic valve replacement were not significantly associated with 1-year mortality.
CONCLUSIONS: In a single center cohort of high-risk patients undergoing TAVR, serum albumin and MELD score are independent predictors of 1-year mortality. A study using a larger cohort to validate risk using albumin and MELD score can be beneficial, in addition to the STS risk score, in predicting TAVR patient outcomes.
Table 1. Multivariate logistic regression model for 1-year post TAVR mortality
Back to 2017 Posters