Abstract 10869: Using All-Cause Mortality to Determine the Best Method for Indexation of Echocardiographic Measures According to Body Size in Obese and Non-Obese Patients

Abstract

Introduction: Body surface area (BSA) is the most widely accepted metric for body size indexation of cardiac measures to improve diagnosis, but its use in obesity is questioned. Furthermore, big data analyses comparing indexation metrics are limited. Hypothesis: All-cause mortality can be used to identify an optimal indexation metric for transthoracic echocardiographic (TTE) measures that will be a better predictor of survival than BSA regardless of obesity (body mass index [BMI] ≥30kg/m2). Methods: Patients (n=11621) with no prior cardiac surgery, with all TTE measures of interest available, were selected from the National Echo Database Australia. Cox regression survival analysis for all-cause mortality (Wald chi-square) was analyzed for TTE measures both unindexed and indexed to different indexation metrics. Results: For both non-obese (n=7684, median BMI 25 kg/m2, age 65 years, 42% female, 4.2 years follow-up, 1586 deaths) and obese patients (n=3937, BMI 34 kg/m2, age 63 years, 42% female, 4.5 years follow-up, 607 deaths), indexation of cardiac sizes (atrial areas, left ventricular mass and diameter, and aortic sinus diameter) to BSA had better prognostic performance (higher chi-square) vs unindexed measures. Indexing by height^2.7 performed worse than BSA. Cardiac output had the best prognostic performance when indexed by weight^3 vs all other metrics (>3-fold higher chi-square). Stroke volume performed better unindexed than when indexed for body size. Conclusions: When using all-cause mortality as the arbiter of appropriateness among both non-obese and obese patients, indexation by BSA improved prognostic performance, cardiac output performed best when indexed by weight^3, and stroke volume performed best unindexed.