To determine the prevalence of low skeletal muscle mass in patients undergoing transcatheter aortic valve replacement (TAVR) and whether skeletal muscle mass measured from preoperative computed tomography (CT) images provides value in predicting postoperative length of stay (LOS).
Background: There are limited data on the use of body composition as a frailty measure in TAVR patients and no studies have determined if this measure predicts LOS.
Methods: We studied 104 consecutive patients who underwent TAVR at Tallahassee Memorial Hospital from 2012 to 2016. Patient demographics, standard frailty measures (hand grip, albumin, and 5-m walk test), clinical comorbidities, echocardiographic data, and Valve Academic Research Consortium II major complications were recorded prospectively. Skeletal muscle index (SMI) [skeletal muscle mass cross-sectional area at L3/height2] was measured from CT images using Slice-O-Matic software (Tomovision, Montreal, Quebec, Canada). Clinical outcomes were assessed and multivariate methods used to determine predictors of LOS.
Results: Sarcopenia was prevalent in men (83%) and women (56%). Patients who suffered from a major complication had significantly longer length of stay (13 vs 4.6days, P<.0001). Skeletal muscle index correlated with age, sex, body mass index, handgrip strength, and previous coronary artery bypass graft surgery, but not major complications. A multivariate model including all univariate predictors of LOS showed SMI, major complications, transapical access, atrial fibrillation, and chronic obstructive pulmonary syndrome as independent predictors of LOS. For every 14-cm2/m2 increase in SMI, there was a 1-day reduction in LOS. None of the standard measures of frailty predicted LOS.
Conclusions: Skeletal muscle index, a measure of sarcopenia readily determined from pre-TAVR CT scans, independently predicts TAVR LOS better than standard frailty testing. Further evaluation of SMI as a frailty measure after TAVR and other cardiovascular procedures is warranted.
Copyright Â© 2016 Elsevier Inc. All rights reserved.