Impact of Body Weight on Outcomes of Patients Undergoing Coronary Artery Bypass Grafting

Overview

All consecutive patients who underwent isolated coronary artery bypass grafting (CABG) at our institution from 2014 to 2020 were included. Patients were divided into 6 groups according to body mass index (BMI): underweight (BMI \<20.0 kg/m2), normal weight (BMI 20.0-24.9 kg/m2), overweight (BMI, 25-29.9 kg/m2), obesity class I (BMI 30-34.9 kg/m2), obesity class II (BMI 35-39.9 kg/m2) and obesity class III (BMI \>40 kg/m2). The long-term mortality was analyzed as primary end-point. The univariable and multivariable analysis was performed using logistic regression modeling.

All consecutive patients aged above eighteen who underwent isolated CABG were enrolled. Patients operated on- and off-pump were included. We reviewed elective, urgent, emergency and salvage surgeries. Patients with any additional concomitant surgical procedures were excluded. Patients were divided into 6 groups according to body mass index (BMI): underweight (BMI \<20.0 kg/m2), normal weight (BMI 20.0-24.9 kg/m2), overweight (BMI, 25-29.9 kg/m2), obesity class I (BMI 30-34.9 kg/m2), obesity class II (BMI 35-39.9 kg/m2) and obesity class III (BMI \>40 kg/m2). The primary endpoint was long-term mortality. The secondary end-point was the rate of sternal wound infections. Summary statistics were calculated. Continuous data are described as the median, with the interquartile range in parentheses, while categorical data are shown as numbers with percentages. Differences between groups for normally distributed, continuous variables were determined using the one-way analysis of variance with Holm-Sidak post-hoc test. The Kruskal Wallis test was used for the analysis of non-normally distributed continuous variables with Dunn's test for post hoc comparison. The chi-squared test was used for the analysis of categorical variables. Bonferoni's correction was used for post-hoc analysis. The univariable analysis was performed using logistic regression. For multivariable analysis a logistic regression model was performed, with mortality or wound infection as the dependent variable, and all other characteristics summarized in the table 1 as independent variables. The backward conditional selection was used for modeling with variables with score statistics \<0.1 included in the model. Significance was assessed at p\<0.05. Multiple imputation was not used to adjust for missing data. The number of valid entries for each variable and alterations in the sample size for specific analyses are described. Analyses were performed using IBM SPSS Version 22.0 and MedCalc Version 19.4.1.