Statistical analysis
The patients who died, or did not complete the follow-up period, were excluded from the study as the analysis of the changes in frailty status after surgery was not possible.
Continuous variables are expressed as median and interquartile interval (IQI) or mean and standard deviation (SD), depending on the normality of their distribution as assessed with Shapiro-Wilk normality test. Categorical data are presented as counts and percentages.
Differences among the three ordinal levels of frailty were analyzed considering the presence of a linear trend between categories. The Mantel Haenszel linear trend test was used for categorical variables, and linear regression models were employed for quantitative ones. Concordance between frailty scales was assessed with the Kappa index.
Chi-square test was used to compare the distribution of frailty groups before and after surgery. Changes in the proportion of patients who score a point in each of the components of the frailty phenotype was analyzed using the Mc Nemar’s test for paired proportions.
All the statistical analysis and graphics were performed using Stata/IC 15.0 Software (Statacorp LP, College Station, TX). A significance level p<0.05 was considered as statistically significative.