5.1 Linear Mixed Models
To assess the HRQoL outcome we performed linear mixed models (LMM) analyses on our data.
The course in time was adjusted for the irregularly spaced postoperative surveys and squared if this resulted in a better model of fit. To determine best model of fit -2 Log Likelihood (-2LL), Akaike’s Information Criterion (AIC) and Schwarz’s Bayesian Criterion (BIC) were used.
For the analysis on predictors the variables sex, age, Krouse staging, ASA, smoker, preoperative EES-Q score, type of surgeryand postoperative antibiotics were included as fixed effects. The variable age was mean-centered.
Covariance type was set at unstructured and calculation was based on repeated measures . This was also determined by using -2LL, AIC and BIC. Estimation method was set to maximum likelihood . Coefficient, standard error, 95% confidence interval and p-value are displayed.
All statistical analyses were performed with IBM SPSS Statistics version 22.0 (SPSS IBM, Inc., Armonk NY, USA).