Chapter 4 40 between fracture and non-fracture related medical consumption, as well as for productivity losses. For the costs, LEFS and the EQ-5D there was no missing data. Therefore, exclusion or extrapolating in case of missing data was not required. A number of subgroup sensitivity analyses were performed. These analysis were conducted to assess differences between employed and unemployed patients and patients with and without complications. The EQ-5D scores were further used to estimate the QALYs. Base case analysis was conducted from a societal perspective, including patient & family costs and productivity loss costs. As costs data are generally skewed and not distributed normally, non-parametric bootstrap re-sampling techniques were performed to estimate cost uncertainties. The medical ethics committee of Maastricht University Medical Center+, Maastricht, the Netherlands approved this study and informed consent was obtained from all patients. Statistical methods Statistical analyses were performed with IBM SPSS Statistics, Version 25.0, Armonk, New York. For the non-parametric bootstrapping (1000 replications) of the costs, Excel for Mac V.16 was used to extract estimates of the mean and 95%-confidence intervals (CIs). Descriptive statistics were used to provide an overview of the demographic data and baseline characteristics for the entire study population. Independent samples ttests were used for normally distributed continuous data and chi-squared tests for categorical variables. Results are presented as either mean ± standard deviation (SD) or as frequencies and percentages. The level of statistical significance was set at ɲ=0.05. Results Baseline characteristics This cohort study included 53 surgically treated trauma patients with peri- and/or intraarticular fractures of the lower extremities (N=1 pelvic, N=3 acetabular, N=28 tibial plateau, N=12 pilon and N=9 calcaneal), 49.1% of which were male, with a median group age of 60.0 years (IQR 47.0-67.0). N=26 (49.1%) were still active employees. Baseline characteristics of the entire population are presented in Table 4.1. Patients with paid job were significantly more male (p=0.04), and significantly younger (p<0.01) compared to the patients without paid job. The ASA-classification and the Charlson score were both significantly lower in patients with paid job compared to the patient