*2.2. Analysis*

We produced a before (*n* = 100 ED patients) and after (N = 100 UCC patients) design, i.e., prior and after the UCC being established. Focus was on dimensions concerning the outcome of treatment at the ED in comparison to the UCC. Aspects studied were LOS, TTP, use of medical services (radiology, laboratory analysis), revisits, hospitalisations, mortality and cost. Descriptive statistics included central tendency and dispersion. LOS was defined as the time spent from admission to discharge from the UCC or the ED or admission to a ward. TTP was defined as the time spent from admission to seeing a physician. Paired-sample calculations of LOS, TTP and laboratory analysis showed a statistical di fference of means. However, the results of normality measured by the Kolmogorov Smirnov test did indicate LOS, TTP and laboratory analysis to be significantly skewed (LOS sig. 0.00, TTP sig. 0.00, and laboratory analysis sig. 0.00). The samples could consequently not be regarded as normally distributed due to outliers. A complementary non-parametric test was therefore conducted using a Wilcoxon signed-rank test.
