*2.7. Statistical Analyses*

For baseline characteristics, variables were reported as mean ± standard deviation or proportions. Shapiro–Wilk tests were used to assess the normality of distributions. Differences between preoperative and postoperative continuous values were evaluated using either the paired Student's t-test (if Gaussian distribution) or the Wilcoxon signedrank test (if non-Gaussian distribution). The correlation between the quality and cost was analyzed using the Pearson's coefficient. A multivariable linear regression model was performed to identify which pre-operative factors (Constant and ASES scores, VAS pain, primary diagnosis), patient characteristics (age, gender, BMI, arm dominance, and tobacco use), and intra-operative factors (patient managemen<sup>t</sup> time in the operating room, surgical procedure, surgical approach, cementation and use of patient specific instrumentation) were independently associated with patient delivered value.

The variables included in the multivariable regression model were identified using the backward selection method with a threshold of significance set at a *p* value < 0.05 (preoperative Constant score, pre-operative VAS pain, tobacco use, and surgical procedure). Statistical analyses were performed using R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria), and *p*-values < 0.05 were considered statistically significant.
