*2.4. Clinical Evaluation*

Patients were evaluated at a minimum follow-up of one year (59 at 1 year and 57 at 2 years). The clinical outcomes included periprosthetic joint infection (PJI), implant revision, and other intra- or postoperative complications. The PROMs included the American Shoulder and Elbow Surgeons (ASES) score [6], the Constant score [7], the Single Assessment Numeric Evaluation (SANE) score [8,9], and the pain on a visual analogue scale (VAS). The ASES score (from 0 worst to 100 best) comprises one pain item and 10 questions relative to patient function/disability. The Constant score (from 0 worst to 100 best) has four different dimensions, including pain, activity, strength, and mobility. The SANE score (from 0 worst to 100 best) was assessed with a single question: "How would you rate your affected shoulder today as a percentage of normal (0% to 100% scale with 100% being normal)" and the pain on VAS (from 0 best to 10 worst) was rescaled to a range of 0–100 points. The reference used for each PROM improvement was the substantial clinical benefit calculated by Simovitch et al. in a combined cohort of 1856 reverse and anatomic TSA (31.5 points for the ASES score, 19.1 points for the Constant score, 32 points for pain on VAS) [3]. Although being a useful metric, we did not include the SANE score improvement in the quality evaluation since its SCB has not been robustly validated in the scientific literature ye<sup>t</sup> [10]. Furthermore, the SANE score has been reported to be moderately/strongly correlated with the ASES score [11]. The authors also used the minimal clinically important differences (MCID) for the aforementioned scores for descriptive analyses (13.6 points for the ASES score, 5.7 points for the Constant score, 16 points for pain on VAS) [3].

#### *2.5. Equation for Patient Value*

The equation used for the calculation of patient value was based on the one previously published by Reilly et al. [2]:

$$\text{Patient value} = \frac{\text{Quality}}{\text{Cost}} = \frac{(\text{Weighted clinical outcomes} + \text{Weighted PROMs improvement})}{\text{Direct cost}}$$

The value equation can therefore be written as follows:

$$\text{Volume} = \frac{\text{W}\_{1}\left(\frac{\text{AP}\_{\text{RISES}}}{\text{SC}\_{\text{RISES}}}\right) + \text{W}\_{2}\left(\frac{\text{AP}\_{\text{CLoS-Tox}}}{\text{SC}\_{\text{CLoS-Tox}}}\right) + \text{W}\_{3}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{SC}\_{\text{DIS-Tox}}}\right) + \text{W}\_{4}\left(\text{P}\_{\text{Inttu-op.Comp}}\right) + \text{W}\_{5}\left(\text{P}\_{\text{P-ol.}-op.\\_\text{Comp}}\right) + \text{W}\_{6}\left(\text{P}\_{\text{P}/I}\right) + \text{W}\_{7}\left(\text{P}\_{\text{V-LoS-Tox}}\right)}{\left(\frac{\text{P}\_{\text{DIS-Tox}}}{\text{DR}\_{\text{CLoS-Tox}}}\right)} + \text{W}\_{7}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{8}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{9}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{10}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{2}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{5}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{D}/I}}\right) + \text{W}\_{10}\left(\frac{\text{AP}\_{\text{DIS-Tox}}}{\text{V}\_{\text{$$

As detailed in the article of Reilly et al. [2], all negative pre- to postoperative improvements were forced to 0, and clinical outcomes were coded as binary depending on the event occurrence. The absence of event resulted in a patient score equaling 0 for that outcome, while the presence of it resulted in a patient score equaling the total weight. The weighting for the clinical outcomes and PROMs was performed by the senior surgeon (AL) according to his strong clinical experience and scientific knowledge. For clarity, a quality of 1.0 indicated an improvement in PROMs which was equal to the defined SCBs and an absence of any complication, PJI or revision. A cost of 1.0 indicated a TSA that cost the exact direct cost reference (see below). The result of the equation (quality/cost) was rounded at the first decimal place and indicated a substantial delivered value if ≥ 1.0 or an unsubstantial delivered value if < 1.0.
