*4.4. Study Limitations*

The current study has some limitations. First, although the German population forecasts provided rather good agreemen<sup>t</sup> between predicted numbers and confirmed past numbers [2,20], they are typically based on hypothetical future assumptions and are, therefore, potentially uncertain. Second, this type of study was based on the procedural growth trajectory for the years 2010–2019, as a longer time frame was not possible due to coding changes after 2009 and the occurrence of the COVID pandemic after 2019, assuming that this trend will continue. However, it is possible, that after the world has overcome all issues regarding the pandemic, new technologies, surgical innovations, or disruptive innovation techniques, such as cartilage regeneration, tissue engineering, or drug therapies that limit the progression of osteoarthritis or enhance fracture recovery or joint-preserving surgery [41]; as well as health-care reforms or reimbursement changes, will meaningfully affect the relationship between supply and demand for shoulder arthroplasty. Especially, it remains to be seen, if the demand of rTSA for proximal humerus fracture will actually see the projected rise in the future, as more and more studies sugges<sup>t</sup> that non-operative treatment may be more favorable in the elderly compared to a surgical approach [37,38,42].

Overall, in contrast to clinical studies, like randomized controlled trials, which draw evidence from (allegedly) artificially generated data, large database studies, like the presented study, describe results of the actual treatment reality. The current study aims to analyze the current practice and its changes over time, giving the reader an unselected view on surgical demand, which cannot be obtained from case series or prospective studies. Therefore, it has to be seen as an important additive or complementary instrument to clinical studies, providing the possibility to examine the potential effects of current scientific evidence on the daily surgical routine on the one side and to analyze treatment trends, whose causal links can be further analyzed by randomized-controlled studies, on the other side.
