Personalized medicine in surgery represents a continuous search for improvement, reducing risks and human errors [21, 22]. As a result of the revolutionary changes in orthopedic and trauma diagnoses brought about by the discovery of X-rays in 1895, as of now, digital technologies and 3D printing are so widespread and well-liked that they are widely used in both trauma and arthroplasty [23, 24]. The main findings of this paper are that a 3D replica of a PHF can assist surgeons in diagnosing, planning, and performing ORIF, increase patient compliance, and improve residents’ satisfaction. Furthermore, surgical simulation can lead to a clear reduction in the duration of the operation.

Scientific attention toward the use of 3D-printed replicas has increased a lot in recent years, however the application in the daily routine of this technology for traumatic joint bone injuries is less widespread and common [7, 23]. This discrepancy may be due to difficulties in organizing the workflow of 3D printing organizations (emergency, radiologic, and orthopedic departments), as well as the availability of equipment and resources [18]. In our experience, having software and hardware already in place has allowed us to convert to an .stl file instantly and automatically after the CT scan: this approach allows the model to be available within a maximum of 11 h, providing time to use it to plan and perform an early surgery when necessary. Moreover, another strong point of this study is that we intended to concretely evaluate the feedback of this procedure both from the point of view of the patient and the surgeons. The results derived from our experience are encouraging: the randomization of the sample allowed us to compare the surgical times, the blood loss, and the quantity of intraoperative X-rays performed. In patients in group A, the lower number of X-ray controls performed and the significant reduction in the duration of the operation, even if not related to a significantly lower blood loss, suggest the advantage that can be derived from a more detailed observation of the fracture. This method allows us to fully know the comminution and dislocation of the fragments, to anticipate the potential bone defects, and therefore to predict the need for a bone graft, transforming a theoretical simulation of surgical planning from the virtual stage to the real stage [25].

At the same time, we did not expect a significant improvement in clinical outcomes in patients who underwent 3D preoperative planning. The substantial overlap of functional outcomes at a 2-year follow-up is likely due to the fact that surgery in each case was performed by an experienced surgeon. A recent review and meta-analysis of PHF treated with plate and screws shows that the final CSS of these patients is 75 ± 15.8 points [9]. In our case, the results of both groups intersected the standard deviation of the aforementioned meta-analysis (group A: CSS = 70 ± 11.9; group B: CSS = 72 ± 5.7), without observing any particular differences.

It is always difficult for patients and their families to fully understand the severity of the fracture and expected outcomes. None of the 20 patients had ever seen a 3D-printed model of any part of their body previously, as evidenced by the answers to our questionnaire. Although we have not been involved in any legal disputes, which is also a result of the absence of complications in this set of patients, we are confident that doctor–patient and family practice communication can become a useful tool for boosting understanding and compliance both before surgery and during recovery. We think the ability to amend the informed consent wording after giving the patient access to a natural-size replica of their fracture and ensuring that the proposed surgery is clearly understood should also lessen the likelihood of a complaint stemming from clinical dissatisfaction.

In general, the interpretation of imaging data is now actually a process of integrating 2D and 3D images, but in complex fractures of the proximal humerus, with difficult-to-understand spatial structures, the analysis and complete determination of the fracture pathological anatomy are more difficult [8, 11, 26]. The occurrence of iatrogenic complications is not only associated with surgical skill and mastery of theoretical knowledge, but also related to preoperative diagnosis and surgical planning [1]. This is especially true for junior surgeons and trainees, where hands-on practical experience goes a long way in successfully performing high-demand surgeries [27, 28]. In addition, a full-touch fracture model appears to be essential to facilitate communication among surgical team members, thus improving performance and collaboration. In particular, the senior surgeons who answered question no. 4 commented that the choice of plate’s holes and the direction of the screws was influenced by the in vitro surgery. Moreover, they greatly appreciated the possibility of handling the sterilized model during surgery, substantially since it will allow them to avoid any inconveniences or diversions that might arise during the interpretation of X-ray and CT images. In answer to question no. 5, they suggested implementing the use of 3D models for other complex and comminuted fractures, advocating not only greater understanding of fractures but also ease of surgical planning.

Also from a didactic point of view this method may represent a useful option to study anatomy regarding the review of surgical techniques and fractures [28]. In particular, the residents were allowed to practice, improving their skills and learning curve, and they aimed for more widespread utilization of modern and innovative tools such as this one, to approach major surgery with more confidence. In place of the more expensive cadaver lab activities, the simulation of the surgical act using 3D printing, virtual, augmented, and mixed reality may be a good alternative [29]. These findings, along with the opinions of the residents, have led us to incorporate this technology into our university’s residency program, as has already been done in other structures [18, 30]. Basically, owning an institutional 3D printer can prove to be an important training tool during residency programs [27, 28, 30].

However, the cost-effectiveness analysis is what determines whether this technology can be developed and is even feasible. A 3D-printed model adds to the already mandatory costs of X-ray or CT examinations. One of the most innovative approaches to measure expenses more accurately and address cost challenges is time-driven activity-based costing. However, this sophisticated instrument is equally suitable in the orthopedic field, at least in the trauma field, owing to the poor reproducibility of surgery [19, 31]. The calculation of the capacity cost rate (CCR), which is the practical capacity of each active operator providing procedures, is not suitable for short interventions and with healthcare workers on fixed salaries [15]. For this reason, we used the more traditional tool of comparing direct and indirect costs, excluding from the result the mandatory expenses incurred by all patients (such as hospitalization, tests, and overhead expenses). In our experience, the sterilization of the devices has not resulted in any additional direct expense, as well as the software used which is free online for noncommercial purposes. Likewise, the hardware (laptop, drivers, and printer) were already owned by our Institute. The direct cost for the reproduction of the “fracture model” and the “reduction model” was EUR 5 each (EUR 10 per patient). The size of the pieces, the type of resin used for printing, and the details of the prototypes can directly affect the final cost. The choice of PLA as the first material for 3D model reconstruction mainly depends on its characteristics. It is inexpensive and represents a valid alternative to petroleum-based materials. It has a low fusion point, and compared with other materials it is more moldable and its processing is less toxic and harmful than other cheap materials (Spectrum Group Ltd.—Pęcice, Polska) [1]. Regarding indirect expenditures (active operating room), we observed a definite reduction in operation length (15%) in patients in group A, resulting in a savings of more than EUR 400 per procedure. Despite these encouraging results, it can take some time for health administrations to approve it in daily routine because it is new and not yet acknowledged as a diagnostic tool for understanding fractures in general. In the future, efficiency in terms of time and costs can be anticipated when taking into account a scenario of model production inside a preoperative procedure inserted in routine practice [18].

One of the limitations of this method, however, is the procedures for printing, segmenting, and immobilizing the 3D model, which can be long, tedious, and complex [32]. The quality of CT images, which may produce cuts as small as 0.9 mm [18], is directly related to the precision of 3D printers. Extruded filament used in modern 3D printers has a thickness of roughly 0.1 mm. In truth, the spatial resolution that can be achieved seldom exceeds this limit and is often around 0.5 mm because of the vibrations that are caused by cartridges. Another limitation of surgical simulation compared with real surgery is represented by the absence of the muscular forces, soft tissue, nerves, and veins, which in the 3D model facilitates the implantation of the plate and screw. Furthermore, the sample examined is certainly too small to reach definitive conclusions on the effectiveness of this tool. Finally, even the cost analysis may not be entirely accurate. The document taken as a reference to calculate the savings on indirect costs is from 2009: given inflation and the general rise in healthcare expenditures, the savings might be even higher.

Finally, even the cost analysis may not be entirely accurate. The most recent document for calculating indirect cost savings is from 2009 and does not consider inflation. Even if the savings highlighted by our experience may seem minimal, and the few minutes saved in the duration of the operation can scarcely guarantee the immediate use of the theater for a new operation, the “lack of production” of the healthcare personnel must still be considered: in fact, since in many countries hospitals are considered “commercial companies,” having staff on duty who are not producing leads to economic losses for the administration. However, one of the main limitations of the capacity cost rate is that it must be applied singly in each institution as salaries are potentially different, which is why we did not explore expenses with this tool. Ultimately, another limitation is that we did not consider the time dedicated to planning in the expense account for both groups. This is because surgical simulation, especially on 3D models, must be strongly considered as education and training, and the human resources involved have subjective qualities of reproducibility, skill, and understanding.