Introduction

Fractures of the proximal femur affect ~ 3.5% of men and 14.6% of women throughout their adult life [1]. Its incidence increases year by year due to the progressive aging of the world’s population, with a projected 100% increase in prevalence in the first half of the 21st century [2,3,4,5].

The primary treatment of hip fractures is surgery, with conservative treatment reserved for those patients at high surgical risk due to their comorbidities or for nondisplaced intracapsular fractures in selected patients [3, 6,7,8,9,10]. Of the surgical implant options available, sliding hip screws and intramedullary nails are the two most frequently used options [9, 11, 12], with the use of cannulated screws being much less frequent [13].

Vascular injury is one of the least frequent and most severe surgical complications, with a prevalence of 0.2–0.49% depending on the series [14]. When it occurs, it increases direct mortality by 6% and favors the appearance of new comorbidities in about 11% of patients [15]. In 42% of the cases, the deep femoral artery was the injured artery, and also the most frequently damaged vessel, followed by its perforating branches, which occurs in 22% of cases [15]. The intrapelvic arteries are traumatized less frequently (8.24% vs. 91.2% of the extrapelvic arteries), with the external iliac artery being the most frequently injured, only 2% of the time.

Pseudoaneurysm is the result of 67% of iatrogenic vascular lesions and is the most important in terms of frequency, followed by laceration with hemorrhage, which occurs 23% of the time [15, 16]. Thus, the symptoms are often not easily identifiable, thereby delaying diagnosis and, in some cases, may even go so far as a hospital discharge without it [16].

Between 30 and 60% of iatrogenic vascular lesions could easily be prevented [15, 16]. Although the traction of the minor trochanter fragment is a cause to consider [17], the drilling of the second cortex is responsible for about 20% of injuries, especially during the placement of the third and fourth nail plate screws and the short nail locking screw.

For many years, instrument kits have carried brand-specific devices designed to control the drilling of the second cortex. Because vascular injury is rare, surgeons often dispense with its use, which usually leads to its disappearance due to loss or its withdrawal due to disuse. This loss also occurs in university hospitals or hospitals with resident surgeons, where due to the infrequent use by experienced surgeons, younger surgeons do not have these aids, which implies an increased risk of vascular injury.

The objective of this article is to describe and evaluate an easily accessible device of universal use that is effective in controlling excessive drilling of the second cortex.

Materials and methods

Design

The design created by the bioengineering team of our hospital can be printed with any domestic 3D FFF (fused filament fabrication) printer and is explained in Fig. 1.

Fig. 1
figure 1

Description of the device. 1.1: screw-on-nut system with a lateral slit that allows drill insertion at any time without removing the motor. 1.2: during drill insertion, the screw rotates on the nut, which, on the one hand, firmly traps the drill and, on the other hand, modifies the length of the device to suit each patient. 1.3: an intraoperative image of the device during a short nail bolt placement. 1.4: detail of how once the first cortex has been drilled anddrill rests on the second cortex (image 1.6 is correlative to moment 1.4), the nut is turned to adjust its length ~1 cm from the end of the motor. In this way, the drill travels further until it passes the second cortex (image 1.5), thereby avoiding excess drilling when coming in contact with the device (image 1.6 and 1.7). The device was printed with medical poly lactic acid (Smartfil Medical, USP VI/ISO 10993-1 certified) and sterilized with a conventional 134° autoclave program. The parameters for the 3D impression are described in Table 1

Full size image
Table 1 Description of the 3D-printing parameters used to print the device
Full size table

After creating the design (3D Builder, Microsoft Corporation, Washington, USA), it was then converted into a conventional three-dimensional.stl format (Meshmixer® version 3.5.474) and configured for 3D printing in.gcode (Cura Software® version 4.3.0). Both documents were shared electronically for direct download via a web link.

For the evaluation of the device, a bi-cortical bone drill simulation system was designed, which is presented and explained in Fig. 2.

Fig. 2
figure 2

Description of the evaluation device. 2.1: in yellow, an overview of a simulated bi-cortical element with a thickness similar to the bone cortex. 2.2: complete lateral view in which the simulation of the two cortexes separated by a space and below the second cortex, represented in red a deformable plastic element in which excessive drilling is recorded. 2.3: in blue, the drill resting on the second cortex is represented. 2.4: the hole in the plastic device. Once drilling is carried out, drilling depth is measured using a screw gauge

Full size image

Device evaluation

Cross-sectional study

Ten subjects were obtained by randomization using a table of random numbers, from a group of 20 trauma residents with < 10 surgeries per year as primary surgeon. Likewise, 10 subjects were obtained by randomization using a table of random numbers from a group of 20 expert surgeons with more than 50 surgeries per year as primary surgeon.

Each subject was asked to perform a bi-cortical drilling (Fig. 2) in triplicate, and in each attempt, the excessive drilling was evaluated to assess the learning curve. Subsequently, the same procedure was carried out, also in triplicate, using the evaluated protection device to study the learning curve of its use.

The excessive drilling in each attempt was evaluated, and differences in drilling lengths between expert surgeons without the device and resident surgeons with the device were defined as the main variable. Meanwhile, the differences between drilling lengths of expert surgeons with and without the device were defined as secondary data and were evaluated to analyze its usefulness in this group, and the same was carried with the resident. Finally, the learning curve of the use of the device was evaluated in both expert surgeons and resident surgeons (Fig. 3).

Fig. 3
figure 3

Schematic description of the materials and methods. Asterisk: differences were statistically significant. In green, the drilling attempts using the device. In red, the drilling attempts without using the device

Full size image

To evaluate the excessive drilling depth, a standard depth gauge for screws was used.

Statistics

The normal behavior of the data was analyzed (Shapiro–Wilk) to define the test used during the hypothesis testing for both the paired data (Student’s t test for paired data vs. Wilcoxon test) and independent data (Student’s t test vs. U of Mann–Whitney). In cases of comparison, multiple quantitative data between two groups were chosen through a variance analysis or Friedman’s test, based on the normal behavior of their data.

The statistics was carried out by an independent researcher with R statistics version 3.5.2.

Results

The expert surgeons drilling without device mean was 3.03 (SD 1.69 95% CI 2.40–3.66 p value = 9.89e−11) while that of residents with the device was 1.76 mm (SD 0.89 95% CI 1.43–2.10 p value = 1.169e−11), meaning that residents with the device performed drill lengths 1.26 mm smaller than those expert surgeons without the device, with this difference being statistically significant (95% CI 0.56–1.97, p value = 0.0007) (Fig. 4, 4.1).

Fig. 4
figure 4

Box diagram of the primary data according to the results. Asterisk: hypothesis test in which the differences were statistically significant

Full size image

The resident surgeon drilling without the device mean was 6.93 mm (SD 2.44 95% CI 6.01–7.84 p value = 1.415e−15), which means a higher drilling length of 5.17 mm compared to drilling with the device, with this increase being statistically significant (95% CI 4.2–6.1 p value = 5.322e−13) (Fig. 4, 4.2).

In the expert surgeon group, the drilling with the device mean was 1.86 mm (SD 0.89 CI 95% 1.53–2.20 p = 3.349e−12 value), this value being 1.17 mm lower than the mean value without the device, which implies a statistically significant improvement in drilling length (95% CI 0.46–1.87 p value = 0.001) (Fig. 4, 4.3).

Residents without the device showed higher drilling lengths than the experts without it, with a mean of 3.09 mm higher, this difference being statistically significant (95% CI 2.8–4.99) (Fig. 4, 4.4). With the use of the device, the differences between experts and residents were 0.1 mm; this difference was not significant [95% CI (−0.36) to (0.56) p value = 0.66] (Fig. 4, 4.5).

There was no difference in drilling length with the aid of the device between the first, second, and third drillings, in both expert surgeons (mean drilling p value = 0.62) and in resident surgeons (p value = 0.22) (Fig. 5, 5.1, 5.2). In drilling without the device, the expert surgeons did not show an improvement between the three attempts, with lengths remaining similar [3.8 (SD 2.29); 2.7 (SD 1.3); 2.6 (SD1.07) for the first, second and third attempts, respectively; p value = 0.21] (Fig. 5, 5.3). Meanwhile, the residents showed improvement in the drilling lengths from the first to the second attempt and did not improve between the second and the third attempt (Fig. 5, 5.4) [p value = 0.03; 8.5 (SD 3.02); 6.3 (SD 2.05); 6.0 (SD 1.33); for first and second attempts, respectively].

Fig. 5
figure 5

Diagram of drilling averages in the learning curve of the device for residents (5.1) and expert surgeons (5.2). 5.3 and 5.4 show the learning curve to avoid excessive drilling with the device in experts and residents, respectively. It should be noted that the first drillings were carried out without the device. This was done to facilitate overcoming the learning curve related to the evaluation device and subsequently analyze the learning curve of the protection device specifically

Full size image

Discussion

The validation data we present obtained from the device show a reduction in the risk of vascular injury in trauma surgery. For less expert surgeons, it allows them to develop their learning curve without the risks of vascular injury exceeding those that an expert surgeon may have during their learning process. The device prevents excessive drilling without reducing the sensation perceived when the second cortex is passed, thereby reducing risks without interfering in the learning process.

In the case of expert surgeons, the device reduces the drilling distance compared to manual drilling. In situations where the risk of injury to vascular and nervous structures is high, this additional safety contribution may be useful, for example, in clavicle fractures to avoid injury to the axillary artery or in tibial plateau fractures to avoid injury to the popliteal artery [18,19,20].

The device has a low learning curve, and the study data show no difference in the results obtained since the first use, making it easy to use. At the same time, thanks to its design, it is easy to obtain thanks to 3D printing. 3D printing has brought about a revolution in orthopedic surgery and traumatology, especially in the field of preoperative planning [21,22,23]. This article is pioneer in describing a new 3D-printing application, which is sharing devices among orthopedic surgeons to find solutions to common problems.

Due to its design, the device can be obtained anywhere in the world with just a click, costs less than surgical gloves, or a scalpel, and can be shared and modified as if it were a conventional text document.

The main limitations of the study are having been carried out in a single center with only 20 participants and using the device designed by our bioengineers for the drilling. Furthermore, we must emphasize that the sample size was calculated before the start of the study and that the central hypothesis showed a statistically significant difference, which implies that a type II error was not made. On the other hand, the evaluation device (Fig. 2) showed similar behavior in terms of excessive drilling to that which has been demonstrated by other studies with elements validated for this purpose [18]. Besides, the possible advantages or disadvantages associated with this evaluation device were the same for both groups, for which the possible existing bias was distributed equally and should not have affected the results.

In conclusion, this device is one example of a new way of sharing surgical solutions among traumatologists through the use of 3D printing and is the first validation study of a device with these characteristics. Due to its simplicity, ease of use, accessibility, necessity, practicality, and efficacy, it is a device that can help resident surgeons develop their learning curve without putting patients at risk and can also help expert surgeons achieve more precision in those cases or situations with a high risk of iatrogenesis.