Hemorrhage from battlefield trauma is a vital cause of early death and the leading cause of preventable death. At present, the most commonly used animal models in trauma hemostasis experiments are pig femoral artery or carotid artery injury hemorrhage models [7, 8], and injury hemorrhage models caused by military weapons are rare. However, gunshot wounds tend to have a cavity area larger than the outlet, rather than being linear, and such complex deep wounds with massive bleeding are often difficult to stop [9, 10]. To address these limitations of existing models, we had successfully established the inguinal bullet wound hemorrhagic model to test the hemostatic effect of the CSD in our previous study [6].

In this study, we further adjusted the method for establishing the groin bullet wound hemorrhagic model by accounting for the guidelines of tactical combat injury treatment. First, we calculated the appropriate sample size through the statistical software and expanded the sample in comparison with the previous study. Second, to simulate the limited first aid time in an actual battlefield, the treatment was started 30 s after the successful establishment. Many other bleeding models also use 30 s as the initial time point [11, 12]. The MCSD and SG groups showed no significant differences in the changes of vital signs within 30 s. Wounds caused by bullets are highly uncontrollable, and free bleeding period of 30 s before hemostasis could not only simulate the battlefield treatment environment, but also further ensure the uniformity of blood loss between the two groups. Thirdly, a clear positive correlation has been observed between the mortality of those who are hit and the duration of first aid [13]. Therefore, we developed a timely care and evaluation protocol based on the “MARCH” strategy. Fourth, to simulate the longer shooting distance in the battlefield, we adjusted the shooting distance to 50 m and used a high-precision sniper rifle for shooting. A color Doppler ultrasound-based marking was also used to improve the accuracy. In addition, we excluded Bama miniature pigs in which the femoral artery was not injured, and we did not attempt shooting again to avoid influencing the data results. Lastly, the hemostatic methods in this experiment showed several differences: (1) The TCCC guidelines recommend that the gauze should be pressurized for at least 3 min to stop bleeding. To simulate the compressive effect of PVA sponge in deep tissues, we maintained an additional 20 kPa pressure while the gauze was packed into the wound [14]. (2) Since the hemostatic material will pull the tissue and induce bleeding if taken out again for observation, we only evaluated the success of hemostasis once in this experiment. Continued bleeding even after the PVA sponge or gauze filled the trajectory was considered to indicate hemostasis failure.

Penetrating wounds and blind wounds caused by bullets or explosive fragments often have narrow entrances and irregular shapes, making it difficult for gauze and dressings to fill wounds and effectively stop bleeding [15]. The CoTCCC recommends three hemostatic devices for the treatment of junctional bleeding: Combat Ready Clamp (CRoC), Attached Emergency Treatment Tool (JETT), and SAM Attached Tourniquet (SJT) and XStat. However, CRoC is bulky and heavy and takes a long time to assemble [16]. On the other hand, both SJT and JETT show low success rates for hemostasis, which requires training to improve [17]. The XStat™ device (RevMedx, Inc.) is a new type of hemostatic device with 92 hemostatic sponges containing chitosan. These sponges expand rapidly after being pushed into the wound, filling the deep and incompressible wound to stop the bleeding [18]. However, XStat™ is composed of many individual micro-sponges, necessitating more time to remove these sponges from the wound. In some cases, complete removal of the sponges necessitated wound enlargement, which further aggravated the wound [19].

Therefore, we independently developed a CSD by combining the characteristics of the PVA sponge [5]. The principle of hemostasis was similar to XStat [20]. However, through practice, we found that the original device could not completely inject the sponge into deep wounds due to the short injection tube. Thus, we improved the device by incorporating a reverse push-type syringe structure to solve this problem. This study showed that the MCSD offered the following advantages.

We found that the use of the MCSD was associated with less post-treatment blood loss, less hemostasis time, more stable vital signs like MAP, more stable hemoglobin level,and a higher success rate than SG. In addition to the rapid hemostatic effect of the PVA sponge in MCSD, the unique structural characteristics of MCSD also played a huge role in conferring these advantages. The MCSD adopts a reverse push-type syringe structure to ensure that the series of PVA sponges can enter the deep wound smoothly. In comparison with gauze, the MCSD, which has a diameter of only 10 mm, id more likely to enter narrow and complex wounds caused by bullets or explosive fragments.

Lander et al. showed that limb movement after the application of hemostatic materials can lead to rebleeding [21]. Although the same experiment was not performed for this study, the MCSD group showed no further bleeding after successful hemostasis despite the process of handling. We considered that this could be attributed to the close compression of the PVA sponge after expanding according to the CT scan results. Moreover, the shape advantages of the prism after expansion made it easier for the dressing to remain stuck and not fall off.

A retrospective study by Warina et al. reported retention of hemostatic material in the wound when XStat was used to treat penetrating wounds [19]. However, no retention was observed in the MCSD group in the present study because of the chain structure of MCSD. Moreover, no significant difference in the removal time was observed between the MCSD and gauze groups. We considered that this may reduce the possibility of wound enlargement by residual material and obviate the need for a secondary extraction.

Medical-grade PVA sponge is widely used in vacuum suction dressings, implantable medical devices, and other medical fields. In this study, TEM and HE staining results showed that the application of MCSD did not lead to significant necrosis of the femoral artery and muscle tissue.

The MCSD can be used single-handed at any time to meet the needs of soldiers for self-rescue and mutual rescue. The compressed PVA sponge can absorb blood and expand, exerting pressure on the surrounding area. We did not even pressurize the wound after injecting the PVA sponge in the experiment. Moreover, the MCSD is sufficiently portable, with a single MCSD weighing only 15 g.

In summary, these findings indicate that the MCSD meets the requirements of war trauma emergency hemostatic dressings, showing a good hemostatic effect, short hemostasis time, good safety performance, and short removal time [22].

Nevertheless, this study also had some limitations. First, although the device is designed to control bleeding at all junctional sites, due to the limitations imposed by the experimental conditions, only the groin region was tested. Further experiments are needed to evaluate hemostasis in different anatomical locations such as the buttock, armpit, and neck. Next, Wang et al. chose to shoot the pistol at close range 3 m away and missed the femoral artery in 36 animals (success rate, 83%) [13]. Similar to the anatomy of humans, the femoral veins and arteries are next to each other in Bama miniature pigs. Therefore, in bleeding models based on femoral artery injury, injury to the femoral vein during establishment of the groin bullet wound model is an unavoidable problem.

Kheirabadi et al. applied a tourniquet in a porcine femoral artery injury bleeding model, and repaired the injured femoral artery and sutured it layer-by-layer to observe complications for 1 week after hemostasis [23]. In the present experiment, the Bama miniature pig model with a bullet penetrating the artery in the inguinal region was more difficult to repair than a simple femoral artery injury model. The bullet also caused comminuted fractures of the pelvis and femur, which affected subsequent functional recovery. However, since the observation time for this study was limited to 3 h, no further damage control surgery was performed.

Furthermore, the XStat, junction tourniquet, etc. have been applied to patients in pre-hospital trauma and bleeding, highlighting the need to apply the MCSD in clinical practice. Lastly, study showed that use of local pro-coagulant hemostatic agents could improve the hemostatic effect [24]. In future studies, we aim to further improve the MCSD by designing different types of injection tubes and evaluate the improved MCSDs in animal explosion models and different anatomical locations such as the buttock, armpit, and neck.