In our series, young males predominate, similar to what has been reported in the literature. In contrast, trauma characteristics, treatment, and results had some differences. The more prevalent wound mechanism was PCI-SW (93%); at admission, a total of 106 patients (44%) had a normal hemodynamic status and 161 (67%) had cardiac tamponade. The high percentage of normal hemodynamic status patients partially explains the low overall mortality in this series (14.6%) compared to what has been reported in the literature [4–6, 18–24]; nonetheless, it is similar to the reports done by Villegas (10.4%) and Duque (13%) in Colombia [25, 26]. Mortality for the PCI-SW cases was 11.7% while in other series it ranges between 13 and 78% [2, 5, 10, 23–25]. For the PCI-GSW, mortality was 41.2%, similar to what is reported in series where type of injury predominate (26–85%) [2, 5, 10, 21, 23–25].

When analyzing the differences in mortality between wound mechanism, we could not identify significant differences in the prevalence for Grades II and III injury for PCI-SW and PCI-GSW cases (46.6 and 41.2% respectively); thus, we inferred that the higher mortality in PCI-GSW was due to not receiving early hospital care rather than injury grade. Another possible explanation for the high mortality is that most of PCI-GSW cases had associated lesions: thoracoabdominal (52.9%), chest (18%), and abdominal (1 case). This observation is consistent with other experiences like those from Asensio et al. [10], Buckman et al. [27], and others who have reported that associated lesions were a poor prognosis factor for PCI patients [28–30]. It is important to remark that in our center a deep analysis was not done to stablish the final cause of death (if heart wound or associated lesions) as this process in Bogotá must be done by specialized centers in forensic medicine.

The degree of hemodynamic compromise in this series also failed to explain the difference in mortality given that 78% of the PCI-SW and 76% of the PCI-GSW were admitted with a hemodynamic status ranging from normal to profound shock. This could be explained probably due to the high prehospital mortality rate in the latter group.

Although the proportion of patients with cardiac tamponade was similar for both types of injuries, it should be noted that when there was cardiac tamponade mortality was higher than the overall mortality (20.5 vs. 14.6%). Some authors consider cardiac tamponade to be a protective factor for patient’s survival [16]. Our results showed that the 2 of the 33 patients that died and had cardiac tamponade also had a normal hemodynamic status; moreover, 66% (n = 22 of 33) had some associated injury in the thorax or abdomen. It is important to point out that, in terms of physiopathology, filling of the pericardium space limits the stroke volume so in response cardiac frequency and right heart filling pressures are elevated through catecholamine production until right heart’s distensibility limit is reached, septum is pushed to the left side, and left side’s function is finally compromised [1]. For this reason, the longer the decompression of the pericardial space is delayed, a poorer prognosis should be expected which may explain our findings. As far as we know, the exact point when tamponade becomes a detrimental factor has not been stablished [27].

This series is unique in that a pericardial window was performed as the diagnostic procedure in 62.3% of the patients. As explained before, no echocardiographic assessment was done in this series, thus the protocol was to perform diagnostic pericardial windows in patients admitted with a PCI and a normal hemodynamic status. The group’s experience with this procedure was satisfactory, and the results were analyzed and published [31]. In none of our cases was the pericardial sac washed to define the management as some groups propose [32], which in hindsight could have prevented the thoracotomy done in the 22 cases (8.4% of the total series) excluded from the analysis for having only the pericardium compromised.

The frequent use of pericardial window partly explains our group’s preference for using sternotomy as the surgical approach and the good results obtained, similar to other series [29, 33]. In this series, a left anterolateral thoracotomy was generally used for resuscitation in patients with evident cardiac tamponade or in extremis. We concur that (i) the advantages of a sternotomy are better access to the right chambers and right pulmonary hilum, allowing cannulation for a cardiopulmonar bypass and (ii) that the left anterolateral thoracotomy facilitates access to posterior structures such as the esophagus, descending aorta, or left hilum [3]. Nevertheless, the choice of surgical approach also depends on the surgeon’s experience, the expected injuries according to the probable trajectory of the wounds, and the evidence of associated lesions.

Different physiological indexes and trauma mechanism have been proposed as predictors of mortality. Among the factors that affect survival rate after a PCI are the type of weapon used, the size of myocardial injury, cardiac injured chamber, compromise of coronary arteries, initial hemodynamic status, associated lesions, and the time elapsed until reaching the hospital, factors with which our series coincide. The overall most common wound location was the ventricles (n = 204), and compromise of the right and left ventricle was observed in 53 and 32% respectively, with a 10–15% mortality. In contrast, of the 18 injuries to the atria, 7 occurred in the right atrium with 71% mortality and 11 in the left atrium with 5 deaths (45.5%). It was surprising that some Grade IV and V patients were admitted with a normal hemodynamic status. Despite an overall survival rate of about 85%, it is not possible to address the “benevolence” of the cardiac injuries as this and other series are based on analyses of patients that received medical attention and thus represent the cases with better prognosis.

The average time between the moment of injury and execution of the surgical procedure was 60 min, and in 73.6% of the cases, this time was less than 120 min. It can be inferred that a large number of victims die before reaching the hospital. In a previous work by Pedraza and Isaza et al. (found in the Universidad del Rosario repository under the title Caracterización de la mortalidad por trauma cardíaco penetrante en Bogotá), 127 autopsies from individuals with heart injuries were performed finding that 71% died within the first hour and that only 51% received some type of medical attention.

The clinical follow-up was done until they were discharged from the hospital. The population’s socioeconomic conditions, the irregular availability of echocardiography, and the poor clinical ambulatory follow-up made it impossible for us to gather information on any residual intracardiac injuries; however, other studies have found a prevalence of valvular incompetence or septal ruptures in 19% of the cases [25, 29]. The hospital currently has stablished protocols for the clinical follow-up with electrocardiogram and strict imaging in PCI patients.

This study has limitations given that (i) it is a series of retrospective cases conducted in just one institution, where the majority of the patients come from a location with one of the highest indices of violence in Bogotá, which gives an atypical demographic profile; and (ii) the design does not permit a more in-depth statistical analysis. Nevertheless, the results permit us to generate new working hypotheses such as the need for revising the classification system of these injuries and the need of analyzing mortality from a forensic perspective.