An Analysis of Traumatic and Non-Traumatic Hemopericardium Using Autopsy Data From a County Medical Examiner

The volume of pericardial blood that overwhelms compensatory homeostatic mechanisms and causes tamponade physiology is not well understood in trauma patients. Based on animal models, it is thought to be <150 mL. In previous human studies, estimates have ranged from 50 to 150 mL. ¹ Several autopsy studies for hemopericardium have been performed, but they did not analyze the volume of hemopericardium (HP). ² Other large studies from trauma registries have focused on the mechanism of injury and outcomes. ³

Given the uncertainty of the volume of HP in the literature, we decided to perform an analysis of autopsy data comparing patients who died with traumatic hemopericardium (>20 mL) to a control group of patients who died from hemopericardium and cardiac tamponade secondary to ruptured aortic dissection or ruptured myocardial infarction. Demographic data, mechanism of death, the primary cause of death, and volume of pericardial blood at autopsy were collected.

The institutional review boards at The University of Texas Health Science Center at Houston and Baylor College of Medicine (Houston, TX) approved this study. Data acquisition was within the jurisdiction of the Harris Institute of Forensic Sciences, which includes the Medical Examiner’s Office. Harris County, TX is the third-most populous county in the United States (exceeding 4.4 million) and includes the Greater Houston area. Analyses were conducted using IBM SPSS Statistics v2213. Percentages were calculated for sex, race, death mechanism, and anatomical locations. Mean and medium, along with a measure of variability (standard deviation and interquartile range), were calculated for age, blood volume, weight, and height. Univariate analyses for discrete variables were performed using chi-squared tests, and Kruskal-Wallis tests were used for continuous variables. For all tests, statistical significance was set at the level of α ≤ .05.

Of 1848 traumatic deaths during the 1-year study period, 54 had HP on autopsy, but HP was not judged to be the primary cause of death for any of these patients. A comparison group of 79 non-trauma patients who died from cardiac tamponade (CT) was also analyzed. Traumatic HP patients were younger (37.1 years [SD 17.1] v. 56.0 years [SD 13.6]; P < .001). There was no difference in sex between the two groups (81.5% v. 72.2%; P = .22). HP resulted from a penetrating mechanism in 72.2% (39/54) of the trauma cohort; aortic dissection was the cause of CT in 53.2% of the non-trauma deaths [Table 1]. Despite not being the primary cause of death, median pericardial blood volume was 150 mL [SD 112; IQR 100-231 mL] in the trauma cohort. CT in the non-trauma cohort resulted from a median volume of 400 mL [SD 224; IQR 300-600 mL]. 63% (34/54) of the trauma patients with HP had more than 100 mL of pericardial blood, while 96% (76/79) of CT cases had more than this volume [Figure 1].

OPEN IN VIEWER

Table 1.

Traumatic vs Non-Traumatic Hemopericardium.

	Total	Traumatic	Non-Traumatic
Total a	133	54	79
Age (mean, SD)	48.3 (17.7)	37.1 (17.1)	56.0 (13.6)
Sex
Male	101 (75.9)	44 (81.5)	57 (72.2)
Female	32 (24.1)	10 (18.5)	22 (27.8)
Race
White	45 (33.8)	16 (29.6)	29 (36.7)
Hispanic	39 (29.3)	12 (22.2)	27 (34.2)
Black	39 (29.3)	25 (46.3)	14 (17.7)
Asian	10 (7.5)	1 (1.9)	9 (11.4)
Weight (lb, median (IQR))	172.0 (147.0-219.5)	162.5 (143.5-191.5)	179.0 (153.0-234.0)
Height (inch, median (IQR))	68.0 (66.0-71.0)	69.0 (66.0-70.8)	68.0 (65.0-71.0)
Death mechanism
Firearm b	34 (63.0)	34 (63.0)	0 (0)
Blunt force c	15 (27.8)	15 (27.8)	0 (0)
Sharp force	5 (9.2)	5 (9.2)	0 (0)
Ruptured aortic dissection	42 (53.2)	0 (0)	42 (53.2)
Ruptured myocardial infarction	37 (46.8)	0 (0)	37 (46.8)
Anatomical location
Ventricle	17 (31.5)	17 (31.5)
Atrium and ventricle	14 (25.9)	14 (25.9)
Aorta	5 (9.3)	5 (9.3)
Atrium	4 (7.4)	4 (7.4)
Pulmonary artery	3 (5.6)	3 (5.6)
Rib	3 (5.6)	3 (5.6)
Right atrium, pulmonary artery	2 (3.7)	2 (3.7)
Aorta, pulmonary artery	1 (1.8)	1 (1.8)
Left ventricle, IVC	1 (1.8)	1 (1.8)
Left ventricle, LAD	1 (1.8)	1 (1.8)
Right atrium, IVC	1 (1.8)	1 (1.8)
Right atrium, aortic transection left SCL	1 (1.8)	1 (1.8)
Right atrium, aorta, and valves	1 (1.8)	1 (1.8)
Blood Volume (ml, median (IQR))	300.0 (150.0-500.0)	150.0 (100.0-231.3)	400.0 (300.0-600.0)

OPEN IN VIEWER

Figure 1.

Volume of HP in traumatic and non-traumatic etiologies.

Despite patients with traumatic HP dying from other causes, most of the patients had considerably more volume than previously believed to cause fatal CT. Sugg et al stated that pericardial pressure of 12 to 22 mmHg would occur at 150 mL of HP, which was predicted to be fatal. ¹ However, in our study, the median value of HP was 150 mL among the trauma patients, and none died from hemopericardium. Furthermore, the volume of HP in non-trauma patients is much larger than the volume identified in trauma patients who died from other causes. Median HP in non-trauma patients was found to be 400 mL. This is similar to volumes found on recent autopsy studies compared with postmortem CT images. ⁴ There are several limitations to the study. As mentioned above, none of the trauma patients died from HP. Therefore, we could not estimate the volume of HP resulting in fatal CP. Our control arm was non-traumatic HP, including acute aortic dissection and myocardial infarction with rupture. These patients represent different etiologies; however, they share the same acute change in physiology involved in our trauma cohort. Given that the non-trauma arm had a higher median value of HP resulting in CT, it may be appropriate to redefine our estimates for the volume of pericardial blood likely to cause death from CT in trauma patients.