ORIGINAL ARTICLE
Screening for Occult Penetrating Cardiac Injuries Andrew J. Nicol, PhD, FCS,∗ Pradeep H. Navsaria, MMed, FCS,∗ Steve Beningfield, FCRad(D)SA,† Martijn Hommes, MD,∗ and Delawir Kahn, ChM, FCS∗
Objective: To determine the sensitivity of emergency department ultrasonography (US) in the diagnosis of occult cardiac injuries. Background: Internationally, US has become the investigation of choice in screening patients for a possible cardiac injury after penetrating chest trauma by detecting blood in the pericardial sac. Methods: Patients presenting with a penetrating chest wound and a possible cardiac injury to the Groote Schuur Hospital Trauma Centre between October 2001 and February 2009 were prospectively evaluated. All patients were hemodynamically stable, had no indication for emergency surgery, and had an US scan followed by subxiphoid pericardial window exploration. Results: There were a total of 172 patients (median age = 26 years; range, 11–65 years). The mechanism of injury was stab wounds in 166 (96%) and gunshot wounds in 6. The sensitivity of US in detecting hemopericardium was 86.7%, with a positive predictive value of 77%. There were 18 falsenegatives. Eleven of these false-negatives had an associated hemothorax and 6 had pneumopericardium. A single patient had 2 negative US examinations and returned with delayed cardiac tamponade. Conclusions: The sensitivity of US to detect hemopericardium in stable patients was only 86.7%. The 2 main factors that limit the screening are the presence of a hemothorax and air in the pericardial sac. A new regimen for screening of occult injuries to make allowance for this is proposed. Keywords: hemopericardium, occult cardiac injury, penetrating cardiac trauma, subxiphoid pericardial window, ultrasound (Ann Surg 2015;261:573–578)
A
penetrating cardiac injury (PCI) is frequently lethal, with an estimated prehospital mortality rate of 86%. Once the patient reaches hospital, clinicians need to promptly recognize this injury, especially as certain patients may be relatively asymptomatic.1,2 The diagnosis of a PCI in the patient who does not have an acute indication for emergency surgery can be difficult. As clinical examination may be unremarkable, many surgeons and emergency medicine physicians rely heavily on ultrasonography (US) of the pericardial sac seeking the presence of fluid (blood) to guide their management in hemodynamically stable patients. US has been reported in recent prospective studies to have a sensitivity of 100% and specificities in the range of 94.7% to 100% for the detection of occult cardiac injuries and has become the investigation of choice worldwide for screening.3–8 Unfortunately, although most of the studies included a large number of patients with penetrating chest trauma, only small numbers with
From the ∗ Trauma Centre, Department of Surgery; and †Department of Radiology, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa. This paper was presented at the 14th European Congress of Trauma and Emergency Surgery, Lyon, France, May 4–7, 2013, and formed part of a PhD dissertation on the “Current Management of Penetrating Cardiac Trauma” awarded at the University of Cape Town, South Africa in 2012. Disclosure: The authors declare no conflict of interest. Supported by the Medical Research Council of South Africa. Reprints: Andrew J. Nicol, FCS, PhD, Trauma Centre, Groote Schuur Hospital, Anzio Rd, Cape Town 7925, South Africa. E-mail: [email protected]. C 2014 Wolters Kluwer Health, Inc. All rights reserved. Copyright ISSN: 0003-4932/14/26103-0573 DOI: 10.1097/SLA.0000000000000713
Annals of Surgery r Volume 261, Number 3, March 2015
actual cardiac injuries were present (Table 1). The largest prospective series in 1999 by Rozycki et al6 included only 29 patients with confirmed cardiac trauma. All these studies have reported no falsenegative results. The experience at our institution in dealing with a large number of PCIs is that although US is an extremely useful screening tool, false-negatives do occur. In 2009, Ball et al9 highlighted the problem of false-negative US examination in patients with a hemothorax being screened for occult cardiac trauma. Five false-negative US examinations were reported in a series of 228 patients, with 2 of these patients dying as a result of the missed cardiac injury. A hemothorax was present in all 5 of these patients and the reason for false-negative US examination was possibly that the hemopericardium had drained through a pericardial laceration into the chest. Alternately, the pericardial fluid may have been discounted as a medially located pleural fluid or have interfered with the interface between the pericardium and the lung. In a retrospective study of 128 PCIs, Harris et al10 noted 6 false-negatives in 56 cardiac US scans, with 3 of these patients presenting with large hemothoraces that required surgery to control bleeding. On this basis, Cull and Bokhari11 from Cook County Hospital in Chicago have suggested that a subxiphoid pericardial window (SPW) should replace echocardiograms in patients with PCI and an associated hemothorax. There have been attempts to define the anatomical region at risk for PCI to recognize high-risk patients. The terms “cardiac proximity,”12 “cardiac box,”13 “cardiac silhouette,”14 and “precordium”15 are frequently mentioned as being more prone to PCI (Fig. 1A–C). The high incidence of PCIs medial to the midclavicular line is well documented, but PCIs do occur outside of this range as well.16 Two thirds of patients with PCIs have skin entry wounds between the left anterior axillary line and the left sternal border, possibly as most of the assailants are right-handed.17 Paradoxically, Degiannis et al18 showed a higher mortality rate for cardiac stab wounds that lay outside of the precordium (25%), rather than where the stab wound was sited over the precordium (4%). It is apparent that a revision of the chest surface markings requiring a high index of suspicion and screening for PCI is required. The high mortality rate from a missed injury emphasizes the need for accurate screening. If US is to be the investigation of choice, then it is essential that physicians dealing with penetrating chest trauma understand the advantages and the limitations of US so that appropriate decisions are made for safe patient management. The aim of this study was to determine the sensitivity and positive predictive value of emergency department US in the screening of cardiac injuries and to propose a new screening regimen.
METHODS All patients presenting to the Groote Schuur Hospital Trauma Centre with penetrating chest wounds and possible cardiac injuries between October 2001 and February 2009 were prospectively included into a cardiac database. This was part of a randomized controlled trial examining whether stable patients with hemopericardium could be managed by drainage-only via an SPW versus a median sternotomy.19 All patients were fully conscious and were either stable or required less than 2 L of crystalloids to achieve hemodynamic stability and www.annalsofsurgery.com | 573
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
TABLE 1. Prospective Series Dealing With the Detection of Hemopericardium on US Study
Design
Ma et al (1995)3 Rozycki et al (1996)4 Rozycki et al (1998)5 Rozycki et al (1999)6 Bokhari et al (2004)7 Tayal et al (2004)8
Prospective single center Prospective single center Prospective single center Prospective multicenter Prospective single center Prospective single center
No. Patients With No. Cardiac Injuries Penetrating Chest Injuries Confirmed at Surgery Sensitivity, % Specificity, % False- False+ Not stated 247 313 261 49 32
6 10 22 29 Not stated 8
100 100 100 100 100 100
99 100 99 97 50 100
0 0 0 0 0 0
1 0 2 7 20 0
FIGURE 1. Chest surface markings currently in use for cardiac injury screening. A, Cardiac proximity. B, Cardiac box. C, Cardiac silhouette or precordium. had no indication for emergency surgery. Patients were excluded if they required emergency surgery for hemodynamic instability or an exsanguinating thoracic bleed, if there was cardiac tamponade, or if they were due to have a laparotomy for associated thoracoabdominal or abdominal injury. All patients were clinically evaluated, and for each patient, an erect chest radiograph and an electrocardiogram were obtained. A central venous catheter was placed at the discretion of the treating physician and the central venous pressure was measured in cm of H2 O. An US scan of the pericardial sac was obtained by an UStrained radiology resident or radiology specialist in the resuscitation room, seeking the presence of fluid in the pericardial sac. If present, the maximum width of fluid in the sac was measured in millimeters. The US result was considered to be positive for hemopericardium if there was any fluid in the pericardial sac. If the visualization was poor, the US report was considered to be equivocal and the reason noted. The presence or absence of a hemothorax on either side was also noted. All patients who had acute hemopericardium confirmed by US, or where there was an equivocal US report, underwent SPW exploration. If the US report was negative but there remained a clinical suspicion of an occult cardiac injury, then the US was repeated at 24 hours. A clinical suspicion of a cardiac injury was recorded if there was a history of shock in the prehospital phase, if the electrocardiogram showed the presence of J-waves,20 if there was a straight left heart border21 on the chest radiograph or enlargement of the cardiac size on a repeat chest radiograph. SPW exploration was performed in an operating theatre under general anesthetic and was considered positive if any blood, blood staining, or blood clots were noted within the pericardial sac. The technique for SPW exploration was to use a small vertical incision 574 | www.annalsofsurgery.com
directly over the xiphoid process and then to divide the linea alba. The xiphoid process was dissected to allow a Langenbeck retractor to be placed underneath it to allow sternal elevation. Dissection was then continued to the diaphragm and the pleura on both sides, and a swab-on-a-stick was then used to wipe away the prepericardial fat pad. The head of the bed was then elevated to improve visualization of the pericardium and the pericardium was grasped between 2 Allis forceps to allow an incision to be made with a scissors into the tented up portion. This incision was then enlarged to 4 to 5 cm.
Definitions A true-positive result was when both the US scan and the SPW exploration confirmed the presence of blood in the pericardial sac. A false-positive result arose when the US scan demonstrated pericardial fluid but SPW exploration showed a negative result. A false-negative result was a normal US scan with blood in the pericardium on SPW exploration. Prehospital shock was noted when a blood pressure of less than 100 mm Hg was recorded at any stage during the prehospital phase. The size of the pericardial effusion was taken at the maximum width of the fluid in the pericardial sac and was measured in millimeters. Permission for the study was obtained from the University of Cape Town Research Ethics Committee. Informed consent for participation in the study was obtained from every patient. Continuous variables were compared by means of the t test or Wilcoxon rank-sum test. χ 2 analysis and the Fisher exact test were used for the analysis of categorical variables, where appropriate. C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Antibiotic Prophylaxis in Clean Surgery
RESULTS A total of 172 patients who had both a preoperative pericardial sac US scan and SPW exploration for a possible cardiac injury between October 2001 and February 2009 were analyzed. The median age of the patients was 26 years (range, 11–65 years) and 168 (96%) were males. The mechanism of the PCI was stab wounds in 166 (96%) and low-velocity gunshot wounds in 6 (4%). Fifty-nine of the 172 patients (34%) were hypotensive in the prehospital phase. The median hemoglobin on presentation was 11.0 g% (range, 4.7–17.0). The US examination was positive for fluid in 152 of the 172 patients with a mean pericardial sac measurement of 8.1 mm (range, 0–30 mm). There were 117 true-positives confirmed on SPW exploration, 35 (19%) false-positives, and 18 false-negatives. These results are illustrated in Table 2. The sensitivity of US in detecting hemopericardium was 86.7%, with a positive predictive value of 77%. Eighteen patients with a false-negative US result showed that 6 patients had pneumopericardium but no fluid was seen on the US scan. Eleven of the remaining 12 patients had left-sided hemothoraces, managed with intercostal drains. Seven of the 18 patients with false-negative results had a repeat US scan performed 24 hours postadmission; in 6, pericardial fluid was then detected by US and confirmed at surgery. There was 1 patient with 2 negative US examinations who was discharged home only to return with a delayed symptomatic pericardial effusion. In 6 patients with initially a negative US result that became positive at 24 hours, the indications for a repeat US scan at 24 hours are shown in Table 3. Two patients had pericardial air that resulted in suboptimal visualization, 2 patients had a straight left heart border on the repeat chest radiograph at 24 hours, another patient had an increase in the size of the heart on the repeat chest radiograph, and the final patient had an elevated central venous pressure of 23 cm H2 O on presentation. The mean diameter of the pericardial effusion on US examination in the 37 patients with a negative SPW exploration result was 6.7 mm [95% confidence interval (CI), 5.0–8.4]. In the 135 patients with a positive SPW exploration result for blood, the mean diameter of the pericardial effusion was 8.5 mm (95% CI, 7.6–9.4). The size of the effusion was a predictor of an occult PCI (P = 0.03), but there was, however, no minimal measurement of effusion that would safely exclude an occult PCI. There were 3 patients with 3-mm effusions
TABLE 2. US as a Screen for Potential Cardiac Injuries SPW +ve SPW –ve Total
US Positive
US Negative
Total
117 35 152
18 2 20
135 37 172
who had a positive SPW exploration result. In these 3 patients with 3-mm effusions, all were found to have cardiac injuries at sternotomy.
DISCUSSION In the 1980s, German surgeons pioneered the use of bedside US for the management of trauma patients, but it has only been since the 1990s that this technology has become more widely accepted and is now a standard practice in trauma centers in the United States.4–6,22–24 The adoption of US into the Advanced Trauma Life Support Course due to the appreciation of the early diagnostic role that it can provide has led to FAST (Focused Assessment with Sonography for Trauma) becoming a catch phrase in the initial assessment and evaluation of these patients. FAST is in essence a rapid assessment to determine the presence of pericardial or peritoneal fluid to aid in identifying the cause for instability, permitting more rapid management.25 In fact, FAST has had a dramatic effect on the management of trauma resuscitation, and with the concept of resuscitation, US is rapidly evolving into an “extension of the physical examination.”26 An international consensus conference defined FAST as an expeditious, focused interrogation of the pericardial and peritoneal spaces looking for free fluid as a marker of injury.27 It has been shown to be extremely reliable in confirming or excluding the peritoneal cavity as the source of blood loss in the unconscious patient who is hemodynamically unstable after sustaining blunt trauma.5,28 The use of US in the resuscitation room has a number of advantages as it is noninvasive, compact, portable, rapid, and cost-effective. However, its reliability in the evaluation of the pericardium forms the basis for our study. The largest prospective study of the role of US in the detection of acute hemopericardium in patients with possible penetrating cardiac wounds was a multicenter trial in 1999.6 Pericardial US was performed in 261 patients. There were 29 cardiac injuries identified, with no false-negatives and 7 false-positive examinations, resulting in a sensitivity of 100% and a specificity of 96.9%. Patients with a negative US result were followed up for a minimum period of 23 hours. Six of their 7 patients with false-positive examinations had an associated hemothorax and on SPW exploration were found to have benign pericardial effusions. All tests have their shortcomings, and it is important for surgeons to be aware of the sensitivity of the tests that they use in guiding patient management. The sensitivity of US at our institution from the present series is 86.7%, with a positive predictive value of 77%. Although this is could be considered good for a screening test, it also indicates that US cannot be relied upon as the sole screening modality. A cardiac injury is potentially rapidly lethal and preferably a test that is oversensitive is required. The 18 US false-negative examinations in our series call into question the entire role of US. There would appear to be 2 main factors limiting the screening sensitivity of US. The first is the presence of pericardial air, and the second a left hemothorax.
TABLE 3. Six False-Negative US Examinations With a Repeat US Scan at 24 Hours US Examination on Presentation
Indication for Repeat US Examination
Repeat US Result
1 2 3
Pneumomediastinum Pneumopericardium Normal
13-mm hemopericardium 5-mm hemopericardium 15-mm hemopericardium
4 5 6
Normal Normal Normal
Suboptimal visualization Suboptimal visualization Increasing heart size on chest radiograph Straight left heart border Straight left heart border Central venous pressure of 23 cm H2 O on admission
Patient No.
C 2014 Wolters Kluwer Health, Inc. All rights reserved.
15-mm hemopericardium 8-mm hemopericardium 6-mm hemopericardium
www.annalsofsurgery.com | 575
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
FIGURE 2. A and B, The “cardiac zone”— the high-risk area for a PCI. Injury in the Cardiac Zone
Ultrasound (US) pericardial sac
Posive US or Pneumopericardium
Equivocal US
Negave US
HT present or Clinical suspicion
SPW
CT scan or SPW
CT scan or Repeat US at 24-h
SPW = subxiphoid pericardial window HT = hemothorax D/C = discharge
Six of the 18 patients with false-negatives had pneumopericardium and 11 of the remaining 12 patients had a left hemothorax. Blood in the pericardial sac is viewed as an echo-lucent dark area and the visceral and parietal layers of the pericardium as distinct echogenic lines. There is little doubt that both air and blood may interfere with the ability to visualize fluid in the pericardial sac because of diminished sound transmission, loss of normal interfaces, and distortion of 576 | www.annalsofsurgery.com
No HT
D/C
FIGURE 3. Management algorithm for the screening of PCI in a hemodynamically stable patient. anatomy. A repeat US scan after 24 hours permitted a further 6 patients with hemopericardium to be identified, and this rescan should be considered in cases where clinical suspicion persists. The epicardial fat pad may occasionally be misdiagnosed as a pericardial effusion. However, the accuracy of US examination has also been shown to be dependent on US experience.29 The false-positive rate of US in our series was 19%. “The limitations of this study are the lack of routine C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
follow-up US in all patients with an initial negative scan, the fact that the experience in US may have varied between the radiologists, and the specific US probe selected was not documented.” The size of the pericardial effusion was found to be statistically significant in predicting the presence of hemopericardium (P = 0.03). The detection of any fluid (blood) in the pericardial sac after penetrating chest trauma is thought to be significant and warrant surgery. There is little mention in the literature about the volume of pericardial fluid or the measurement of the effusion. Some healthy patients do have a certain amount of pericardial fluid and this may potentially cause a false-positive result. Patel et al30 published a retrospective study on surgeon-performed sonography. Twenty patients were found to have a cardiac injury at surgery, and they described a true-positive result as the presence of at least 5 mm of fluid and a cardiac injury identified at surgery. There is no discussion as to where this 5-mm figure was derived from and what the significance is of having less than 5 mm of fluid. In our series, there were patients who had a positive result for SPW exploration and yet they had less than 5 mm of fluid present on the US scan. In 3 patients, the size of the pericardial effusion was noted to be 3 mm and in all of them, a cardiac injury was documented at sternotomy. As a result, it would appear that there is no minimum “safe” limit and the presence of any fluid in the pericardial sac should be an indicator of a PCI. There have been very few studies that have specifically investigated transthoracic echocardiography (ECHO) in the diagnosis of penetrating cardiac trauma. Meyer et al12 compared the sensitivity and specificity of ECHO against the gold standard of SPW exploration in 1995. In their prospective study, 105 hemodynamically stable patients with PCI in the “proximity of the heart” were included. Eighty-nine patients had a true-negative ECHO result, as performed and interpreted by the trauma surgeon. There were 9 cardiac injuries identified at surgery, with ECHO confirming 5 of these. They also had 7 falsepositives and 4 false-negatives. Their sensitivity of ECHO was 56%, with 93% specificity for detecting a cardiac injury. This study also highlighted the presence of a hemothorax having an adverse influence on the ability of ECHO to detect hemopericardium. In 1990, Jimenez et al31 published a prospective study comparing ECHO with SPW exploration in the diagnosis of the occult cardiac injury after penetrating chest trauma. Seventy-three patients were included, all having undergone ECHO and a subsequent SPW exploration. There were 9 cardiac injuries detected. The sensitivity of ECHO for diagnosing an injury was 90%, with a specificity of 97%; there was a single false-negative result. Grewal et al14 in 1995 reported on the evaluation of SPW exploration in the diagnosis of the occult cardiac injury. In that study, there were 2 patients with negative ECHO findings who had cardiac injuries. The one case presented with delayed cardiac tamponade after discharge, and the second with an episode of hypotension while still in the hospital. Bolton et al32 documented 5 cases where ECHO failed to detect a cardiac injury. There are numerous case reports of where ECHO has failed to detect the cardiac injury in asymptomatic patients.33 A 23-year-old man from Turkey was admitted 6 months after his initial injury with a traumatic aorto-left atrial fistula and features of cardiac tamponade. The initial clinical evaluation and ECHO on admission to hospital showed negative findings.34 From Korea, a 22-year-old woman presented with a stab wound to the chest with a left-sided hemothorax. The initial ECHO examination showed a small amount of fluid in the pericardial sac without any shunt or injury to the myocardium. It was only on follow-up ECHO performed the next day after clinical deterioration that a ventricular septal defect was demonstrated, which was repaired surgically.35 From Spain, a 32-year-old man with a precordial stab wound and normal ECHO examination on presentation, who 4 days later experienced a cardiac arrest due to delayed cardiac C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Antibiotic Prophylaxis in Clean Surgery
tamponade. Emergency surgery revealed a laceration to midsegment of the left anterior descending coronary artery that required suturing.36 There are 2 cases of cardiac tamponade in the United States that presented 4 and 14 days after chest stab wounds with initial normal admission evaluation of US, ECHO, and computed tomographic (CT) scan of the chest.37 It would appear that ECHO is as sensitive as US in the detection of the occult cardiac injuries. The major concern is the high number of false-negative results and consequently our view that US cannot be a stand-alone screening test for cardiac injuries. A further consideration is the availability of the technology in the trauma and emergency departments in hospitals and the available expertise to provide the service afterhours. CT is being used in many centers as the primary imaging modality in hemodynamically stable patients with penetrating chest trauma. This has been driven by the speed and excellent image reconstruction of multidetector CT.38 There is very little data on the role of CT in screening for a cardiac injury. Nagy et al39 published a retrospective study in 1996 on 60 stable patients reporting a sensitivity and specificity of 100% and 96.6%, respectively. Unfortunately, only 2 of the 60 patients had a confirmed cardiac injury. It does, however, seem logical to include CT in a screening protocol, particularly when the US result is equivocal, a hemothorax is present, or when clinical suspicion of a PCI is high. However, further studies on the sensitivity of CT are required and one does need to be aware of the radiation dose in young patients and the fact that US is more frequently available in the emergency department. It has been well described that despite skin entry wounds outside the previously mentioned surface anatomical landmarks, these injuries can cause PCI. In a 15-month postmortem study of PCI in Cape Town, 507 wounds in the thorax resulted in 240 cardiac injuries; of those with cardiac injuries, only 33% were precordial and 23% were from the posterior chest wall.40 Evans et al41 observed that cardiac injuries may occur whether the entrance wound is in the posterior chest wall or in the subcostal or subxiphoid abdominal region. There are numerous case reports in the literature that warn of cardiac injuries occurring from remote sites.33,42 A prospective study of 66 patients by Buckman et al15 showed that the probability of a cardiac injury varied according to the mechanism of trauma. They found that 80% of stab wounds with associated PCI were located precordially but only 46% of gunshots were located precordially. They felt that any gunshot wound to the torso carried the potential of a cardiac injury.
NEW SCREENING REGIMEN We would, therefore, like to propose that the high-risk area for cardiac stab wounds is the region extending from the right anterior axillary line anteriorly across the precordium to the posterior left chest ending in the midline at the spinous processes, with the superior margin being the supraclavicular areas and the inferior margin being the costal margins, including the upper epigastrium. The authors call this area the “cardiac zone” (Figs. 2A, B). Any gunshot wound to the neck, shoulders, chest, back, and abdomen should be considered to have a possible cardiac injury and should be screened appropriately. Screening in hemodynamically stable patients should include an erect chest radiograph if possible, an electrocardiogram, and an US scan of the pericardial sac. The screening US must be performed immediately on admission to the emergency department, as an occult PCI may cause rapid deterioration in the patient’s condition and is potentially fatal. If the screening US scan is positive for blood in the pericardial sac, then the patient must undergo SPW exploration. The presence of any fluid in the pericardial sac should be regarded as a positive US result as there is no particular measurement in millimeters that safely excludes a PCI. If the screening US scan is equivocal, then the patient should either undergo SPW exploration or a CT scan of chest. If the initial screening US scan shows negative result www.annalsofsurgery.com | 577
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
but there is a hemothorax, then a CT scan of the chest should be obtained immediately or the US scan repeated at 24 hours (Fig. 3). If the US scan shows negative result but there is clinical suspicion of a PCI due to the presence of J-waves on the electrocardiogram, or a straight left heart border on chest radiograph, then a CT chest scan should be obtained or else the US scan of the pericardial sac should be repeated at 24 hours. If there is pneumopericardium present on chest radiograph, US scan, or CT scan, then SPW exploration should be undertaken. The patient with a negative US result and no hemothorax, no pneumopericardium, or clinical suspicion of a PCI may be discharged home.
REFERENCES 1. Ivatury RR, Rohman M, Steichen FM, et al. Penetrating cardiac injuries: twenty-year experience. Am Surg. 1987;53:310–317. 2. Demetriades D, van der Veen BW. Penetrating injuries of the heart: experience over two years in South Africa. J Trauma. 1983;23:1034–1041. 3. Ma OJ, Mateer JR, Ogata M, et al. Prospective analysis of a rapid trauma ultrasound examination performed by emergency physicians. J Trauma. 1995;38:879–385. 4. Rozycki GS, Feliciano DV, Schmidt JA, et al. The role of surgeon-performed ultrasound in patients with possible cardiac wounds. Ann Surg. 1996;223:737– 746. 5. Rozycki GS, Ballard RB, Feliciano DV, et al. Surgeon-performed ultrasound for the assessment of truncal injuries: lessons learned from 1540 patients. Ann Surg. 1998;228:557–567. 6. Rozycki GS, Feliciano DV, Ochsner MG, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma. 1999;46:543–552. 7. Bokhari F, Nagy K, Roberts R, et al. The ultrasound screen for penetrating truncal trauma. Am Surg. 2004;70:316–321. 8. Tayal VS, Beatty MA, Marx JA, et al. FAST (focused assessment with sonography in trauma) accurate for cardiac and intraperitoneal injury in penetrating anterior chest trauma. J Ultrasound Med. 2004;23:467–472. 9. Ball CG, Williams BH, Wyrzykowski AD, et al. A caveat to the performance of pericardial ultrasound in patients with penetrating cardiac wounds. J Trauma. 2009;67:1123–1124. 10. Harris DG, Papagiannopoulos KA, Pretorius J, et al. Current evaluation of cardiac stab wounds. Ann Thorac Surg. 1999;68:2119–2122. 11. Cull JD, Bokhari F. Penetrating chest injury in the setting of a hemothorax: should pericardial windows replace echocardiograms? Am Surg. 2013;79: E56–E57. 12. Meyer DM, Jessen ME, Grayburn PA. Use of echocardiography to detect occult cardiac injury after penetrating thoracic trauma: a prospective study. J Trauma. 1995;39:902–909. 13. Asensio JA, Roldan G, Petrone P, et al. Cardiac trauma. Trauma. 2001;3:69–77. 14. Grewal H, Ivatury RR, Divakar M, et al. Evaluation of subxiphoid pericardial window used in the detection of occult cardiac injury. Injury. 1995;26:305–310. 15. Buckman RF, Jr, Badellino MM, Mauro LH, et al. Penetrating cardiac wounds: prospective study of factors influencing initial resuscitation. J Trauma. 1993;34:717–727. 16. Jones EW, Helmsworth J. Penetrating wounds of the heart. Thirty years’ experience. Arch Surg. 1968;96:671–682. 17. Demetriades D. Cardiac wounds. Experience with 70 patients. Ann Surg. 1986;203:315–317. 18. Degiannis E, Loogna P, Doll D, et al. Penetrating cardiac injuries: recent experience in South Africa. World J Surg. 2006;30:1258–1264. 19. Nicol AJ, Navsaria PH, Hommes M, et al. Sternotomy or drainage for a hemopericardium after penetrating trauma. A randomized controlled trial. Ann Surg. 2014;259:438–442.
578 | www.annalsofsurgery.com
20. Nicol AJ, Navsaria PH. The J-wave: a new electrocardiographic sign of an occult cardiac injury. Injury. 2014;45:112–115. 21. Nicol AJ, Navsaria PH, Beningfield S, et al. A straight left heart border: a new radiological sign of a hemopericardium. World J Surg. 2014;38: 211–214. 22. Boulanger BR, Brenneman FD, McLellan BA, et al. A prospective study of emergent abdominal sonography after blunt trauma. J Trauma. 1995;39:325– 330. 23. Kirkpatrick AW, Sirois M, Laupland KB, et al. Prospective evaluation of handheld focused abdominal sonography for trauma (FAST) in blunt abdominal trauma. Can J Surg. 2005;48:453–460. 24. Boulanger BR, McLellan BA, Brenneman FD, et al. Emergent abdominal sonography as a screening test in a new diagnostic algorithm for blunt trauma. J Trauma. 1996;40:867–874. 25. Dente CJ, Rozycki GS. Surgeon performed ultrasound in trauma and surgical critical care. In: Feliciano DV, Mattox KL, Moore EE, eds. Trauma. 7th ed. New York: McGraw-Hill Medical; 2013:301–321. 26. Gillman LM, Ball CG, Panebianco N, et al. Clinician performed resuscitative ultrasonography for the initial evaluation and resuscitation of trauma. Scand J Trauma Resusc Emerg Med. 2009;17:34. 27. Scalea TM, Rodriguez A, Chiu WC, et al. Focused assessment with sonography for trauma (FAST): results from an international consensus conference. J Trauma. 1999;46:466–472. 28. Kirkpatrick AW, Sirois M, Laupland KB, et al. Hypotension after blunt abdominal trauma: the role of emergent abdominal sonography in surgical triage. J Trauma. 1996;41:815–820. 29. Blaivas M, DeBehnke D, Phelan MB. Potential errors in the diagnosis of pericardial effusion on trauma ultrasound for penetrating injuries. Acad Emerg Med. 2000;7:1261–1266. 30. Patel AN, Brennig C, Cotner J, et al. Successful diagnosis of penetrating cardiac injury using surgeon-performed sonography. Ann Thorac Surg. 2003;76:2043– 2047. 31. Jimenez E, Martin M, Krukenkamp I, et al. Subxiphoid pericardiotomy versus echocardiography: a prospective evaluation of the diagnosis of occult penetrating cardiac injury. Surgery. 1990;108:676–680. 32. Bolton JW, Bynoe RP, Lazar HL, et al. Two-dimensional echocardiography in the evaluation of penetrating intrapericardial injuries. Ann Thorac Surg. 1993;56:506–509. 33. Pathi V, Sutherland FW, Ireland A, et al. Breaking the rules! Cardiac injury from remote entry sites. J Accid Emerg Med. 1998;15:115–116. 34. Belgi A, Mete A, Avsar O, et al. Late presentation of complex cardiac defect following penetrating cardiac trauma: case report. Turk Kardiyol Dern Ars. 2011;31:1–4. 35. Jeon K, Lim WH, Kang SH, et al. Delayed diagnosis of traumatic ventricular septal defect in penetrating chest injury: small evidence on echocardiography makes big difference. J Cardiovasc Ultrasound. 2010;18:28–30. 36. Enriquez SG, Fernandez CG, Entem FR, et al. Delayed pericardial tamponade after penetrating chest trauma. Eur J Emerg Med. 2005;12:86–88. 37. Simmons JD, Haraway AN, Schmieg RE, Jr, et al. Is there a role for secondary thoracic ultrasound in patients with penetrating injuries to the anterior mediastinum? Am Surg. 2008;74:11–14. 38. De Vries CS, Africa M, Gebremariam FA, et al. The imaging of stab injuries. Acta Radiol. 2010;51:92–106. 39. Nagy KK, Gilkey SH, Roberts RR, et al. Computed tomography screens stable patients at risk for penetrating cardiac injury. Acad Emerg Med. 1996;3:1024– 1027. 40. Oakland C, Vivian J. Penetrating cardiac injuries [Letter to the Editor]. BMJ. 1987;295:502. 41. Evans J, Gray LA, Jr, Rayner A, et al. Principles for the management of penetrating cardiac wounds. Ann Surg. 1979;189:777–784. 42. Claassen CW, O’Connor JV, Gens D, et al. Penetrating cardiac injury: think outside the box. J Trauma. 2010;68:E71–E73.
C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Screening for Occult Penetrating Cardiac Injuries Andrew J. Nicol, PhD, FCS,∗ Pradeep H. Navsaria, MMed, FCS,∗ Steve Beningfield, FCRad(D)SA,† Martijn Hommes, MD,∗ and Delawir Kahn, ChM, FCS∗
Objective: To determine the sensitivity of emergency department ultrasonography (US) in the diagnosis of occult cardiac injuries. Background: Internationally, US has become the investigation of choice in screening patients for a possible cardiac injury after penetrating chest trauma by detecting blood in the pericardial sac. Methods: Patients presenting with a penetrating chest wound and a possible cardiac injury to the Groote Schuur Hospital Trauma Centre between October 2001 and February 2009 were prospectively evaluated. All patients were hemodynamically stable, had no indication for emergency surgery, and had an US scan followed by subxiphoid pericardial window exploration. Results: There were a total of 172 patients (median age = 26 years; range, 11–65 years). The mechanism of injury was stab wounds in 166 (96%) and gunshot wounds in 6. The sensitivity of US in detecting hemopericardium was 86.7%, with a positive predictive value of 77%. There were 18 falsenegatives. Eleven of these false-negatives had an associated hemothorax and 6 had pneumopericardium. A single patient had 2 negative US examinations and returned with delayed cardiac tamponade. Conclusions: The sensitivity of US to detect hemopericardium in stable patients was only 86.7%. The 2 main factors that limit the screening are the presence of a hemothorax and air in the pericardial sac. A new regimen for screening of occult injuries to make allowance for this is proposed. Keywords: hemopericardium, occult cardiac injury, penetrating cardiac trauma, subxiphoid pericardial window, ultrasound (Ann Surg 2015;261:573–578)
A
penetrating cardiac injury (PCI) is frequently lethal, with an estimated prehospital mortality rate of 86%. Once the patient reaches hospital, clinicians need to promptly recognize this injury, especially as certain patients may be relatively asymptomatic.1,2 The diagnosis of a PCI in the patient who does not have an acute indication for emergency surgery can be difficult. As clinical examination may be unremarkable, many surgeons and emergency medicine physicians rely heavily on ultrasonography (US) of the pericardial sac seeking the presence of fluid (blood) to guide their management in hemodynamically stable patients. US has been reported in recent prospective studies to have a sensitivity of 100% and specificities in the range of 94.7% to 100% for the detection of occult cardiac injuries and has become the investigation of choice worldwide for screening.3–8 Unfortunately, although most of the studies included a large number of patients with penetrating chest trauma, only small numbers with
From the ∗ Trauma Centre, Department of Surgery; and †Department of Radiology, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa. This paper was presented at the 14th European Congress of Trauma and Emergency Surgery, Lyon, France, May 4–7, 2013, and formed part of a PhD dissertation on the “Current Management of Penetrating Cardiac Trauma” awarded at the University of Cape Town, South Africa in 2012. Disclosure: The authors declare no conflict of interest. Supported by the Medical Research Council of South Africa. Reprints: Andrew J. Nicol, FCS, PhD, Trauma Centre, Groote Schuur Hospital, Anzio Rd, Cape Town 7925, South Africa. E-mail: [email protected]. C 2014 Wolters Kluwer Health, Inc. All rights reserved. Copyright ISSN: 0003-4932/14/26103-0573 DOI: 10.1097/SLA.0000000000000713
Annals of Surgery r Volume 261, Number 3, March 2015
actual cardiac injuries were present (Table 1). The largest prospective series in 1999 by Rozycki et al6 included only 29 patients with confirmed cardiac trauma. All these studies have reported no falsenegative results. The experience at our institution in dealing with a large number of PCIs is that although US is an extremely useful screening tool, false-negatives do occur. In 2009, Ball et al9 highlighted the problem of false-negative US examination in patients with a hemothorax being screened for occult cardiac trauma. Five false-negative US examinations were reported in a series of 228 patients, with 2 of these patients dying as a result of the missed cardiac injury. A hemothorax was present in all 5 of these patients and the reason for false-negative US examination was possibly that the hemopericardium had drained through a pericardial laceration into the chest. Alternately, the pericardial fluid may have been discounted as a medially located pleural fluid or have interfered with the interface between the pericardium and the lung. In a retrospective study of 128 PCIs, Harris et al10 noted 6 false-negatives in 56 cardiac US scans, with 3 of these patients presenting with large hemothoraces that required surgery to control bleeding. On this basis, Cull and Bokhari11 from Cook County Hospital in Chicago have suggested that a subxiphoid pericardial window (SPW) should replace echocardiograms in patients with PCI and an associated hemothorax. There have been attempts to define the anatomical region at risk for PCI to recognize high-risk patients. The terms “cardiac proximity,”12 “cardiac box,”13 “cardiac silhouette,”14 and “precordium”15 are frequently mentioned as being more prone to PCI (Fig. 1A–C). The high incidence of PCIs medial to the midclavicular line is well documented, but PCIs do occur outside of this range as well.16 Two thirds of patients with PCIs have skin entry wounds between the left anterior axillary line and the left sternal border, possibly as most of the assailants are right-handed.17 Paradoxically, Degiannis et al18 showed a higher mortality rate for cardiac stab wounds that lay outside of the precordium (25%), rather than where the stab wound was sited over the precordium (4%). It is apparent that a revision of the chest surface markings requiring a high index of suspicion and screening for PCI is required. The high mortality rate from a missed injury emphasizes the need for accurate screening. If US is to be the investigation of choice, then it is essential that physicians dealing with penetrating chest trauma understand the advantages and the limitations of US so that appropriate decisions are made for safe patient management. The aim of this study was to determine the sensitivity and positive predictive value of emergency department US in the screening of cardiac injuries and to propose a new screening regimen.
METHODS All patients presenting to the Groote Schuur Hospital Trauma Centre with penetrating chest wounds and possible cardiac injuries between October 2001 and February 2009 were prospectively included into a cardiac database. This was part of a randomized controlled trial examining whether stable patients with hemopericardium could be managed by drainage-only via an SPW versus a median sternotomy.19 All patients were fully conscious and were either stable or required less than 2 L of crystalloids to achieve hemodynamic stability and www.annalsofsurgery.com | 573
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
TABLE 1. Prospective Series Dealing With the Detection of Hemopericardium on US Study
Design
Ma et al (1995)3 Rozycki et al (1996)4 Rozycki et al (1998)5 Rozycki et al (1999)6 Bokhari et al (2004)7 Tayal et al (2004)8
Prospective single center Prospective single center Prospective single center Prospective multicenter Prospective single center Prospective single center
No. Patients With No. Cardiac Injuries Penetrating Chest Injuries Confirmed at Surgery Sensitivity, % Specificity, % False- False+ Not stated 247 313 261 49 32
6 10 22 29 Not stated 8
100 100 100 100 100 100
99 100 99 97 50 100
0 0 0 0 0 0
1 0 2 7 20 0
FIGURE 1. Chest surface markings currently in use for cardiac injury screening. A, Cardiac proximity. B, Cardiac box. C, Cardiac silhouette or precordium. had no indication for emergency surgery. Patients were excluded if they required emergency surgery for hemodynamic instability or an exsanguinating thoracic bleed, if there was cardiac tamponade, or if they were due to have a laparotomy for associated thoracoabdominal or abdominal injury. All patients were clinically evaluated, and for each patient, an erect chest radiograph and an electrocardiogram were obtained. A central venous catheter was placed at the discretion of the treating physician and the central venous pressure was measured in cm of H2 O. An US scan of the pericardial sac was obtained by an UStrained radiology resident or radiology specialist in the resuscitation room, seeking the presence of fluid in the pericardial sac. If present, the maximum width of fluid in the sac was measured in millimeters. The US result was considered to be positive for hemopericardium if there was any fluid in the pericardial sac. If the visualization was poor, the US report was considered to be equivocal and the reason noted. The presence or absence of a hemothorax on either side was also noted. All patients who had acute hemopericardium confirmed by US, or where there was an equivocal US report, underwent SPW exploration. If the US report was negative but there remained a clinical suspicion of an occult cardiac injury, then the US was repeated at 24 hours. A clinical suspicion of a cardiac injury was recorded if there was a history of shock in the prehospital phase, if the electrocardiogram showed the presence of J-waves,20 if there was a straight left heart border21 on the chest radiograph or enlargement of the cardiac size on a repeat chest radiograph. SPW exploration was performed in an operating theatre under general anesthetic and was considered positive if any blood, blood staining, or blood clots were noted within the pericardial sac. The technique for SPW exploration was to use a small vertical incision 574 | www.annalsofsurgery.com
directly over the xiphoid process and then to divide the linea alba. The xiphoid process was dissected to allow a Langenbeck retractor to be placed underneath it to allow sternal elevation. Dissection was then continued to the diaphragm and the pleura on both sides, and a swab-on-a-stick was then used to wipe away the prepericardial fat pad. The head of the bed was then elevated to improve visualization of the pericardium and the pericardium was grasped between 2 Allis forceps to allow an incision to be made with a scissors into the tented up portion. This incision was then enlarged to 4 to 5 cm.
Definitions A true-positive result was when both the US scan and the SPW exploration confirmed the presence of blood in the pericardial sac. A false-positive result arose when the US scan demonstrated pericardial fluid but SPW exploration showed a negative result. A false-negative result was a normal US scan with blood in the pericardium on SPW exploration. Prehospital shock was noted when a blood pressure of less than 100 mm Hg was recorded at any stage during the prehospital phase. The size of the pericardial effusion was taken at the maximum width of the fluid in the pericardial sac and was measured in millimeters. Permission for the study was obtained from the University of Cape Town Research Ethics Committee. Informed consent for participation in the study was obtained from every patient. Continuous variables were compared by means of the t test or Wilcoxon rank-sum test. χ 2 analysis and the Fisher exact test were used for the analysis of categorical variables, where appropriate. C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Antibiotic Prophylaxis in Clean Surgery
RESULTS A total of 172 patients who had both a preoperative pericardial sac US scan and SPW exploration for a possible cardiac injury between October 2001 and February 2009 were analyzed. The median age of the patients was 26 years (range, 11–65 years) and 168 (96%) were males. The mechanism of the PCI was stab wounds in 166 (96%) and low-velocity gunshot wounds in 6 (4%). Fifty-nine of the 172 patients (34%) were hypotensive in the prehospital phase. The median hemoglobin on presentation was 11.0 g% (range, 4.7–17.0). The US examination was positive for fluid in 152 of the 172 patients with a mean pericardial sac measurement of 8.1 mm (range, 0–30 mm). There were 117 true-positives confirmed on SPW exploration, 35 (19%) false-positives, and 18 false-negatives. These results are illustrated in Table 2. The sensitivity of US in detecting hemopericardium was 86.7%, with a positive predictive value of 77%. Eighteen patients with a false-negative US result showed that 6 patients had pneumopericardium but no fluid was seen on the US scan. Eleven of the remaining 12 patients had left-sided hemothoraces, managed with intercostal drains. Seven of the 18 patients with false-negative results had a repeat US scan performed 24 hours postadmission; in 6, pericardial fluid was then detected by US and confirmed at surgery. There was 1 patient with 2 negative US examinations who was discharged home only to return with a delayed symptomatic pericardial effusion. In 6 patients with initially a negative US result that became positive at 24 hours, the indications for a repeat US scan at 24 hours are shown in Table 3. Two patients had pericardial air that resulted in suboptimal visualization, 2 patients had a straight left heart border on the repeat chest radiograph at 24 hours, another patient had an increase in the size of the heart on the repeat chest radiograph, and the final patient had an elevated central venous pressure of 23 cm H2 O on presentation. The mean diameter of the pericardial effusion on US examination in the 37 patients with a negative SPW exploration result was 6.7 mm [95% confidence interval (CI), 5.0–8.4]. In the 135 patients with a positive SPW exploration result for blood, the mean diameter of the pericardial effusion was 8.5 mm (95% CI, 7.6–9.4). The size of the effusion was a predictor of an occult PCI (P = 0.03), but there was, however, no minimal measurement of effusion that would safely exclude an occult PCI. There were 3 patients with 3-mm effusions
TABLE 2. US as a Screen for Potential Cardiac Injuries SPW +ve SPW –ve Total
US Positive
US Negative
Total
117 35 152
18 2 20
135 37 172
who had a positive SPW exploration result. In these 3 patients with 3-mm effusions, all were found to have cardiac injuries at sternotomy.
DISCUSSION In the 1980s, German surgeons pioneered the use of bedside US for the management of trauma patients, but it has only been since the 1990s that this technology has become more widely accepted and is now a standard practice in trauma centers in the United States.4–6,22–24 The adoption of US into the Advanced Trauma Life Support Course due to the appreciation of the early diagnostic role that it can provide has led to FAST (Focused Assessment with Sonography for Trauma) becoming a catch phrase in the initial assessment and evaluation of these patients. FAST is in essence a rapid assessment to determine the presence of pericardial or peritoneal fluid to aid in identifying the cause for instability, permitting more rapid management.25 In fact, FAST has had a dramatic effect on the management of trauma resuscitation, and with the concept of resuscitation, US is rapidly evolving into an “extension of the physical examination.”26 An international consensus conference defined FAST as an expeditious, focused interrogation of the pericardial and peritoneal spaces looking for free fluid as a marker of injury.27 It has been shown to be extremely reliable in confirming or excluding the peritoneal cavity as the source of blood loss in the unconscious patient who is hemodynamically unstable after sustaining blunt trauma.5,28 The use of US in the resuscitation room has a number of advantages as it is noninvasive, compact, portable, rapid, and cost-effective. However, its reliability in the evaluation of the pericardium forms the basis for our study. The largest prospective study of the role of US in the detection of acute hemopericardium in patients with possible penetrating cardiac wounds was a multicenter trial in 1999.6 Pericardial US was performed in 261 patients. There were 29 cardiac injuries identified, with no false-negatives and 7 false-positive examinations, resulting in a sensitivity of 100% and a specificity of 96.9%. Patients with a negative US result were followed up for a minimum period of 23 hours. Six of their 7 patients with false-positive examinations had an associated hemothorax and on SPW exploration were found to have benign pericardial effusions. All tests have their shortcomings, and it is important for surgeons to be aware of the sensitivity of the tests that they use in guiding patient management. The sensitivity of US at our institution from the present series is 86.7%, with a positive predictive value of 77%. Although this is could be considered good for a screening test, it also indicates that US cannot be relied upon as the sole screening modality. A cardiac injury is potentially rapidly lethal and preferably a test that is oversensitive is required. The 18 US false-negative examinations in our series call into question the entire role of US. There would appear to be 2 main factors limiting the screening sensitivity of US. The first is the presence of pericardial air, and the second a left hemothorax.
TABLE 3. Six False-Negative US Examinations With a Repeat US Scan at 24 Hours US Examination on Presentation
Indication for Repeat US Examination
Repeat US Result
1 2 3
Pneumomediastinum Pneumopericardium Normal
13-mm hemopericardium 5-mm hemopericardium 15-mm hemopericardium
4 5 6
Normal Normal Normal
Suboptimal visualization Suboptimal visualization Increasing heart size on chest radiograph Straight left heart border Straight left heart border Central venous pressure of 23 cm H2 O on admission
Patient No.
C 2014 Wolters Kluwer Health, Inc. All rights reserved.
15-mm hemopericardium 8-mm hemopericardium 6-mm hemopericardium
www.annalsofsurgery.com | 575
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
FIGURE 2. A and B, The “cardiac zone”— the high-risk area for a PCI. Injury in the Cardiac Zone
Ultrasound (US) pericardial sac
Posive US or Pneumopericardium
Equivocal US
Negave US
HT present or Clinical suspicion
SPW
CT scan or SPW
CT scan or Repeat US at 24-h
SPW = subxiphoid pericardial window HT = hemothorax D/C = discharge
Six of the 18 patients with false-negatives had pneumopericardium and 11 of the remaining 12 patients had a left hemothorax. Blood in the pericardial sac is viewed as an echo-lucent dark area and the visceral and parietal layers of the pericardium as distinct echogenic lines. There is little doubt that both air and blood may interfere with the ability to visualize fluid in the pericardial sac because of diminished sound transmission, loss of normal interfaces, and distortion of 576 | www.annalsofsurgery.com
No HT
D/C
FIGURE 3. Management algorithm for the screening of PCI in a hemodynamically stable patient. anatomy. A repeat US scan after 24 hours permitted a further 6 patients with hemopericardium to be identified, and this rescan should be considered in cases where clinical suspicion persists. The epicardial fat pad may occasionally be misdiagnosed as a pericardial effusion. However, the accuracy of US examination has also been shown to be dependent on US experience.29 The false-positive rate of US in our series was 19%. “The limitations of this study are the lack of routine C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
follow-up US in all patients with an initial negative scan, the fact that the experience in US may have varied between the radiologists, and the specific US probe selected was not documented.” The size of the pericardial effusion was found to be statistically significant in predicting the presence of hemopericardium (P = 0.03). The detection of any fluid (blood) in the pericardial sac after penetrating chest trauma is thought to be significant and warrant surgery. There is little mention in the literature about the volume of pericardial fluid or the measurement of the effusion. Some healthy patients do have a certain amount of pericardial fluid and this may potentially cause a false-positive result. Patel et al30 published a retrospective study on surgeon-performed sonography. Twenty patients were found to have a cardiac injury at surgery, and they described a true-positive result as the presence of at least 5 mm of fluid and a cardiac injury identified at surgery. There is no discussion as to where this 5-mm figure was derived from and what the significance is of having less than 5 mm of fluid. In our series, there were patients who had a positive result for SPW exploration and yet they had less than 5 mm of fluid present on the US scan. In 3 patients, the size of the pericardial effusion was noted to be 3 mm and in all of them, a cardiac injury was documented at sternotomy. As a result, it would appear that there is no minimum “safe” limit and the presence of any fluid in the pericardial sac should be an indicator of a PCI. There have been very few studies that have specifically investigated transthoracic echocardiography (ECHO) in the diagnosis of penetrating cardiac trauma. Meyer et al12 compared the sensitivity and specificity of ECHO against the gold standard of SPW exploration in 1995. In their prospective study, 105 hemodynamically stable patients with PCI in the “proximity of the heart” were included. Eighty-nine patients had a true-negative ECHO result, as performed and interpreted by the trauma surgeon. There were 9 cardiac injuries identified at surgery, with ECHO confirming 5 of these. They also had 7 falsepositives and 4 false-negatives. Their sensitivity of ECHO was 56%, with 93% specificity for detecting a cardiac injury. This study also highlighted the presence of a hemothorax having an adverse influence on the ability of ECHO to detect hemopericardium. In 1990, Jimenez et al31 published a prospective study comparing ECHO with SPW exploration in the diagnosis of the occult cardiac injury after penetrating chest trauma. Seventy-three patients were included, all having undergone ECHO and a subsequent SPW exploration. There were 9 cardiac injuries detected. The sensitivity of ECHO for diagnosing an injury was 90%, with a specificity of 97%; there was a single false-negative result. Grewal et al14 in 1995 reported on the evaluation of SPW exploration in the diagnosis of the occult cardiac injury. In that study, there were 2 patients with negative ECHO findings who had cardiac injuries. The one case presented with delayed cardiac tamponade after discharge, and the second with an episode of hypotension while still in the hospital. Bolton et al32 documented 5 cases where ECHO failed to detect a cardiac injury. There are numerous case reports of where ECHO has failed to detect the cardiac injury in asymptomatic patients.33 A 23-year-old man from Turkey was admitted 6 months after his initial injury with a traumatic aorto-left atrial fistula and features of cardiac tamponade. The initial clinical evaluation and ECHO on admission to hospital showed negative findings.34 From Korea, a 22-year-old woman presented with a stab wound to the chest with a left-sided hemothorax. The initial ECHO examination showed a small amount of fluid in the pericardial sac without any shunt or injury to the myocardium. It was only on follow-up ECHO performed the next day after clinical deterioration that a ventricular septal defect was demonstrated, which was repaired surgically.35 From Spain, a 32-year-old man with a precordial stab wound and normal ECHO examination on presentation, who 4 days later experienced a cardiac arrest due to delayed cardiac C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Antibiotic Prophylaxis in Clean Surgery
tamponade. Emergency surgery revealed a laceration to midsegment of the left anterior descending coronary artery that required suturing.36 There are 2 cases of cardiac tamponade in the United States that presented 4 and 14 days after chest stab wounds with initial normal admission evaluation of US, ECHO, and computed tomographic (CT) scan of the chest.37 It would appear that ECHO is as sensitive as US in the detection of the occult cardiac injuries. The major concern is the high number of false-negative results and consequently our view that US cannot be a stand-alone screening test for cardiac injuries. A further consideration is the availability of the technology in the trauma and emergency departments in hospitals and the available expertise to provide the service afterhours. CT is being used in many centers as the primary imaging modality in hemodynamically stable patients with penetrating chest trauma. This has been driven by the speed and excellent image reconstruction of multidetector CT.38 There is very little data on the role of CT in screening for a cardiac injury. Nagy et al39 published a retrospective study in 1996 on 60 stable patients reporting a sensitivity and specificity of 100% and 96.6%, respectively. Unfortunately, only 2 of the 60 patients had a confirmed cardiac injury. It does, however, seem logical to include CT in a screening protocol, particularly when the US result is equivocal, a hemothorax is present, or when clinical suspicion of a PCI is high. However, further studies on the sensitivity of CT are required and one does need to be aware of the radiation dose in young patients and the fact that US is more frequently available in the emergency department. It has been well described that despite skin entry wounds outside the previously mentioned surface anatomical landmarks, these injuries can cause PCI. In a 15-month postmortem study of PCI in Cape Town, 507 wounds in the thorax resulted in 240 cardiac injuries; of those with cardiac injuries, only 33% were precordial and 23% were from the posterior chest wall.40 Evans et al41 observed that cardiac injuries may occur whether the entrance wound is in the posterior chest wall or in the subcostal or subxiphoid abdominal region. There are numerous case reports in the literature that warn of cardiac injuries occurring from remote sites.33,42 A prospective study of 66 patients by Buckman et al15 showed that the probability of a cardiac injury varied according to the mechanism of trauma. They found that 80% of stab wounds with associated PCI were located precordially but only 46% of gunshots were located precordially. They felt that any gunshot wound to the torso carried the potential of a cardiac injury.
NEW SCREENING REGIMEN We would, therefore, like to propose that the high-risk area for cardiac stab wounds is the region extending from the right anterior axillary line anteriorly across the precordium to the posterior left chest ending in the midline at the spinous processes, with the superior margin being the supraclavicular areas and the inferior margin being the costal margins, including the upper epigastrium. The authors call this area the “cardiac zone” (Figs. 2A, B). Any gunshot wound to the neck, shoulders, chest, back, and abdomen should be considered to have a possible cardiac injury and should be screened appropriately. Screening in hemodynamically stable patients should include an erect chest radiograph if possible, an electrocardiogram, and an US scan of the pericardial sac. The screening US must be performed immediately on admission to the emergency department, as an occult PCI may cause rapid deterioration in the patient’s condition and is potentially fatal. If the screening US scan is positive for blood in the pericardial sac, then the patient must undergo SPW exploration. The presence of any fluid in the pericardial sac should be regarded as a positive US result as there is no particular measurement in millimeters that safely excludes a PCI. If the screening US scan is equivocal, then the patient should either undergo SPW exploration or a CT scan of chest. If the initial screening US scan shows negative result www.annalsofsurgery.com | 577
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 3, March 2015
Nicol et al
but there is a hemothorax, then a CT scan of the chest should be obtained immediately or the US scan repeated at 24 hours (Fig. 3). If the US scan shows negative result but there is clinical suspicion of a PCI due to the presence of J-waves on the electrocardiogram, or a straight left heart border on chest radiograph, then a CT chest scan should be obtained or else the US scan of the pericardial sac should be repeated at 24 hours. If there is pneumopericardium present on chest radiograph, US scan, or CT scan, then SPW exploration should be undertaken. The patient with a negative US result and no hemothorax, no pneumopericardium, or clinical suspicion of a PCI may be discharged home.
REFERENCES 1. Ivatury RR, Rohman M, Steichen FM, et al. Penetrating cardiac injuries: twenty-year experience. Am Surg. 1987;53:310–317. 2. Demetriades D, van der Veen BW. Penetrating injuries of the heart: experience over two years in South Africa. J Trauma. 1983;23:1034–1041. 3. Ma OJ, Mateer JR, Ogata M, et al. Prospective analysis of a rapid trauma ultrasound examination performed by emergency physicians. J Trauma. 1995;38:879–385. 4. Rozycki GS, Feliciano DV, Schmidt JA, et al. The role of surgeon-performed ultrasound in patients with possible cardiac wounds. Ann Surg. 1996;223:737– 746. 5. Rozycki GS, Ballard RB, Feliciano DV, et al. Surgeon-performed ultrasound for the assessment of truncal injuries: lessons learned from 1540 patients. Ann Surg. 1998;228:557–567. 6. Rozycki GS, Feliciano DV, Ochsner MG, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma. 1999;46:543–552. 7. Bokhari F, Nagy K, Roberts R, et al. The ultrasound screen for penetrating truncal trauma. Am Surg. 2004;70:316–321. 8. Tayal VS, Beatty MA, Marx JA, et al. FAST (focused assessment with sonography in trauma) accurate for cardiac and intraperitoneal injury in penetrating anterior chest trauma. J Ultrasound Med. 2004;23:467–472. 9. Ball CG, Williams BH, Wyrzykowski AD, et al. A caveat to the performance of pericardial ultrasound in patients with penetrating cardiac wounds. J Trauma. 2009;67:1123–1124. 10. Harris DG, Papagiannopoulos KA, Pretorius J, et al. Current evaluation of cardiac stab wounds. Ann Thorac Surg. 1999;68:2119–2122. 11. Cull JD, Bokhari F. Penetrating chest injury in the setting of a hemothorax: should pericardial windows replace echocardiograms? Am Surg. 2013;79: E56–E57. 12. Meyer DM, Jessen ME, Grayburn PA. Use of echocardiography to detect occult cardiac injury after penetrating thoracic trauma: a prospective study. J Trauma. 1995;39:902–909. 13. Asensio JA, Roldan G, Petrone P, et al. Cardiac trauma. Trauma. 2001;3:69–77. 14. Grewal H, Ivatury RR, Divakar M, et al. Evaluation of subxiphoid pericardial window used in the detection of occult cardiac injury. Injury. 1995;26:305–310. 15. Buckman RF, Jr, Badellino MM, Mauro LH, et al. Penetrating cardiac wounds: prospective study of factors influencing initial resuscitation. J Trauma. 1993;34:717–727. 16. Jones EW, Helmsworth J. Penetrating wounds of the heart. Thirty years’ experience. Arch Surg. 1968;96:671–682. 17. Demetriades D. Cardiac wounds. Experience with 70 patients. Ann Surg. 1986;203:315–317. 18. Degiannis E, Loogna P, Doll D, et al. Penetrating cardiac injuries: recent experience in South Africa. World J Surg. 2006;30:1258–1264. 19. Nicol AJ, Navsaria PH, Hommes M, et al. Sternotomy or drainage for a hemopericardium after penetrating trauma. A randomized controlled trial. Ann Surg. 2014;259:438–442.
578 | www.annalsofsurgery.com
20. Nicol AJ, Navsaria PH. The J-wave: a new electrocardiographic sign of an occult cardiac injury. Injury. 2014;45:112–115. 21. Nicol AJ, Navsaria PH, Beningfield S, et al. A straight left heart border: a new radiological sign of a hemopericardium. World J Surg. 2014;38: 211–214. 22. Boulanger BR, Brenneman FD, McLellan BA, et al. A prospective study of emergent abdominal sonography after blunt trauma. J Trauma. 1995;39:325– 330. 23. Kirkpatrick AW, Sirois M, Laupland KB, et al. Prospective evaluation of handheld focused abdominal sonography for trauma (FAST) in blunt abdominal trauma. Can J Surg. 2005;48:453–460. 24. Boulanger BR, McLellan BA, Brenneman FD, et al. Emergent abdominal sonography as a screening test in a new diagnostic algorithm for blunt trauma. J Trauma. 1996;40:867–874. 25. Dente CJ, Rozycki GS. Surgeon performed ultrasound in trauma and surgical critical care. In: Feliciano DV, Mattox KL, Moore EE, eds. Trauma. 7th ed. New York: McGraw-Hill Medical; 2013:301–321. 26. Gillman LM, Ball CG, Panebianco N, et al. Clinician performed resuscitative ultrasonography for the initial evaluation and resuscitation of trauma. Scand J Trauma Resusc Emerg Med. 2009;17:34. 27. Scalea TM, Rodriguez A, Chiu WC, et al. Focused assessment with sonography for trauma (FAST): results from an international consensus conference. J Trauma. 1999;46:466–472. 28. Kirkpatrick AW, Sirois M, Laupland KB, et al. Hypotension after blunt abdominal trauma: the role of emergent abdominal sonography in surgical triage. J Trauma. 1996;41:815–820. 29. Blaivas M, DeBehnke D, Phelan MB. Potential errors in the diagnosis of pericardial effusion on trauma ultrasound for penetrating injuries. Acad Emerg Med. 2000;7:1261–1266. 30. Patel AN, Brennig C, Cotner J, et al. Successful diagnosis of penetrating cardiac injury using surgeon-performed sonography. Ann Thorac Surg. 2003;76:2043– 2047. 31. Jimenez E, Martin M, Krukenkamp I, et al. Subxiphoid pericardiotomy versus echocardiography: a prospective evaluation of the diagnosis of occult penetrating cardiac injury. Surgery. 1990;108:676–680. 32. Bolton JW, Bynoe RP, Lazar HL, et al. Two-dimensional echocardiography in the evaluation of penetrating intrapericardial injuries. Ann Thorac Surg. 1993;56:506–509. 33. Pathi V, Sutherland FW, Ireland A, et al. Breaking the rules! Cardiac injury from remote entry sites. J Accid Emerg Med. 1998;15:115–116. 34. Belgi A, Mete A, Avsar O, et al. Late presentation of complex cardiac defect following penetrating cardiac trauma: case report. Turk Kardiyol Dern Ars. 2011;31:1–4. 35. Jeon K, Lim WH, Kang SH, et al. Delayed diagnosis of traumatic ventricular septal defect in penetrating chest injury: small evidence on echocardiography makes big difference. J Cardiovasc Ultrasound. 2010;18:28–30. 36. Enriquez SG, Fernandez CG, Entem FR, et al. Delayed pericardial tamponade after penetrating chest trauma. Eur J Emerg Med. 2005;12:86–88. 37. Simmons JD, Haraway AN, Schmieg RE, Jr, et al. Is there a role for secondary thoracic ultrasound in patients with penetrating injuries to the anterior mediastinum? Am Surg. 2008;74:11–14. 38. De Vries CS, Africa M, Gebremariam FA, et al. The imaging of stab injuries. Acta Radiol. 2010;51:92–106. 39. Nagy KK, Gilkey SH, Roberts RR, et al. Computed tomography screens stable patients at risk for penetrating cardiac injury. Acad Emerg Med. 1996;3:1024– 1027. 40. Oakland C, Vivian J. Penetrating cardiac injuries [Letter to the Editor]. BMJ. 1987;295:502. 41. Evans J, Gray LA, Jr, Rayner A, et al. Principles for the management of penetrating cardiac wounds. Ann Surg. 1979;189:777–784. 42. Claassen CW, O’Connor JV, Gens D, et al. Penetrating cardiac injury: think outside the box. J Trauma. 2010;68:E71–E73.
C 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
