|
|
||||||||
Emergency Radiology |
1 From the Division of Emergency Medicine (J.R.R., N.A.K.) and Department of Radiology (J.P.M.), University of California, Davis Medical Center, PSSB 2100, 2315 Stockton Blvd, Sacramento, CA 95817; and Department of Radiology, General Hospital of Chinese P.L.A., Beijing, China (L.W.). Received April 26, 2001; revision requested June 1; revision received July 30; accepted September 17. Address correspondence to J.R.R. (e-mail: jrrichards@ucdavis.edu).
| ABSTRACT |
|---|
|
|
|---|
MATERIALS AND METHODS: Imaging findings were recorded prospectively in 744 consecutive children who underwent emergency US from January 1995 to October 1998; free fluid and parenchymal abnormalities of specific organs were also noted. Patients with intraabdominal injuries were identified retrospectively. Computed tomographic (CT) findings, intraoperative findings, and clinical outcome were compared with the initial US findings. Sensitivity, specificity, and positive and negative predictive values were calculated for patients who underwent CT, laparotomy, or both after US.
RESULTS: Seventy-five (10%) of 744 patients had intraabdominal injuries, and US depicted free fluid in 42 of them. US had 56% sensitivity, 97% specificity, 82% positive predictive value, and 91% negative predictive value for detection of hemoperitoneum alone. US helped identify parenchymal abnormalities that corresponded to actual organ injury without accompanying free fluid in nine patients (12%). Inclusion of identification of parenchymal organ injury at US increased the sensitivity of US to 68%, with an accuracy of 92%.
CONCLUSION: US for blunt abdominal trauma in children is highly accurate and specific, but moderately sensitive, for detection of intraabdominal injury.
© RSNA, 2002
Index terms: Abdomen, CT, 70.12112, 70.12115, 80.12112, 80.12115 Abdomen, hemorrhage, 79.41 Abdomen, injuries, 70.41, 80.41 Abdomen, US, 70.1298, 80.1298 Children, injuries, 70.41, 80.41 Trauma, 70.41, 80.41
| INTRODUCTION |
|---|
|
|
|---|
Conversely, ultrasonography (US) is rapid, noninvasive, and relatively inexpensive. US has been described as an accurate method for detection of hemoperitoneum in adults, but less is known about its accuracy for detection of hemoperitoneum or specific organ injuries in children (1215). The purpose of this study was to assess the accuracy of emergency abdominal US in the detection of both hemoperitoneum and parenchymal organ injury in children.
| MATERIALS AND METHODS |
|---|
|
|
|---|
US was performed for the sole purpose of detecting intraabdominal injury from blunt trauma in 744 pediatric patients (350 boys [47%] and 394 girls [53%]; mean age, 10.1 years ± 4.9 [SD]; age range, 2 weeks to 16 years) who presented to the emergency department from January 1995 to October 1998. These did not include every pediatric patient who presented to the emergency department with blunt traumatic injury. Mechanisms of injury are detailed in Table 1.
|
US was performed by registered diagnostic medical sonographers soon after the patient arrived in the emergency department, usually within 30 minutes. Median time to complete US scanning was 5 minutes (range, 310 minutes). The left and right upper quadrants were scanned for the presence of free fluid, with attention paid to the splenorenal and hepatorenal areas. The parenchyma of both the liver and spleen were also evaluated to detect irregularities suggestive of hematomas and lacerations. The epigastrium was scanned to evaluate the left lobe of the liver and the pancreas. Both the right and left flanks were scanned to detect free fluid and evaluate the kidneys. The pelvis was then evaluated for free fluid. Renal perfusion was not assessed with Doppler US.
US was performed before CT, diagnostic peritoneal lavage, or laparotomy, and results were considered positive when any abnormality was detected that could have resulted from trauma: free fluid was assumed to represent hemoperitoneum, and irregularities within the parenchyma of solid organs or subcapsular collections of fluid were assumed to be lacerations or hematomas. True injury was confirmed by means of laparotomy or CT. The decision to perform CT or diagnostic peritoneal lavage or to proceed to surgery was made by the attending trauma surgeon on the basis of mechanism of injury, physical findings, vital signs, and US findings.
CT of the abdomen was performed (9800 or HiSpeed CTi; GE Medical Systems, Milwaukee, Wis) with either 7- or 10-mm incremental scans from diaphragm to pelvis. Iohexol (Omnipaque 300; Nycomed Amersham, Princeton, NJ) was administered intravenously at a dose of 1 mL per kilogram of body weight. Contrast material was not administered orally. No CT was performed later than 6 hours after the initial US examination.
All US studies were interpreted initially by the faculty, fellow, or resident radiologist on call. A final faculty interpretation was performed if the initial reading was performed by a resident or fellow. Images were printed and reviewed immediately. A data sheet that detailed the findings was completed at this time. All examination results were recorded for final interpretation by faculty radiologists. Radiologists did not perform the US examinations, and serial US examinations were not performed. Patients with intraabdominal injuries were identified retrospectively. Final reports were obtained for all patients who underwent CT or surgical intervention, and findings were compared with those on the prospective data sheets. Patients were followed up until discharge from the hospital, and results of return visits to the emergency department and trauma clinic were obtained by means of chart review (J.R.R., L.W.).
Fifteen exploratory laparotomies and five diagnostic peritoneal lavages were performed directly after US, whereas 24 laparotomies and one diagnostic peritoneal lavage were performed after both US and CT. Three hundred ninety patients underwent CT after initial US. Thus, 39 injuries were confirmed by means of laparotomy, and the remaining 36 were confirmed by means of CT.
Data are reported as the mean ± SD unless otherwise stated. Sensitivity, specificity, positive and negative predictive values, and 95% CIs were calculated for patients who underwent CT, laparotomy, or both.
| RESULTS |
|---|
|
|
|---|
|
|
|
|
|
|
|
|
|
|
| DISCUSSION |
|---|
|
|
|---|
Partrick and co-workers (18) determined in their prospective study of 230 children that emergency US obviated further CT in a certain subset of patients in stable condition, which resulted in a substantial total cost saving. Coley and colleagues (19) compared US with CT for detection of hemoperitoneum from blunt injury in 107 pediatric patients. They did not search for parenchymal abnormalities. This group determined that the sensitivity and specificity of US relative to CT were 55% and 83%, respectively, and they concluded that US was not accurate enough to replace CT in this setting. In our study, sensitivity was similar for detection of free fluid, but specificity was higher.
Luks et al (20) examined 259 children admitted for blunt trauma who underwent US, and intraabdominal injury was detected in 81. Their calculated sensitivity, specificity, and accuracy for detection of hemoperitoneum, but not parenchymal injury, were 89%, 96%, and 94%, respectively. In three of nine patients with initial negative US results, intraabdominal injury was detected at repeat US.
Thourani and associates (21) analyzed results at surgeon-performed US in 192 children with blunt abdominal trauma; free fluid was detected in eight. There were two were false-positive results and no false-negative results in their study. Sensitivity and specificity were 80% and 100%, respectively. A smaller study with different results was published by Mutabagani and colleagues (22), who performed focused abdominal sonography for trauma, or FAST, in 46 children. Sensitivity and specificity were 30% and 100%, respectively, and there were nine false-negative US results. They concluded that US could not be relied on as a screening examination for intraabdominal injury in children.
In our study, nine patients had false-positive US results with the finding of minimal to moderate free fluid in the pelvis. This may represent physiologic free fluid, especially since seven of the nine patients were girls. Holmes et al (23) determined only 17% of children with isolated free fluid detected at CT had intraabdominal injury in their series of 527 patients.
A limitation in many of these studies has been the use of clinical observation to determine patient outcome. Patients discharged after negative US findings and no further imaging or procedures were considered to have true-negative results. Some of these patients may have had minor injuries that were missed at US, and this would affect the sensitivity and specificity calculated for US. These values may change with a higher or lower frequency of injuries than the 10% in our study. One limitation in our study is that the decision to clinically observe the patient, perform CT after US, or proceed to surgery was often arbitrary and subjective; potential bias regarding accuracy of US exists. One way to avoid this bias would be to conduct a randomized blinded study, which may be difficult to justify ethically. In our analysis, we chose to omit those patients who were observed clinically after US but did not undergo CT, laparotomy, or both.
Solid organ injuries may result in hemoperitoneum that may not be detectable at US. These injuries may be depicted at US as aberrations of the normal parenchymal architecture of such organs as the liver, spleen, and kidney. Hematomas may be identified as mixed echogenic or, less commonly, cystic areas in a subcapsular or intraparenchymal distribution (24). Certain injuries, such as subcapsular hematomas or bowel perforations, may not result in appreciable hemoperitoneum and may be missed at US.
Detection of solid organ injury requires more effort and skill than simply searching for free fluid. Other authors have addressed the potential of US for the detection of actual solid organ injuries. Krupnick and associates (25) compared CT with US in a blinded study, and 12 of 32 pediatric splenic injuries were missed at US, with seven injuries having no associated free fluid. Richards and colleagues (26) determined the most commonly identified parenchymal abnormality for hepatic injury was a discrete hyperechoic focus, followed by a diffuse hyperechoic pattern. In their study, US had a sensitivity of 98% for detection of high-grade hepatic injuries.
Scanning solely for hemoperitoneum will lead to missing a number of intraabdominal injuries. Shanmuganathan and co-workers (27) identified a group of patients with visceral injury, and more than a quarter of them had no hemoperitoneum detected at screening US. In addition to helping scrutinize the parenchyma of the liver, spleen, and kidney for the detection of abnormalities, serial US over time may help in detection of hemoperitoneum as the peritoneal cavity fills with blood. Siniluoto et al (28) detected two of five splenic injuries with serial US, and Henderson and co-workers (29) identified four patients whose initial US results were negative for hemoperitoneum and later became positive at serial examinations. In our study, we did not perform serial US, but this may have improved the overall sensitivity of US.
US after blunt trauma has been demonstrated to be cost-effective in several studies (12,16,17) and has essentially replaced diagnostic peritoneal lavage as a tool for detection of intraabdominal injury. Further developments include the use of special US contrast materials to enhance parenchymal abnormalities and small hand-held US units that can easily fit in the clinicians pocket (30,31).
We conclude that US is accurate and specific for detection of intraabdominal injury but is only moderately sensitive and that searching for parenchymal solid organ abnormalities appears to improve the sensitivity of US. As more experience is gained in the use and interpretation of US studies, it is likely that there will be greater interest in the detection of actual organ injuries in addition to free fluid. On the basis of the results in the current study and in prior studies, we believe that emergency US might be used as a triage tool for patients with blunt abdominal trauma whose condition is unstable. We believe that those patients in unstable condition with free fluid, parenchymal organ injury, or both detected at US could be triaged immediately to surgery. We also believe that patients in stable condition with free fluid or parenchymal injury detected at US should undergo CT unless their condition deteriorates. Those with negative US results may undergo clinical observation or CT at the discretion of the clinician, on the basis of physical and laboratory findings.
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J G Browning, A G Wilkinson, and T Beattie Imaging paediatric blunt abdominal trauma in the emergency department: ultrasound versus computed tomography Emerg. Med. J., October 1, 2008; 25(10): 645 - 648. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Levy and V. E. Noble Bedside Ultrasound in Pediatric Emergency Medicine Pediatrics, May 1, 2008; 121(5): e1404 - e1412. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Christie-Large, D. Michaelides, and S. James Focused assessment with sonography for trauma: the FAST scan Trauma, April 1, 2008; 10(2): 93 - 101. [Abstract] [PDF] |
||||
![]() |
M. Valentino, C. Serra, P. Pavlica, A. M. Morselli Labate, M. Lima, S. Baroncini, and L. Barozzi Blunt Abdominal Trauma: Diagnostic Performance of Contrast-enhanced US in Children--Initial Experience Radiology, March 1, 2008; 246(3): 903 - 909. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Beck-Razi, D. Fischer, M. Michaelson, A. Engel, and D. Gaitini The Utility of Focused Assessment With Sonography for Trauma as a Triage Tool in Multiple-Casualty Incidents During the Second Lebanon War J. Ultrasound Med., September 1, 2007; 26(9): 1149 - 1156. [Abstract] [Full Text] [PDF] |
||||
![]() |
O. Catalano, R. Lobianco, M. M. Raso, and A. Siani Blunt Hepatic Trauma: Evaluation With Contrast-Enhanced Sonography: Sonographic Findings and Clinical Application J. Ultrasound Med., March 1, 2005; 24(3): 299 - 310. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Murphy, A. Ghosh, and K Mackway-Jones Ultrasound or computed tomography in paediatric blunt abdominal trauma Emerg. Med. J., November 1, 2002; 19(6): 554 - 556. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |