Accuracy of Point-of-Care Ultrasound for Diagnosis ofSkull Fractures in Children
WHAT’S KNOWN ON THIS SUBJECT: Head injuries and concern forskull fracture are common in pediatrics. Point-of-care ultrasoundis an imaging tool that can be used to diagnose fractures.However, there are scant data regarding the accuracy of point-of-care ultrasound in skull fracture diagnosis.
WHAT THIS STUDY ADDS: Clinicians with focused point-of-careultrasound training are able to diagnose skull fractures inchildren with high specificity. Ultrasound may be valuable todiagnose skull fractures in children at the point of care.
abstractOBJECTIVE: To determine the test performance characteristics forpoint-of-care ultrasound performed by clinicians compared withcomputed tomography (CT) diagnosis of skull fractures.
METHODS: We conducted a prospective study in a convenience sampleof patients #21 years of age who presented to the emergency de-partment with head injuries or suspected skull fractures that re-quired CT scan evaluation. After a 1-hour, focused ultrasoundtraining session, clinicians performed ultrasound examinations toevaluate patients for skull fractures. CT scan interpretations by at-tending radiologists were the reference standard for this study. Point-of-care ultrasound scans were reviewed by an experienced sonologistto evaluate interobserver agreement.
RESULTS: Point-of-care ultrasound was performed by 17 clinicians in69 subjects with suspected skull fractures. The patients’mean age was6.4 years (SD: 6.2 years), and 65% of patients were male. Theprevalence of fracture was 12% (n = 8). Point-of-care ultrasoundfor skull fracture had a sensitivity of 88% (95% confidence interval[CI]: 53%–98%), a specificity of 97% (95% CI: 89%–99%), a positivelikelihood ratio of 27 (95% CI: 7–107), and a negative likelihood ratio of0.13 (95% CI: 0.02–0.81). The only false-negative ultrasound scan wasdue to a skull fracture not directly under a scalp hematoma, butrather adjacent to it. The k for interobserver agreement was 0.86(95% CI: 0.67–1.0).
CONCLUSIONS: Clinicians with focused ultrasound training were ableto diagnose skull fractures in children with high specificity. Pediatrics2013;131:e1757–e1764
AUTHORS: Joni E. Rabiner, MD,a Lana M. Friedman, MD,b
Hnin Khine, MD,a Jeffrey R. Avner, MD,a and James W.Tsung, MD, MPHb
aDivision of Pediatric Emergency Medicine, Department ofPediatrics, Children’s Hospital at Montefiore/Albert EinsteinCollege of Medicine, Bronx, New York; and bDivision of PediatricEmergency Medicine, Department of Emergency Medicine, MountSinai School of Medicine/Mount Sinai Medical Center, New York,New York
KEY WORDSultrasound, skull fracture, head trauma, pediatrics, emergencymedicine
ABBREVIATIONSCI—confidence intervalCT—computed tomographyED—emergency departmentPEM—pediatric emergency medicine
Dr Rabiner conceptualized and designed the study, collecteddata, performed statistical analyses, drafted the manuscript,and approved the final manuscript as submitted; Dr Friedmanconceptualized and designed the study, collected data,performed initial statistical analyses, drafted the initialmanuscript, and approved the final manuscript as submitted;Drs Avner and Khine co-supervised the overall conduct of thestudy at 1 of the 2 sites, critically reviewed the manuscript, andapproved the final manuscript as submitted; and Dr Tsungconceptualized and designed the study, supervised the overallconduct of the study at 1 of the 2 sites, assisted with data andstatistical analysis, critically reviewed the manuscript, andapproved the final manuscript as submitted.
www.pediatrics.org/cgi/doi/10.1542/peds.2012-3921
doi:10.1542/peds.2012-3921
Accepted for publication Mar 6, 2013
Address correspondence to Joni E. Rabiner, MD, Children’sHospital at Montefiore, Albert Einstein College of Medicine, 3415Bainbridge Ave, Bronx, NY 10467. E-mail: [email protected]
Head trauma is one of the most com-mon childhood injuries, accounting for.600 000 emergency department (ED)visits, 60 000 hospitalizations, and 6000deaths in children annually in theUnited States.1 It is estimated that 16%of children with nontrivial head inju-ries may have skull fractures, and thepresence of a skull fracture is associ-ated with a fourfold increased risk ofan underlying intracranial injury.2 Thegold standard diagnostic test to eval-uate for skull fracture and intracranialhemorrhage is head computed to-mography (CT), which is highly sensi-tive for identification of children withintracranial injuries requiring acuteintervention.3 However, CT imaging ex-poses developing brains to ionizingradiation4–7 and may require sedationin young children. Clinicians caring forchildren need to decide, on the basisof the risks and benefits, whether toperform a head CT in a child presentingwith a closed head injury.
Point-of-care ultrasound is an imagingmodality used by a variety of medicalspecialties,8–10 and it is widely acceptedas a diagnostic tool for use in the ED.11
Multiple studies show that ultrasoundfor fracture diagnosis has good accu-racy when used by clinicians12–15 aswell as when used by radiologists.16,17
In addition, ultrasound is well toleratedby children, even in areas of injury.12,15
Emerging data suggest that ultrasounddiagnosis of skull fractures in childrenis promising,13,15,17–20 and additional in-vestigation of its utility is warranted.Our principal objectivewas to determinethe test performance characteristics forpoint-of-care ultrasound performed byclinicians compared with CT scan for thediagnosis of skull fractures in children.
METHODS
Study Design and Setting
This was a prospective observationalstudy conducted from September 2010toMarch 2012 in 2 urban, level II trauma
center pediatric EDs with a combinedannual census of 100 000 patients. Aconvenience sample of patients #21years of age with head injuries re-quiring CT scan for suspected fractureand/or intracranial injury, who pre-sented when a trained study physicianwas available, was eligible for inclusionin this study. Written informed consentwas obtained from the patient or parent/guardian, and written assent was ob-tained frompatients$7 years of age. Thisstudy was approved by the hospitals’ in-stitutional review boards, and it adheredto the STARD (Standards for Reportingof Diagnostic Accuracy) criteria for re-search involving diagnostic accuracy.21
The methods were similar to those thathave been published elsewhere.15,22
Selection of Participants
Inclusion criteria included patient age of#21 years with a head trauma and/orsuspected skull fracture requiring radio-graphic evaluationwith a head CT scan asrecommended by the treating pediatricemergency medicine (PEM) physician.Patients were excluded if they presentedwith completed radiologic studies, a con-firmedskull fracture, an open fracture, orif urgent intervention was required.
Ultrasound Technique
Before the start of the study, all en-rolling PEM attending and fellow physi-cians attended a 30-minute didacticsession to learn how to use ultrasoundto evaluate the skull for fracture andto standardize the method in which
bedside ultrasound was performed byparticipating physicians, followed bya 30-minute hands-onpractical session.A reference manual complete with in-structions and images was availablethroughout the study. All study sonol-ogists except for one were novices tomusculoskeletal ultrasoundat the startof the study. We defined an experiencedsonologist as having performed $25musculoskeletal ultrasound examina-tions, which is the minimum recom-mended number of scans for ultrasoundcredentialing per American College ofEmergency Physicians Emergency Ul-trasound Guidelines.11
SonoSite ultrasound systems (SonoSiteInc, Bothell, WA) with high-frequencylinear transducer probes (10–5 MHz)were used to perform focused ultra-sound examinations to evaluate forskull fracture. Ultrasound gel was lay-ered onto the ultrasound probe, andthen the probe was lightly applied tothe scalp to avoid pressure on the in-jured skull. The transducer was placedover the area of soft tissue swelling,hematoma, point of impact, or pointof maximal tenderness (Fig 1). Scanswere performed in 2 perpendicularplanes, and still pictures and videoclips were recorded in each orienta-tion. Skull suture lines were differen-tiated from skull fractures by followingsuspected sutures to a fontanelle. Ifa suspected fracture crossed a sutureline or fontanelle, the contralateral areaon the skull was imaged for comparison.
FIGURE 1A, Linear transducer probe placement for skull ultrasound. B, Corresponding ultrasound image withskull fracture.
e1758 RABINER et alby guest on February 16, 2016Downloaded from
The sagittal, coronal, and metopicsutures can be traced to the anteriorfontanelle, and the lambdoid suturescan be traced to the posterior fonta-nelle. The squamous sutures, however,may be difficult to follow to an openfontanelle, but sonologists were en-couraged to scan the contralateralarea of the skull for comparison. A di-agram of suture anatomy was includedin the study reference manual.
Enrollment Protocol
Before performing the point-of-careultrasound, enrolling PEM physiciansfilledoutdatacollection formstorecordclinical characteristics, including scalphematoma and location, loss of con-sciousness, vomiting, altered mentalstatus or a Glasgow coma scale score,15, and/or palpable skull fracture.The PEM physician also determinedand recorded his or her impression ofthe clinical likelihood of skull fracturebefore the ultrasound (#1%, 2%–25%,26%–50%, 51%–75%, 76%–98%, or$99%).
A positive skull ultrasound was definedas the enrolling PEM physician’s de-termination of a cortical disruption orirregularity visualized on the point-of-care ultrasound (Fig 1B). The enrollingsonologist recorded the point-of-careultrasound findings (positive or nega-tive for skull fracture) on the datacollection sheet immediately after theprocedure and before reviewing anyradiographic imaging studies. All testperformance characteristics were an-alyzed on the basis of the enrolling PEMphysician’s determination of the pres-ence or absence of skull fracture.
A PEM physician with expertise in ul-trasonography (J.W.T.), who has .10years of point-of-care ultrasound clini-cal and teaching experience, reviewedall recorded ultrasound scans to pro-vide a measure of agreement and toclassify diagnostic errors made byenrolling PEM physicians. The expert
PEM sonologist was blinded to thepatient’s clinical findings, the enrollingsonologist’s ultrasound interpretation,and radiographic imaging. The time toperform the point-of-care ultrasoundwas determined from the time stampson the first and last images recordedfor each patient.
After completion of the point-of-careultrasound, all patients received ahead CT as per the discretion of thetreating PEM attending physician. Thegold standard for skull fracture wasdefined as “fracture” or “cortical ir-regularity” as documented in the at-tending radiologist’s report of the headCT. The radiologists were blinded to thepoint-of-care ultrasound examinationresults. Patients without definite frac-ture on CT scan in the ED received astructured telephone follow-up at least1 week after the initial ED visit to as-certain outcomes.
Our primary outcomewas to determinethe test performance characteristicsof point-of-care ultrasound for skullfracture performed and interpreted bytrained PEM physicians compared withthe diagnosis of fracture on CT scanwith clinical follow-up. Our secondaryobjectives were to compare interob-server agreement between enrollingPEM physicians and an expert PEMsonologist and to compare skull frac-ture with clinical assessment, findings,and follow-up. Last, we combined ourdata with published studies that usedsimilar methodology and performeda pooled-analysis for accuracy of point-of-careultrasound fordiagnosisof skullfracture in children.
Statistical Analysis
Data were analyzed by using SPSSStatistics (IBM, Armonk, NY) and aredescribed by using sensitivity, speci-ficity, positive and negative predictivevalues, positive and negative likelihoodratios, and 95% confidence intervals(CIs). Descriptive statistical analyses
were used for categorical data. k Val-ues were calculated as a measure ofinterobserver agreement.
By using the method of Arkin andWachtel,23 a sample size of 60 patientswould be needed to obtain a 95% CI (SD:5%) with an estimated 96% specificityfor ultrasound diagnosis of skull frac-tures based on the study by Weinberget al.15
RESULTS
Sixty-nine patients with a mean age of6.4 years (SD: 6.2 years; range: 7 days to21 years) were enrolled. Patient de-mographic and clinical information ispresented in Tables 1 and 2. The studyflowchart is presented in Fig 2.
Skull fracturewaspresent onCT scan in8 (12%) patients. The test performancecharacteristics for point-of-care ultra-sound diagnosis of skull fracturescomparedwith CT imagingwith 95%CIsand k values for interobserver agree-ment between enrolling physiciansand an experienced PEM sonologist arepresented in Table 3. The diagnostictest results for each point-of-care ul-trasound performed compared withthe reference standard imaging testand the interobserver agreement be-tween enrolling PEM physicians andthe expert PEM sonologist for each point-of-care ultrasound are shown in Fig 3.
TABLE 1 Patient DemographicCharacteristics
n (%)
Male 45 (65)Scalp hematoma 43 (62)Frontal 9 (13)Temporal 8 (12)Temporal and parietal 2 (3)Parietal 11 (16)Parietal and occipital 1 (1)Occipital 11 (16)Location not noted 1 (1)
Loss of consciousness 9 (13)Vomiting 22 (32)GCS ,15 or altered mental status 8 (12)Palpable fracture 4 (6)
N = 69. GCS, Glasgow coma scale.
ARTICLE
PEDIATRICS Volume 131, Number 6, June 2013 e1759by guest on February 16, 2016Downloaded from
Seventeen PEM physicians performedamean of 4.1 ultrasound scans (SD: 4.0;range: 1–13 scans) and a median of 2scans each (interquartile range: 1–7).It took PEM physicians a median of 68seconds (interquartile range: 39–148seconds) to obtain the skull ultrasoundimages necessary to make a diagnosis.One PEM physician, who enrolled 12patients, had previous experience inskull ultrasound before the start of thestudy, and this PEM physician did notenroll any patients with skull fractureaccording to ultrasound or CT scan.
Of the 4 (6%) patientswhohad reportedpalpable skull fractures on physicalexamination before ultrasound or ra-diographic imaging, 2 patients hadparietal skull fractures, 1 patient hadnegative ultrasound and CT imagingstudies,and1patienthadanultrasoundpositive for fracture as determined bythe enrolling and expert PEM physicianand a CT scan that was negative forfracture.
The anatomic locations of the skullfractures were as follows: 5 parietal(7%), 1 frontal (1%), 1 temporal/parietal(1%), and1parietal/occipital (1%). Four(6%) patients had intracranial hemor-rhage on CT: 2 (3%) with epidural he-matoma, both of whom had associatedskull fractures; 1 (1%) with a subduralhematoma; and 1 (1%) with subarach-noid hemorrhage not associated withskull fractures. On telephone follow-up, there was no change in clinicalstatus among patients who had nega-tive imaging studies at the initial EDvisit. Seven patients (10%) did nothave telephone follow-up after the
initial ED visit. However, all 7 of thesepatients had a negative ultrasoundand head CT for fracture at the initialED visit; these patients were includedin the analysis categorized as “fractureabsent.”
By using point-of-care ultrasound, frac-ture was diagnosed by the enrollingsonologist in 9 patients (13%). Overall,there were 3 (4%) discordant resultsbetween point-of-care ultrasound andradiographic imaging, with 1 false-negative result and 2 false-positiveresults. The false-negative case wasa 7-month-oldmalewho presentedwitha temporal scalp hematoma after headtrauma. The clinical likelihood of skullfracture before the ultrasound was26% to 50%. Point-of-care ultrasoundwas performed over the scalp hema-toma, and no skull fracture was visu-alized by the enrolling PEM physicianor the expert PEM sonologist. On CTscan, the patient was found to havea parietal nondepressed skull fractureadjacent to but not directly under-neath the scalp hematoma (Fig 4).This patient was admitted for obser-vation and did not require any addi-tional intervention.
The first false-positive case was due toan error by the PEM physician early inthe study. The PEM physician inter-preted the ultrasound as positive andthe expert PEM sonologist interpretedthe ultrasound as negative for fracture(Fig 5A); the CT scan was also negativefor fracture. The second false-positivecase was a minimally displaced skullfracture in the temporal fossa that wasvisualized on ultrasound. This skullfracture was confirmed on clip reviewby the expert PEM sonologist but notvisualized by CT scan (Fig 5B).
Combining our data with the results ofother published studies15,18,19 for a totalof 185 patients, the pooled fracturerate was 27%. Skull ultrasound forfractures had a combined sensitivity of94% and a specificity of 96% (Table 4).
FIGURE 2STARD (Standards for Reporting of Diagnostic Accuracy) flowchart.
e1760 RABINER et alby guest on February 16, 2016Downloaded from
DISCUSSION
We have demonstrated in the largestcohort of patients to date that witha 1-hour, focused musculoskeletal ultra-sound training session, novice sonolo-gists are able to quickly and accuratelydiagnose skull fractures with highspecificity. Previous data on ultrasoundby radiologists for skull fracture di-agnosis revealed high accuracy.17,20
In addition, studies of ultrasound byclinicians with focused training have
also revealed rapid and accurate di-agnosis of skull fractures with point-of-care ultrasound.15,18,19 In our study,as with most ultrasound applications,the specificity was higher than the sen-sitivity (Table 3).
Clinical assessment may not be com-pletely reliable for predicting skullfractures and intracranial injuries inchildren.24 In our data, 2 of 39 (5%)patients assessed to have a 2% to 25%likelihood of fracture and 3 of 9 (33%)
assessed to have a 26% to 50% likeli-hood of fracture after obtaining thehistory and physical examination hadconfirmed skull fractures (Table 2). Inaddition, of the 4 patients in our studywho had reported palpable skull frac-tures on physical examination, only 2(50%) had confirmed skull fracture byCT scan.
In current practice, head CT serves asthe gold standard diagnostic test toevaluate for skull fractures and in-tracranial bleeding after head trauma.However, there are several advantagesof using point-of-care ultrasound inthe detection of skull fractures. First,ultrasound can be performed rapidly,which can allow earlier detectionof skull fracture as a marker forsuspected intracranial injury and neu-rosurgical consultation. Second, point-of-care ultrasound has the potential toreduce CT use and ionizing radiationexposure in children. The estimatedlifetime risk of cancer from a head CT issubstantially higher for children thanfor adults because of a longer latencyperiod and the greater sensitivity ofdeveloping organs to radiation.4–7 How-ever, intracranial injury may occur with-out skull fracture, and clinicians mustuse clinical judgment or decisionrules25–28 for obtaining CT scan re-gardless of the presence or absence ofskull fracture. In addition, ultrasoundcan also be performed in young chil-dren without the need for sedation.
Point-of-care ultrasound for skull frac-turesmay be especially useful in placeswithout access to CT scan. It has beenestimated by the World Health Orga-nization that up to two-thirds of theworld’s population does not have ac-cess to diagnostic imaging technol-ogy,29 and portable ultrasound may be
TABLE 3 Test Performance Characteristics for Point-of-Care Ultrasound Diagnosis of Skull Fractures
N Fractures, n (%) Sensitivity, % Specificity, % PPV NPV LR+ LR2 k
Data are test performance characteristics (95% CI). LR+, likelihood ratio of a positive test; LR2, likelihood ratio of a negative test; NPV, negative predictive value; PPV, positive predictive value.
FIGURE 3A, Diagnostic results for eachpoint-of-careultrasoundperformedbyaPEMphysician comparedwith thereference standard of head CTwith clinical follow-up. B, Agreement between enrolling PEM physiciansand an experienced PEMsonologist for each point-of-care ultrasound performed. Each block representsa unique patient enrolled by a PEM physician sonographer (x-axis) with the number of ultrasound scansperformed by each PEM physician (y-axis), and each block is color-coded to show the test resultcompared with the reference standard (A) or the agreement with the expert sonologist (B). The blocksare arranged vertically in chronological order, with the first ultrasound scan at the bottom. Thesensitivity and specificity (A) and k values for interobserver agreement (B) for point-of-care skullultrasound are given for the first, second, third, fourth, and fifth to thirteenth ultrasound scansperformed by physicians. Neg, negative; Pos, positive; Sens, sensitivity; Spec, specificity.
ARTICLE
PEDIATRICS Volume 131, Number 6, June 2013 e1761by guest on February 16, 2016Downloaded from
implemented in these resource-scarcelocations.30 In addition, ultrasoundmaybe useful for triage in mass casualtydisasters31 or in austere environments.32
Last, ultrasound may be used in pedia-tricians’offices or in urgent care centersfor patients with suspected isolated skullfracture without ready access to CT scan.
Ultrasound may diagnose minimally ornondisplaced skull fractures that canbe missed on CT scan. Recent researchhas revealed that ultrasound has su-perior sensitivity to radiography incertain types of fractures,33 and it hasbeen shown to detect nondisplacedfractures as small as 1 mm.34 Ourstudy included a case of a 16-year-oldmale who presented with a boggyfrontal scalp hematoma after an as-sault. Skull ultrasound performed bya novice sonologist was interpretedas positive for fracture and confirmedon expert review (Fig 5B). The CT was
read as negative for skull fracture, andthe patient was discharged from theED. On telephone follow-up, the patientwas asymptomatic.
Knowledge of suture anatomy is es-sential in performing ultrasound ex-aminations of infant skulls.18,19 A sutureappears symmetric and regular andleads to a fontanelle, whereas a frac-ture is jagged andmay be displaced. Allenrolling sonologists in our study weretaught to differentiate sutures fromskull fractures by following suturesto a fontanelle. If a suspected fracturecrossed a suture or fontanelle, thecontralateral area of the skull wasimaged for comparison. No errors inour study were due to sutures.
There have been several recent studiespublished on ultrasound for diagnosisof skull fractures in children that in-volvedsmallsamplesizesofchildren.15,18,19
Our study adds the largest cohort tothe current literature. In addition, pool-ing our data with these similar studiesto form a cohort of 185 patients revealsultrasound to be highly sensitive andspecific for diagnosing skull frac-tures in children (Table 4). The studyby Weinberg et al15 looked at fracturedetection for all bones and included asmall subset of patients with suspectedskull fracture. In the study by Riera andChen,19 few enrolling sonologists withno formalized skull ultrasound trainingperformed skull ultrasound. Parri et al18
reported a very high prevalence of skull
fracture because they enrolled patientswith localizing evidence of trauma. How-ever, all of these studies used cliniciansonologists who performed blindedpoint-of-care ultrasound imaging andcompared skull ultrasound with CT asthe reference standard.
Skull ultrasound may be particularlyuseful in well-appearing patients withsuspected isolated skull fracture on thebasis of history and physical examina-tionand lowrisk forclinically importanttraumatic brain injury. The questionremains whether the absence of skullfracture on ultrasound in selectedpatients with head injury in the pres-ence of single isolated risk factors forintracranial bleeding can obviate theneed for CT scan. Two children in ourstudy, one with isolated scalp hematomaand another with isolated loss of con-sciousness, had no skull fracture detec-ted on ultrasound or CT scan but weresubsequently found to have intracranialhemorrhage. Thus, caution is warrantedin using ultrasound to rule out intra-cranial injury, and additional research isneeded to fully answer this question.
Our study has several limitations. Ourstudy population consisted of a conve-nience sample of patients enrolledwhen a trained physicianwas available,but the prevalence of skull fracturesof 12% in our study is similar toother studies.15,19,24 Ultrasound is anoperator-dependent modality, but be-cause a novice group of sonologistswas trained to perform skull ultrasoundwith such high specificity, we believethat our results may be generalizable toother clinicians with focused training.Last, there was a limitation in our ul-trasound scanning technique. Our onlyfalse-negative result was due to a skullfracture that was adjacent to but notdirectly beneath the scalp hematoma,and therefore this fracture was missedon ultrasound but confirmed on CT scan(Fig 4). We now recommend scanningthe areas around the scalp hematoma
FIGURE 4Coronal CT scan of a skull fracture adjacent to scalphematoma missed on point-of-care ultrasound.
FIGURE 5A, Negative skull ultrasound (thick arrow) incorrectly interpreted as a fracture (thin arrow) by the PEMphysician. B, Skull fracture (arrow) visualized on point-of-care ultrasound and not detected by head CT.
e1762 RABINER et alby guest on February 16, 2016Downloaded from
if a skull fracture is not visualized di-rectly beneath it, similar to the methodproposed by Riera and Chen.19
CONCLUSIONS
Clinicians with focused, point-of-care ul-trasound training were able to diagnoseskull fractures in children with headtrauma with high specificity and highnegative predictive value. In addition, al-most perfect agreement was observed
between novice and experienced sonol-ogists. Pooled analysis of publishedstudies for skull fracture reveals highspecificities with variable sensitivities.Future research is needed to determineif ultrasound can reduce the use of CTscans in children with head injuries.
ACKNOWLEDGMENTSWe thank our sonologists for patientrecruitment and enrollment into the
study: Daniel Fein,MD; Inna Elikashvili,DO; Rene Forti, MD; Sylvia Garcia, MD;Charles Murphy, MD; Lorraine Ng,MD; Audrey Paul, MD; Ruby Rivera,MD; Catherine Sellinger, MD; VaishaliShah, MD; Cary Siegel, MD; LouisSpina, MD; Christopher Strother, MD;Ee Tay, MD; and Adam Vella, MD. Wealso thank Romita Almonte, MD, forher assistance with data collectionand management.
REFERENCES
1. Faul M, Xu L, Wald MM, Coronado VG.Traumatic Brain Injury in the United States:Emergency Department Visits, Hospital-izations and Deaths, 2002-2006. Atlanta, GA:Centers for Disease Control and Pre-vention, National Center for Injury Pre-vention and Control; 2010
2. Quayle KS, Jaffe DM, Kuppermann N, et al.Diagnostic testing for acute head injury inchildren: when are head computed tomog-raphy and skull radiographs indicated?Pediatrics. 1997;99(5). Available at: www.pediatrics.org/cgi/content/full/99/5/e11
3. Schutzman SA, Greenes DS. Pediatric minorhead trauma. Ann Emerg Med. 2001;37(1):65–74
4. Brenner DJ, Hall EJ. Computed tomography—an increasing source of radiation expo-sure. N Engl J Med. 2007;357(22):2277–2284
5. Brenner DJ. Estimating cancer risks frompediatric CT: going from the qualitative tothe quantitative. Pediatr Radiol. 2002;32(4):228–231; discussion 242–244
6. Frush DP. Radiation, CT, and children: thesimple answer is . . . it’s complicated. Ra-diology. 2009;252(1):4–6
7. Brody AS, Frush DP, Huda W, Brent RL;American Academy of Pediatrics Section onRadiology. Radiation risk to children fromcomputed tomography. Pediatrics. 2007;120(3):677–682
8. Ma OJ, Mateer JR, Blavias M. EmergencyUltrasound. 2nd ed. New York, NY: McGrawHill Publishing; 2008
9. Kendall JL, Hoffenberg SR, Smith RS. His-tory of emergency and critical care ultra-sound: the evolution of a new imagingparadigm. Crit Care Med. 2007;35(suppl 5):S126–S130
10. Moore CL, Copel JA. Point-of-care ultraso-nography. N Engl J Med. 2011;364(8):749–757
11. American College of Emergency Physicians.Emergency ultrasound guidelines. AnnEmerg Med. 2009;53(4):550–570
12. Chien M, Bulloch B, Garcia-Filion P, YoussfiM, Shrader MW, Segal LS. Bedside ultra-sound in the diagnosis of pediatric claviclefractures. Pediatr Emerg Care. 2011;27(11):1038–1041
13. Hübner U, Schlicht W, Outzen S, Barthel M,Halsband H. Ultrasound in the diagnosis offractures in children. J Bone Joint Surg Br.2000;82(8):1170–1173
14. Patel DD, Blumberg SM, Crain EF. The utilityof bedside ultrasonography in identifyingfractures and guiding fracture reduction inchildren. Pediatr Emerg Care. 2009;25(4):221–225
15. Weinberg ER, Tunik MG, Tsung JW. Accuracyof clinician-performed point-of-care ultra-sound for the diagnosis of fractures inchildren and young adults. Injury. 2010;41(8):862–868
16. Williamson D, Watura R, Cobby M. Ultra-sound imaging of forearm fractures inchildren: a viable alternative? J Accid EmergMed. 2000;17(1):22–24
17. Moritz JD, Berthold LD, Soenksen SF, AlzenGF. Ultrasound in diagnosis of fractures inchildren: unnecessary harassment or use-ful addition to X-ray? Ultraschall Med. 2008;29(3):267–274
18. Parri N, Crosby BJ, Glass C, et al. Ability ofemergency ultrasonography to detect pe-diatric skull fractures: a prospective, ob-servational study. J Emerg Med. 2013;44(1):135–141
19. Riera A, Chen L. Ultrasound evaluation ofskull fractures in children: a feasibilitystudy. Pediatr Emerg Care. 2012;28(5):420–425
20. Steiner S, Riebel T, Nazarenko O, et al. Skullinjury in childhood: comparison of ultra-sonography with conventional X-rays andcomputerized tomography [in German].Rofo. 1996;165(4):353–358
21. Bossuyt PM, Reitsma JB, Bruns DE, et al.Towards complete and accurate reportingof studies of diagnostic accuracy: theSTARD initiative. BMJ. 2003;326(7379):41–44
22. Rabiner JE, Khine H, Avner JR, Friedman LM,Tsung JW. Accuracy of point-of-care ultra-sonography for diagnosis of elbow frac-tures in children. Ann Emerg Med. 2013;61(1):9–17
23. Arkin CF, Wachtel MS. How many patientsare necessary to assess test performance?JAMA. 1990;263(2):275–278
24. Dietrich AM, Bowman MJ, Ginn-Pease ME,Kosnik E, King DR. Pediatric head injuries:can clinical factors reliably predict an
TABLE 4 Pooled-Data Analysis of Point-of-Care Ultrasound for Skull Fracture Diagnosis
Study (Reference) N Fractures, n (%) Sensitivity, % Specificity, % LR+ LR2
Weinberg et al (15) 21 2 (10) 100 (20–100) 100 (79–100) Infinity (2.1–infinity) 0 (0–2.15)Riera and Chen (19) 40 5 (13) 60 (17–93) 94 (79–99) 10.5 (2.3–48.2) 0.42 (0.15–1.25)Parri et al (18) 55 35 (64) 100 (88–100) 95 (75–100) 13.8 (3.0–64.6) 0.02 (0–0.24)Rabiner et al 69 8 (12) 88 (53–98) 97 (89–99) 26.7 (6.7–106.9) 0.13 (0.02–0.81)Total pooled data 185 50 (27) 94 (84–98) 96 (92–98) 25.4 (10.7–60.2) 0.06 (0.02–0.19)
Data are test performance characteristics (95% CI). LR+, likelihood ratio of a positive test; LR2, likelihood ratio of a negative test.
ARTICLE
PEDIATRICS Volume 131, Number 6, June 2013 e1763by guest on February 16, 2016Downloaded from
abnormality on computed tomography?Ann Emerg Med. 1993;22(10):1535–1540
25. Greenes DS, Schutzman SA. Clinical signifi-cance of scalp abnormalities in asymp-tomatic head-injured infants. PediatrEmerg Care. 2001;17(2):88–92
26. Kuppermann N, Holmes JF, Dayan PS, et al;Pediatric Emergency Care Applied Re-search Network (PECARN). Identificationof children at very low risk of clinically-important brain injuries after head trauma:a prospective cohort study. Lancet. 2009;374(9696):1160–1170
27. Palchak MJ, Holmes JF, Vance CW, et al. Adecision rule for identifying children at lowrisk for brain injuries after blunt head
trauma. Ann Emerg Med. 2003;42(4):492–506
28. Schutzman SA, Barnes P, Duhaime AC, et al.Evaluation and management of childrenyounger than two years old with appar-ently minor head trauma: proposed guide-lines. Pediatrics. 2001;107(5):983–993
29. World Health Organization. Essential di-agnostic imaging. Available at: www.who.int/eht/en/DiagnosticImaging.pdf. AccessedFebruary 12, 2013
30. Spencer JK, Adler RS. Utility of portableultrasound in a community in Ghana. J Ul-trasound Med. 2008;27(12):1735–1743
31. Ma OJ, Norvell JG, Subramanian S. Ultra-sound applications in mass casualties and
extreme environments. Crit Care Med.2007;35(suppl 5):S275–S279
32. Sargsyan AE, Hamilton DR, Jones JA, et al.FAST at MACH 20: clinical ultrasoundaboard the International Space Station. JTrauma. 2005;58(1):35–39
33. Cho KH, Lee SM, Lee YH, Suh KJ. Ultrasounddiagnosis of either an occult or missedfracture of an extremity in pediatric-agedchildren. Korean J Radiol. 2010;11(1):84–94
34. Grechenig W, Clement HG, Fellinger M,Seggl W. Scope and limitations of ultraso-nography in the documentation of frac-tures—an experimental study. Arch OrthopTrauma Surg. 1998;117(6–7):368–371
e1764 RABINER et alby guest on February 16, 2016Downloaded from
