23

Seat Belt Injuries:Radiologic Findingsand ClinicalCorrelation1Curtis W. Hayes, MD

William F. Conway, MD, PhD

James W. Walsh, MD

Lynn Coppage, MD

Alfred S. Gervin, MD

The seat belt syndrome consists of skeletal, soft-tissue, and visceral in-

juries associated with use of two- and three-point restraints in patients

involved in motor vehicle accidents. Skin abrasions of the neck, chest,

and abdomen-the classic seat belt sign-indicate internal injury in30% of cases. Neck abrasions are associated with injuries to the carot-

id artery, larynx, and cervical spine; chest abrasions, with fractures of

the sternum, ribs, and clavicles and injuries to the heart and thoracic

aorta; and abdominal abrasions, with mesenteric tears, bowel perfora-tion and hematoma, Chance fractures, and injuries to the abdominal

aorta. The seat belt sign should prompt a diligent search for relatedinjuries.

U INTRODUCTIONLap-type seat belts became standard equipment in automobiles in the United Statesin i 964 . The three-point variety of seat belt (combined lap and shoulder belt) be-came standard in the auto industry in i 973 . Theoretically, seat belts protect thewearer during a collision in three ways: They (a) prevent ejection from the auto,(b) allow a more controlled deceleration during the initial phase of the collision,

and (C) reduce the severity of impact between the wearer and the car interior.There is no doubt that the use of seat belts, especially the three-point variety,

lowers morbidity and mortality associated with motor vehicle accidents. A study

performed by Volvo, a Swedish car manufacturer, in 1 967 demonstrated that,

Index terms: Abdomen. injuries, 70.4 1 #{149}Arteries, injuries. 56.4 1 #{149}Bones, injuries. 40.4 1 #{149}Fractures, 40.4 1

Neck, injuries. 27.491 #{149}Pelvis, injuries, 80.41

RadloGraphics 1991; 1 1 :23-36

I From the Departments of Radiology (C.W.H., W.F.C., J.W.W., L.C.) and General Surgery, Trauma Division (A.S.G.),

Medical College ofVirginia/Virginia Commonwealth University, 12th and Marshall Sts. Box 615, MCV Station, Rich.

mond, VA 23298. Received July 26, 1990; revision requested August 29 and received September 17; accepted Sep-

tember 18. Address reprint requests to C.W.H.� RSNA. 1991

24 U RadioGraphics U Hayes et al Volume 11 Number 1

Table 1Patterns of Injuries Seen In Motor Vehicle Accidents

Location Unrestrained Two-Point Restraint Three-Point Restraint

Face, head, and neck Fractures . . . Injuries to thecarotid artery,larynx

Abdomen Liver and spleenlacerations and rupture

Hollow viscus injuries(mesenteric tears,perforations) ,

abdominal walldisruptions

Hollow viscus injuries(mesenteric tears,perforations),abdominal walldisruptions

Chest, spine Fractures of thethoracolumbar spine,pelvis, and extremities

Lumbar spinefractures (Chance)

Fractures of theclavicles, ribs,lumbar spine (Chance),and sternum

when only lap belts were used, fatal injuriesoccurred throughout the entire range of

speeds, as low as 12 mph (1). With three-

point restraints, no deaths occurred at speeds

less than 60 mph.

The first seat belt-related injury-a smallbowel injury in a trooper wearing a lapbelt-was described by Kulowski and Rost

(2) in 1 956. A pattern of injuries distinctlydifferent from those seen in unrestrained oc-

cupants has since been recognized in seatbelt wearers. Garrett and Braunstein (3)were the first to use the term seat belt syn-

drome in describing the pattern of abdomi-

nal and skeletal injuries related to seat belt

use. In this article, we review the common

and unusual features of seat belt injuries,

their radiologic appearance, and clinical and

operative correlations.

U PATI�ERNS OF INJURYUnrestrained occupants involved in motor

vehicle accidents usually sustain injuries tothe face, head, neck, and extremities second-

aiy to collision with the interior of the car,for example, the windshield, dashboard, andsteering wheel. In moderate and severe colli-sions, devastating abdominal and thoracic in-

juries related to crushing forces are corn-

mon. According tojordan and Beall (4), so!-

id organ injuries-rupture or lacerations ofthe liver and spleen, in particular-outnurn-

ber hollow viscus injuries in unrestrained oc-cupants by two to one.

Two-point restraints or lap belts reduce in-juries largely by preventing ejection from the

auto. Crush injuries are decreased except in

very severe cases. Injuries caused by lap beltsinclude mesentenic tears and perforations ofthe small bowel and colon. Proposed mecha-nisms for these injuries are shearing forces

due to rapid deceleration against the smallarea of the lap belt and increased intralu-

minal pressure secondary to compression of

a bowel loop between the seat belt and

spine, which produces “blowouts” (5). In

the spine, distractive forces cause character-istic horizontal fractures involving the ante-

nor and posterior elements of the lumbar

spine. Soft-tissue injuries attributed to lapbelts include various abrasions (seat beltsign) and rupture of the abdominal wall mus-

des and penitoneum. These three injuries-hollow viscus injury, flexion-distraction frac-tunes of the lumbar spine, and abdominalwall disruptions-are the classic seat belt in-juries.

With three-point restraints, injuries mostfrequently involve the thoracic cage: thenibs, sternum, and clavicles. Catastrophic

thoracic injuries such as aortic transection,ventricular blowouts, and massive lung lac-erations and contusions usually occur onlywith very severe, high-speed collisions. Ho!-low-viscus abdominal injuries also occur.Fractures of the lumbar spine are less corn-monly found with three-point restraints than

with lap belts. Cervical spine injuries, carot-id and subclavian vascular injuries, and Ia-ryngeal trauma have been attributed to thecrossing shoulder strap.

Table 1 summarizes the patterns of injuriesfound in restrained and unrestrained victimsof motor vehicle accidents. The followingsections elaborate on specific seat belt inju-

nies and their radiologic findings.

1. 2.

- -...-�

Figures 1-3. (1) Abdominal seat belt sign. Alarge anterior abdominal wall abrasion extendsfrom flank to flank in a victim of a high-speed mo-tor vehicle accident. (2) Thoracic seat belt sign.Diagonal abrasions (arrows) extend from the rightshoulder to the left side of the chest in the pa-tient, who was seated on the passenger side of theauto. (3) Neck seat belt sign. Diagonal abrasionsare present along the course of the shoulder strap.

January 1991 Hayes et al U RadioGrapbics U 25

3.

U TYPES OF INJURIES

. Skin AbrasionsA linear abrasion across the lower abdomenextending from flank to flank is the classic

seat belt sign (Fig i) . Although this finding

is not an incontrovertible sign of significant

internal injuries, we have found that they are

present in approximately 30% of the patients

with a seat belt sign.

A diagonal abrasion extending across thechest from a three-point restraint (Fig 2) is

often associated with underlying rib frac-

tunes, clavicular fracture, or sternal fracture.

In severe collisions, aortic transection or

ventricular rupture should be considered, al-though these injuries are, in our experience,

less common in patients who wore seat belts

than in unrestrained occupants in compara-

ble accidents.A diagonal neck abrasion on the side of the

shoulder strap (Fig 3) may be associated

with vascular injury, tracheal or laryngeal in-

jury, or, occasionally, cervical spine injury,

including transverse process fractures.

. Skeletal InjuriesFractures of the thoracolumbar spine are

common seat belt injuries in patients whowore two-point and, to a lesser extent, three-point restraints. The pattern of injury is fre-

quently different from that of fractures sus-

tamed by unrestrained victims, and can beexplained by the three-column concept ofthe structure of the spine by Denis (6) . Ac-cording to this theory, the anterior column isformed by the anterior longitudinal liga-ment, anterior annulus fibrosus, and anteriorpart of the vertebral body. The middle col-

umn is formed by the posterior longitudinal

ligament, posterior annulus fibrosus, andposterior wall of the vertebral body. The pos-

a. b.

26 U RadioGrapbics U Hayes et al Volume 11 Number 1

Figure 4. Thoracolumbar junction flexion-distraction injury. (a) Anteroposterior radiograph shows an-terolateral compression of the superior end plate of L- 1 (large arrow). Injury to posterior elements may besuspected due to splaying of spinous processes (small arrows). (b) Lateral radiographic findings confirmdistraction of posterior elements (arrow).

tenor column is formed by the posterior

bony elements and the posterior ligamentous

complex. A flexion force on an unrestrainedoccupant has a fulcrum centered at the mid-die column, or posterior half of the vertebralbody. This produces compression of the an-tenor vertebral body and, with more severe

forces, distraction of the posterior elements.Moderate forces produce simple, neurologi-cally stable, compression fractures in these

cases. Severe forces produce flexion-distrac-tion injuries and fracture dislocations due to

failure of all three columns; such injuries are

unstable and commonly associated with neu-

rologic deficits. They occur most often at thethoracolumbar junction (Fig 4) . The anteriorcompression may be mild or moderate. More

important is the disruption of the posteriorelements, which must be searched for care-fully.

Restrained occupants, especially those

wearing lap belts, show characteristic hori-zontal fractures, frequently at L-2 or L-3.These fractures were first reported by

Chance (7) in 1948. Several variations ofthese injuries were reported by Smith and

Kaufer (8) in i969. These may involve a sin-gle vertebral level (body and posterior dc-ments) , two levels, purely soft-tissue disrup-iion, or a combination. The mechanism of in-

jury is the shifting of the fulcrum of the force

to the anterior abdominal wall, caused by theseat belt, thus producing distraction on all

three columns (anterior, middle, and poste-nor) of the spine. These fractures show little

or no anterior compression of the involvedvertebral body (Fig 5) . Neurologic deficits

are infrequent.In some cases, distractive forces produce

purely ligamentous disruption at the in-volved level, leading to a lumbar subluxa-tion/dislocation or “Chance equivalent” in-

jury. This injury usually involves disruption

of the soft-tissue components of the posteni-or and middle columns of the spine (posteni-or ligamentous complex, posterior longitu-dinal ligament, and posterior annulus fibro-

sus). Radiographic findings of this unstable

injury may be subtle: mild widening of the

posterior aspect of the affected disk space,

widened facet joints, and splaying of spinousprocesses. On the anteroposterior view, the

latter abnormality produces the “empty holesign” (Fig 6). Chance fractures and their hg-

6a. 6b.Figures 5, 6. (5) Chance fracture of L-2. (a) Anteroposterior radiograph shows characteristic splaying ofthe spinous processes (double arrow), with fractures extending through both transverse processes andpedicles (curved arrows) . (b) Lateral radiograph shows the fracture extending through the posteroinferiorcorner of L-2 (arrow), into the L2-3 disk space. (6) Ligamentous Chance injury at L3-4. (a) Anteroposte-nor radiograph shows mild widening of the involved disk space (large arrows) and an ‘ ‘empty hole sign”(small arrows), denoting splaying of the spinous processes. (b) Lateral radiograph more clearly showssplaying of the spinous processes (double arrow) and widening of the facet joints (curved arrow).

January 1991 Hayes et a! U RadloGrapbics U 27

Figure 7. Rib fractures, three-point restraint pattern. (a) Posteroanterior radiograph shows the diag-onal course of the shoulder strap. (b) Diagonal seat belt sign (arrowheads) is seen in the same patient.(c) Oblique radiograph shows fracture of the right first rib (arrow). (d) Anteroposterior radiograph showsfractures of the left fourth through ninth ribs (arrows). The sternum was also fractured, but there was noaortic or cardiac injury.

28 U RadioGrapbics U Hayes et al Volume 11 Number 1

January 1991 Hayes et a! U RadioGraphics U 29

Figure 8 Sternal fracture. Lateral radiographdemonstrates a slightly depressed fracture of onlythe anterior cortex of the sternum (arrows) imme-diately beneath the site of the shoulder strap.

amentous equivalents are frequently seen inchildren and other back-seat passengers,

since two-point restraints are still used in the

back seats of most autos. New legislation

making back-seat three-point restraints man-

datory may decrease the number of such in-

juries.

Anterolateral compression fractures have

been reported in occupants wearing shoul-der restraints. These injuries occur at thethoracolumbar junction and may be due to

a roll-out mechanism of combined flexion

and rotation about the axis of the shoulder

strap (9).

Rib fractures are common in all victims of

motor vehicle accidents of moderate or se-

vere force, regardless of whether restraintsare used. With three-point restraints, frac-

tunes tend to align along the course of the di-

agonal strap (Fig 7) . The frequency of severeflail chest with multiple fractures is said tobe decreased with seat belt usage, however(i 0) . Arndt (1 1) has also described fracturesof the transverse processes of C-7 or T- 1 , as-

sociated with usage of shoulder harnesses.Sternal fractures are common in patients

with seat belt injuries. It is theorized that the

deceleration force is absorbed by the ster-num through the small area of the shoulder

strap, accounting for the relatively low fre-

quency of aortic transections or ventricularblowouts, except in victims of accidents ofextremely severe force. We have observedfractures involving only the anterior cortex

of the sternum immediately beneath theshoulder strap (Fig 8), which appears to be

characteristic of this mechanism. Sternalfractures in unrestrained occupants or thosewith lap belts only (due to impact againstthe steering wheel or dashboard) are oftenassociated with myocardial contusion. Todate, we have been unable to document any

cases of myocardial contusion associated

with sternal fractures caused by shoulder

straps.

Clavicle fractures are commonly associat-ed with use of three-point restraints duringmotor vehicle accidents.

. Soft-Tissue and Visceral InjuriesTransverse tears of the rectus abdominis mus-

dc and anterior peritoneal tears are becom-ing more frequently recognized as part of theseat belt syndrome. These injuries occur im-

mediately beneath the site of the lap portion

of the seat belt and may be quite subtle ra-diologically (Fig 9) . Computed tomography(CT) may show only abrupt thinning of theinvolved muscle at the level of injury, and

this finding may easily be overlooked. Thisinjury is frequently associated with hollow

viscus injury (i 2) and may provide a clue to

the gravity of the injury.

e. I.

Figure 9. Abdominal wall disruption. (a-d) CT images demonstrate rectus muscle disruption in a 14-year-old victim of a high-speed motor vehicle accident. Images show normal rectus abdominis muscles su-periorly (straight arrows in a) , virtually absent rectus muscles in the midportion of the abdomen (arrowsin b, c) , and normal-appearing muscles inferiorly (arrows in d) . Subcutaneous flank hemorrhage was alsoseen (curved arrow in a). An abdominal seat belt sign was present. (e) Diagram illustrates transverse tearsof the rectus abdominis muscles at the site of abdominal abrasions (arrows). (0 Intraoperative view showsan extensive tear of the anterior abdominal wall (arrows).

30 U RadloGraphics U Hayes et al Volume 11 Number 1

a. b.

c.

January 1991 Hayes et al U RadioGraphics U 31

The most common significant injuries en-

countered in restrained patients are mesen-

tenic tears and frank ruptures of the small

bowel and colon. Vascular tears occur on the

mesenteric side of the bowel, more frequent-

ly in the small bowel than in the colon. Per-

forations nearly always occur on the antimes-

enteric border. Jejunal and ileal perforations

are most common, with large intestine andduodenal ruptures less frequent. In an exam-

ination of case reports, Williams and Kirkpat-

rick (1 3) found a higher rate of hollow vis-

cus injury in patients who wore two- versus

three-point restraints. Sixty-three of 82 inju-

ries involved the small bowel; 1 9 involved

the colon. Mesenteric tears and bowel perfo-

rations were equally common. The true fre-quency of these injuries cannot be deter-mined from this study, however, since it rep-resented an analysis of case reports.

The efficacy of CT in the evaluation of

blunt abdominal trauma and potential hol-

low viscus injuries has been studied by sever-

Figure 10. Small bowel perforation. CT scansobtained with oral and intravenous administrationof contrast material show unequivocal extralu-minal gas (arrows in a) , mesenteric thickening(arrow in b) , and moderate free fluid in the pelvis(arrow in c).

a! authors (1 4- 1 7) . In their 1 989 study of5 1 patients (37 ofwhom were in motor vehi-

dc accidents) , Rizzo et al (1 8) examined thefrequency of the following CT findings: cx-

traluminal air, free intraperitoneal fluid,

thickened or infiltrated mesentery, and

thickened bowel wall (Figs 10, 1 1). Extralu-

minal air was observed in only nine of 16

cases of perforation (56%) . Two cases

showed extravasation of orally administered

contrast material. In all cases of bowel perfo-

rations, there were moderate to large

amounts of free intraperitoneal fluid. Thick-

ened bowel wall (75%) and mesentenic infil-tration (88%) were also found in most cases

of perforation. From this study and others,

one may conclude that (a) free gas or extrav-asated contrast material are relatively specif-

ic but insensitive (56%) signs of bowel per-

foration; (b) the presence of moderate orlarge amounts of free intraperitoneal fluid,when not associated with solid organ lacera-tion, should raise the suspicion of bowel per-foration or mesenteric vascular tear; and (c)

a localized mesenteric hematoma or the find-ing of a focal high-attenuation clot adjacentto bowel are helpful in localizing the site ofinjury. In the above situation, CT and diag-nostic penitoneal lavage should be used ascomplementary diagnostic examinations.

The relative roles of CT and diagnosticpenitoneal lavage in the evaluation of blunttrauma have been examined by a number ofauthors and remain controversial (19-25).

In cases in which there is a high suspicionfor bowel injury (seat belt sign or Chance

32 U RadioGraphics U Hayes et al Volume 11 Number 1

Figure 15. Intraperitoneal bladder rupture. Cys-togram obtained with the patient supine demon-strates intraperitoneal extravasation of contrastmaterial from the bladder, with contrast materialclearly outlining the right paracolic gutter and in-ferior margin of the liver.

January 1991 Hayes et a! U RadloGrapblcs U 33

fractures of the spine) , judicious use of both

tests is warranted. Because small bowel lac-

erations may initially be asymptomatic and

since delay produces significant morbidity,

rapid evaluation, with both tests if necessary,

is recommended. When hollow viscus injuryis suspected, CT should be performed prior

to diagnostic peritoneal lavage, since the

important CT signs of intraperitoneal fluid

or free air may be related to the lavage it-

self (26).The frequency of liver and splenic injuries

is reduced with the use of two- and three-

point restraints. Severe forces and crush inju-ries continue to produce the usual comple-

ment of liver lacerations (Fig 1 2) , splenic

lacerations and ruptures (Fig 1 3) , and pan-

creatic (Fig 14) and renal injuries.

Significant injuries may be produced withless severe forces when there is improper po-

sitioning of the two- or three-point restraint

(27,28). Proper usage dictates that the lapbelt be positioned across the anterior iliacspine so that the forces are dissipated

through the stable, bony pelvis. If the beltmigrates up onto the abdominal wall (sec.

ondary to obesity, slouching, or “submanin-ing”) , the load of a sudden deceleration maybe dissipated over the abdominal compart-ment, resulting in visceral injuries. Place-ment of the shoulder strap under the arm bya passenger-side occupant produces a force

directly across the liver, with potentially

devastating effects.Intraperitoneal rupture of a distended

bladder may occur due to lap belt pressureduring an accident (Fig 1 5) . Urethral injuryhas occasionally been reported as well. Sev-

eral cases of uterine injury have also been re-ported (29), including a case in which nip-ture of a gravid uterus left a 6th-month fetusfree in the abdominal cavity. Fortunately,such reports are rare.

Cases of diaphragmatic rupture have beenreported in seat belt wearers (30) (Fig 16).The mechanism of the injury may be second-aiy to faulty placement of the belt or slidingof the occupant under the belt in the earlystages of the impact, leading to upward ab-

dominal pressure from the belt and subse-quent herniation of abdominal contents intothe thoracic cavity.

common carotid angiogram demonstrates a transection (arrow) below the bifurcation. (b) CT scan of thesame patient shows extensive gas in the soft tissues of the anterior and lateral aspects of the neck second-ary to laryngeal fracture.

I, � --�.--- ----�---- �----- - --- -- ---C

proximal descending aorta, with a well-definedflap (arrow) at the lower extent of the injuredsegment.

34 U RadioGrapbics U Hayes et a! Volume 11 Number 1

. Vascular InjuriesWe have seen one case of transection of thecommon carotid artery in a patient involvedin a motor vehicle accident while using a

three-point seat belt (Fig 1 7a) . The same pa-tient also had a fracture of the larynx that ne-cessitated a tracheostomy (Fig 1 7b) . Similar

cases, as well as injuries to the internal carot-

id artery, subclavian artery, and superior

vena cava, have also been reported In the lit-erature (31-35).

The force of collision required to producethoracic aortic tears appears to be greater inrestrained occupants than in unrestrainedpersons. The shoulder strap may allow a

more controlled deceleration of the chestduring the collision, which may decrease theshearing force on the aorta somewhat. Still,aortic injuries are relatively common with

severe forces (Fig i 8) (36).

Seat belt-related injuries involving the ab-dominal aorta have been reported (37,38).We observed one case in which distal ab-dominal aortic occlusion occurred due to anear-complete transection (Fig 19). This pa-

tient also sustained a Chance fracture at the

same level. The mechanism of injury waspresumably distraction and compression ofthe aorta between the seat belt and spine.

U CONCLUSIONSGiven the same force of impact during the

accident, a seat belt wearer (especiallythree-point type belt) will usually sustainless severe injuries than a nonwearer. Seat

January 1991 Hayes et a! U RadioGrapbics U 35

Table 2Seat Belt Synd rome: Physical Signs and Related Injuries

Location of

Abrasion Affected Structures

Neck Injuries to the carotid artery, larynx, cervicalspine, transverse process

Chest Injuries to the sternum, ribs, clavicle, heart(myocardial contusion) , thoracic aorta

Abdomen Injuries to the small and large bowel(mesenteric tears, perforations, hema-toma) , vertebrae (Chance fractures,especially L-2, L-3), abdominalaorta and branches

Figure 19. Distal abdominal aortic occlusionfollowing seat belt injury. Anteroposterior ab-dominal aortogram shows complete occlusion justproximal to the aortic bifurcation. At surgery,complete transection with only intact adventitiawas found. A Chance fracture was present at L-4.

belts prevent ejection of the wearer from thevehicle and decrease the severity of the oc-cupant’s impact with the steering wheel,

dashboard, and other interior structures.However, a distinctly different pattern of in-juries occurs in seat belt wearers. In our cx-

perience, the scat belt sign-abdominal,chest, or neck bruises and abrasions at the

site of belt contact-is associated with sig-

nificant internal injuries in approximately

30% of cases. Chance fractures and their liga-mentous equivalents, large and small bowelperforations or mesentenic tears, and abdom-inal wall disruptions are more common thansolid organ injury. Occasional significantvascular injuries also occur. Table 2 summa-nizes the various injuries associated with dif-ferent seat belt abrasions. The history of a re-strained occupant in a significant motor ye-hide accident or the presence of the telltaleseat belt sign should prompt a diligentsearch for these related injuries.

U REFERENCES1 . Bohlin NI. A statistical analysis of 28,000

accident cases with emphasis on occupantrestraint. Proceedings of the Eleventh StappCar Crash Conference. Soc Automotive EngTrans 1967; 76:2981-2994.

2. KulowskiJ, Rost WB. Intra-abdominal inju-ry from safety belt in auto accident. ArchSung 1956; 73:970-971.

3. GarrettjW, Braunstein PW. The seat beltsyndrome.JTrauma 1962; 2:220-237.

4. Jordan GL, Beall AC. Diagnosis and manage-ment of abdominal trauma. Curr Prob Surg1971 ; November:3-62.

5. WilliamsJS, Lies BA, Hale HW. The automo-tive safety belt: in saving a life may produceintra-abdominal injuries. J Trauma 1966;6:303-313.

6. Denis F. The three column spine and its sig-nificance in the classification of acute thora-columbar spinal injuries. Spine i983; 8:817-83 1.

7. Chance GQ. Note on the type of flexionfracture ofthe spine. BrJ Radiol 1948;21:452-453.

8. Smith WS, Kaufer H. Patterns and mecha-nisms of lumbar injuries associated with lapseat belts.J Bonejoint Surg [Ami 1969; Si:239-254.

9. Miniaci A, McLaren AC. Anterolateral corn-pression fracture of the thoracolumbar

36 U RadioGrapbics U Hayes et a! Volume 11 Number 1

spine: a seat belt injury. Clin Orthop Rd Res1989; 240:153-156.

10. Newman RJ. Chest wall injuries and theseat belt syndrome. Injury I 984 ; 1 6: 1 10-113.

1 1 . Arndt RD. Cervical-thoracic transverse pro-cess fracture: further observation on the

seatbelt syndrome.J Trauma 1975; 15:600-

602.i 2 . Johnstone BR, Waxman BP. Transverse dis-

ruption ofthe abdominal wall: a tell-talesign of seat belt related hollow viscus injury.Aust N ZJ Sung 1987; 57:455-460.

1 3 . Williams JS, Kirkpatrick JR. The nature of

seat belt injuries. J Trauma 197 i � 1 1:207-

216.14. Cook DE, WalshJW, Vick CW, Brewer WH.

Upper abdominal trauma: pitfalls in CT diag-nosis. Radiology 1986; 159:65-69.

1 5. DonohueJH, Federle MP, Griffiths BG,

Trumkey DD. Computed tomography in thediagnosis of blunt intestinal and mesenteric

injunies.JTrauma 1987; 27:11-17.i6. Orwig D, Federle MP. Localized clotted

blood as evidence of visceral trauma on CT:the sentinel clot sign. AJR 1989; 153:747-

749.

1 7. SherckJP, Oakes DD. Intestinal injuriesmissed by computed tomography. J Trauma1990; 30:1-7.

18. Rizzo MJ, Federle MP, Griffiths BG. Boweland mesenteric injury following blunt ab-dominal trauma: evaluation with CT. Radiol-ogy 1989; 173:143-148.

i 9. Wing VW, Federle MP, Morris JA, Jeffrey RB,Bluth R. The clinical impact of CT for bluntabdominal trauma. AJR 1985; 14�:1191-

1i94.20. Fabian TC, Mangiante EC, White 1), Patter-

son CR, Boldreghini S, Britt LG. A prospec-tive study of 9 1 patients undergoing bothcomputed tomography and peritoneal la-

vage following blunt abdominal trauma. JTrauma 1986; 26:602-608.

2 1 . Powell RW, Green JB, Ochsner G, Barttel-

bort SW, Shackford SR, Sise MJ. Peritoneallavage in pediatric patients sustaining bluntabdominal trauma: a reappraisal. J Trauma1987; 27:6-10.

22. Gomez GA, Alvarez R, Plasencia G, et al.Diagnostic penitoneal lavage in the manage-ment of blunt abdominal trauma: a reassess-ment.J Trauma 1987; 27:1-5.

23. Keaney PA, Vahey T, Burney RE, Glazer G.Computed tomography and diagnostic pen-

toneal lavage in blunt abdominal trauma.

ArchSurg 1989; 124:344-347.

24. Bulas DI, Taylor GA, Eichelberger MR. Thevalue of CT in detecting bowel perforationin children after blunt abdominal trauma.AJR 1989; 153:561-564.

25. Taylor GA, Guion CJ, Potter BM, Eichel-

bergen MR. CT of blunt abdominal trauma

inchildren.AJR 1989; 153:555-559.26. Kane NM, Dorfman GS, CronanjJ. Efficacy

of CT following penitoneal lavage in abdom-inal trauma.J Comput Assist Tomogr 1987;

11:998-1002.

27. Sato TB. Effects of seat belts and injuries re-

sulting from improper use. J Trauma 1987;27:754-758.

28. StatesJD, Huelke DF, Dance M, et al. Fatal

injuries caused by underarm use of shoulderbelts.JTrauma 1987; 27:740-745.

29. Rubovitz FE. Traumatic rupture of the preg-nant uterus from seat belt injury. Am J Ob-stetGynecol 1964; 90:828-829.

30. Bengqvist D, Dahlgren S, Hedelin H. Rup-tune of the diaphragm in patients wearingseatbelts.JTrauma 1978; 18:781-783.

3 1 . Ernest A, Robertson HJ, Bercien ML, KlineDG. Occult carotid artery injury related to

automobile seat belts. Ann Emerg Med1988; 17:1091-1094.

32. Benito MC, Garcia F, Fernandez-Guero L, etal. Lesion of the internal carotid arterycaused by a car safety belt. J Trauma 1990;30:116-117.

33. Costa MC, RobbsJV. Nonpenetrating sub-clavian artery trauma. J Vasc Sung i 988;8:71-75.

34. van de Wal HJCM, DraaismaJMT, VincentJG,Gonis RJA. Rupture of the supnadiaphrag-matic inferior vena cava by blunt decelerat-ing trauma: case report. J Trauma 1990;

30:1 1 1-1 13.

35. Guertlier AT. Blunt laryngeal trauma asso-

ciated with shoulder harness use. Ann EmergMed 1988; 17:838-839.

36. Arajarvi E, Santavirta S. Chest injuries sus-

tamed in severe traffic accidents by seatbeltwearen.JTrauma 1989; 29:37-41.

37. Warnian RK, ShoenutJP, lannicello CM,Sharma GP, Trenholm BG. Seatbelt injury

to the abdominal aorta. J Trauma 1988;28:1 505-i 507.

38. DaJee H, Richardson 1W, lype MO. Seat beltaorta: acute dissection and thrombosis of theabdominal aorta. Surgery 1979; 85:263-

267.