Secondary Logo

Journal Logo


Atlanto-Occipital Injuries in Three Children Treated With Halo Immobilization

Parent-Weiss, Nicole M. CO, OTR, FAAOP; Weiss, David B. MD; Farley, Frances A. MD

Author Information
JPO Journal of Prosthetics and Orthotics: January 2007 - Volume 19 - Issue 1 - p 22-28
doi: 10.1097/JPO.0b013e31802d4726
  • Free

Atlanto-occipital dislocations (AOD) are rare and usually fatal injuries.1 Fatalities are generally associated with transsection of the medulla oblongata or the spinomedullary junction.2 AOD was first reported in 1908 by Blackwood,3 who described a single case study on a 34-year-old patient. AOD commonly occurs in patients involved in high-energy collisions but because of the high mortality associated with this injury, absolute incidence can not be determined or reported.4,5 However, more aggressive and rapid emergency response care is credited with increasing survival rates, especially among children.6 Postmortem analysis shows AOD is not as rare as previously suspected.7 Alker et al.8 reported that 19 cases of 312 involved in car accidents sustained AOD. Bucholz and Burkhead9 reported 26 of 112 patients involved in fatal motor vehicle accidents sustained cervical spine injuries, and 9 of those 26 (33%) sustained AOD. Twenty-five percent of patients dying of cervical spine injuries are found to have AOD.10 Additional postmortem analysis of pediatric trauma from 1985 to 1990 showed 18% sustained AOD.10

For a variety of reasons, most of those who survive AOD are children.5,7 Children are more than twice as likely as adults to sustain AOD,9 in part because of the small size of occipital condyles and the horizontal orientation plane of the atlanto-occipital joint.4,11 The increased occurrence may also be attributable to the laxity of the supporting ligaments.2,4,10 Anatomical examination of the injury site reveals that the craniovertebral junction is formed by the occipital-axial joints. The occipital condyles and superior articulating facets of the axis provide ball-and-socket motion at the craniovertebral junction. However, the atlas is only loosely connected to the occiput by weak anterior, posterior, and lateral occipital-axial membranes. Thus, the stability is provided mainly by the ligamentous connections between the occiput and the axis.12 These ligamentous connections include the tectorial membrane, the cruciate ligament, the apical dental ligament, and the paired alar ligaments. The tectorial membrane and the alar ligaments are the most structurally important.13 AOD denotes separation of the joint between the occiput and the first cervical vertebrae (atlas).4

There are several accepted methods for define and diagnosing AOD. One of the earliest methods defines AOD as an injury in which there is an increased abnormal distance between a line formed from the basion and opisthion in relation to a line between the anterior and posterior arch of C1. If the ratio between the two is greater than one, the diagnosis suggests anterior AOD.13 With the occurrence of missed diagnoses and overlooked dislocations that have spontaneously reduced, the more common measurement of the distance from the tip of the dens and the basion is used. With this method, a distance of more than 12.5 mm suggests AOD. One third of survivors of AOD have had the injury overlooked on first examination.7,14,15 This may be attributable in part to the occurrence of spontaneous reduction.16 Powers et al.13 report that successful management of this serious injury requires early radiographic identification (Figure 1).

Figure 1.:
Radiograph of AOD.

AOD can be classified into three types based on the direction in which the occiput is displaced in relation to the atlas: anterior, posterior, or longitudinal.2 The most common is the anterior dislocation. Lateral AOD is rare.12,17 Most authors consider hyperextension to be the mechanism of injury.4,5,7,11–13,18 Although hyperextension may be considered the abnormal motion primarily responsible for AOD, it is the combination of dynamic forces that results in hyperextension and lateral rotary flexion4,10 that truly lead to AOD. Anterior AOD is the most common type among children injured in auto accidents. Werne19 agreed that hyperextension leads to severance of the tectorial membrane and alar ligaments, resulting in excessive hyperextension that causes dislocation.

Treatment recommendations vary from reduction with internal fixation4,20,21 to trial of halo immobilization only,21 followed by surgical reduction if conservative halo methods fail. Most authors agree that a halo device should be used whether for conservative treatment or postoperative immobilization.5,13,21 The halo cervicothoracic orthosis (CTO) is considered the most secure method of immobilization as compared with other conventional cervical orthoses.22 Successful conservative treatment with halo immobilization is challenging and requires the most intimate fit of both ring and vest to maintain anatomical position. The minute amounts of migration of the vest in relation to the torso that commonly occur with even the best of vest fits may result in loss of anatomic reduction.

We present three cases of children who sustained AOI just short of a dislocation. All three were treated with a halo, and one had a cervical fusion. Our hypothesis is that halo immobilization is adequate for the subtle injuries. Frank dislocations require posterior fusion along with halo immobilization.



A 6-year-old boy was involved in a high-speed motor vehicle accident. He was not in a child safety seat or booster seat and was restrained in the back seat by a lap belt with no shoulder harness. His injury included an occiput-C1 injury in which magnetic resonance imaging showed a 4-mm hematoma between the clivus and the foramen magnum, causing elevation of the tectorial ligament. He also sustained three small bowel perforations, a right frontal depressed skull fracture with traumatic brain injury, and an L3-4 chance fracture. His operative treatment included exploratory laparotomy and repair of his small bowel. He also underwent open reduction and internal fixation with posterior spinal fusion of the L3-4 chance fracture. His postoperative treatment included a custom thoracolumbosacral orthosis (TLSO), to stabilize his L3-4 injury, attached to a halo to address his AOI (Figures 2).

Figure 2.:
Halo attached to TLSO.

Initially he was placed in a modified prefabricated halo vest to stabilize his cervical spine pending operative treatment of his lumbar injury. The bracket width was decreased and the circumference was modified to allow appropriate compression to prevent superior migration of the vest (Figure 3). This prevention of distraction forces was essential to maintain alignment of the occiput and cervical spine. A halo ring predrilled with multiple holes was used. Six pins were inserted (two anterior and four posterior), and pin torque was brought to 6 inch lbs.

Figure 3.:
Decreased width of halo bracket.

The patient underwent operative treatment for his lumbar fracture during which he was stabilized in a Mayfield adapter (Depuy Acromed, Raynham, MA) with removal of the posterior vest to allow access. After surgery, he was molded for a TLSO to be attached to his halo. The molding procedure consisted of removal of the anterior vest with manual stabilization of the cervical spine, specifically preventing any distraction forces. The anterior chest was molded with a halo bracket molded into the plaster cast. The uprights were then attached to the anterior cast to maintain alignment during the rolling of the patient to the prone position for posterior TLSO casting. He was then placed prone, lying on the anterior plaster cast. The posterior vest was removed with similar manual stabilization of the cervical spine. The posterior chest was molded and removed once set, followed by replacement of the posterior vest. The patient was then rolled supine, and the anterior mold was removed and the anterior vest replaced.

Once fabricated, the TLSO, with halo brackets attached, was attached to his superstructure following removal of the vest with constant manual stabilization of the cervical spine during transition. The halo system was equipped with distraction adjustment components. They were simply reversed to allow increased compressive forces with adjustment and to prevent distraction.

In 2002, Steinmetz et al.23 highlighted the shortcomings of pediatric vest fit and the need for augmentation with straps to secure a pediatric vest when addressing AOD. Although the attachment of the halo to a rigid TLSO secured over the waist and iliac crest is believed to have been enough support to maintain position of the halo and prevent distraction, groin straps were attached to the halo for secondary security and prevention of distraction. The straps were designed to be removable and replaced for hygiene with a plastic rivet securing them to the TLSO and overlapping Velcro to allow removal. The straps were well padded, and no skin complications occurred. The patient remained in the halo TLSO for 8 weeks, magnetic resonance imaging showed normal cranial cervical junction with resorption of the hematoma. The halo was removed at this time. One year after surgery plain radiographs demonstrated normal alignment of the cervical spine.


The 3-year-old sibling of the patient in case 1 was involved in the same high-speed motor vehicle accident. He was restrained in the back seat by only a lap belt with no shoulder restraint and was not positioned in a child safety seat or booster seat. His injuries included AOI with 1 cm hematoma from the clivus to the foramen magnum and a longitudinal split in the tectorial membrane and a widening of the occiput-C1 space with fluid in the joint. He also sustained a small bowel perforation, traumatic brain injury, and bilateral iliac wing fractures. Operative treatment included exploratory laparotomy and repair of a small bowel perforation.

This patient also underwent an occiput-C2 posterior spinal fusion with postoperative placement of a halo vest. A pediatric ring with posterior offset to allow occipital access and multiple holes was used. Eight-pin fixation (Figure 4) was used initially (four anterior and four posterior) with initial pin torque set at 3 inch lbs. The patient experienced pin loosening complications 10 weeks after halo application related to a fall that resulted in pin loosening but no dislodgement. The pins were retorqued to 2 inch lbs but not removed. Two additional posterior pins were set to 3 inch lbs. The halo vest was a modified prefabricated vest with bracket width reduced from standard 7-inch width to a 3-inch width. The thoracic straps were modified to apply appropriate compression, and the length was reduced significantly to avoid impingement during sitting and forward flexion, per the physician's request. Groin straps similar to those previously discussed were attached to the inferior aspect of the halo vest (Figure 5) to further securely prevent distraction.

Figure 4.:
Eight-pin fixation on pediatric ring.
Figure 5.:
Groin straps attached to decrease superior migration.

Six months later, the child underwent a revision occiput-C1 fusion with plate and screws and repeat bone grafting because of demonstrated motion at occiput-C1. The halo was modified to accommodate growth and weight gain. Straps were lengthened and plastic replaced on the vest to accommodate increased torso length. Two anterior pins were retorqued to 2 inch lbs with less than one revolution of the pin. Two anterior pins were added with initial torque of 3 inch lbs. Halo immobilization lasted an additional 3 months after the second surgery, for a total of 9 months of halo treatment. One year later, the cervical fusion appeared solid on plain radiographs.


A 4-year-old girl was involved in a high-speed motor vehicle accident. She was restrained in the back seat with a lap belt and shoulder harness. Her injuries involved bilateral mandible fractures, odontoid fracture, occiput-C2 injury, C1-2 interspinous, transverse ligament and posterior longitudinal ligament injury, as well as a C1 lateral mass fracture with approximately 4 mm hematoma between clivus and the tectorial membrane consistent with partial tear of the tectorial membrane.

Treatment included an attempt at conservative management of AOI with halo immobilization only. Ring design included multiple pin holes with less than 1 cm of clearance between skin and ring. Eight-pin fixation with initial pin torque of 4 inch lbs was used. A modified prefabricated vest was attached to the ring and superstructure. The bracket width was reduced to 4.5 inches. Circumference and plastic were modified to apply appropriate compression and length. Firm contact was achieved at the deltopectoral groove. Groin straps were attached, and distraction bolts were reversed to apply a compressive force.

Conservative treatment was unsuccessful, and despite several adjustments, superior migration of the vest repeatedly occurred. With superior migration of the vest, a distracting force is allowed, which in the case of this injury was detrimental to healing and stabilization. The patient underwent operative odontoid screw fixation and continuation of halo immobilization after surgery. Radiographs 6 weeks after surgery demonstrated stable alignment in flexion and extension of both the dens fracture and occiput-C1 articulation.


The fit of any halo vest to a small child requires ultimate attention to detail and the principles of appropriate fit criteria. Specifically related to halo treatment with a diagnosis of AOD, it has been suggested that groin straps are necessary to maintain compression of the atlanto-occipital joint and ultimately prevent distraction forces. The study of Steinmetz et al.23 highlighted the shortcomings of pediatric halo vest fit and the subsequent need for augmentation with groin straps to assure an adequate fit. However, in the instance of AOI, there are numerous details involved with the fitting of small children in halo vests that may make this modification unnecessary. Specifically, decreasing the width of the halo bracket, increasing the overall intimacy of fit of the vest, and decreasing the length so it does not extend below the costal margin all assist with increasing the stability of the vest fit. The design of the halo vest to function optimally and methods to customize and appropriately fit a small child with a halo become the main challenges for the orthotist. It is proposed that with this intimate fitting of the halo vest, modifications of adding groin straps may be unnecessary. Proper fit may achieve the same goal.

Current literature documents the use of a halo in children as young as 7 months of age,24 with more common reporting of use in children 3 to 5 years of age.25–27 The documentation shows successful use of the halo, provided that the appropriate components are used and applied.28 Complication rates are reported to occur at rates similar to those of adults.29 Extensive research has been done on location of safe zones for pin placement30–33 and pin torque for pediatric halo application. Determination of safe pin placement in children necessitates a computed tomography scan to determine appropriate and safe placement. There are no safe zones for pediatric skulls.

Currently, the authors use computed tomography scanning to determine pin placement and refer to the method of pin torque determination with the method described by Mandabach et al.,26 in which 1 inch pound per year of life is an appropriate starting point. We also further determine appropriate torque by the theory of Garfin et al.30 of multiple pin sites with low torque and a guideline of 4 to 6 inch lbs for children 3 to 5 years of age. Well-calibrated torque wrenches are necessary to allow minute adjustment of torque pressures. Adjustability of the torque wrench varies with manufacturer. The more adjustability that is present, the more precise the orthotist can be in determining exact pin torque. Some torque wrenches allow adjustability to the 1/10 inch pound. Clinical judgment plays an essential role in final determination of pediatric pin torque. Considerations must also be made for appropriate halo ring fit. Pediatric rings should be designed with multiple pin site holes to allow for multiple pin placement when necessary. Ring design should also include offset above the occiput to allow surgical access when necessary. Current pediatric ring designs that meet all these specifications are available only in custom rings. When paired with careful evaluation and determination of whether occipital access is needed, standard rings can be used.

Studies have been performed on the importance of halo vest fit in adults.34 Similar criteria apply to the fit of a vest in the pediatric patient. Pediatric halo vest criteria can be classified into these general areas of importance: appropriate circumference, length and bracket width, along with firm contact in the deltopectoral groove and interscapular pressure directed anteriorly. The circumference remains an important aspect of halo vest fit in children. There must be adequate circumferential pressure without compromising respiration. The material of the vest should slightly conform around the thorax to provide close contact with compression and avoid edge pressure. Connections between the anterior and posterior aspects of the vest should be at minimum semirigid to prevent shear forces between the two vest pieces with patient motion. The length of the halo vest should not extend below the costal margin to avoid superiorly directed pressure during sitting and forward or lateral flexion. Accommodations should be made for a protruding abdomen when necessary. This can be done with a higher anterior cutout or heat flare modification.

Bracket width recently has become a controversial topic. Current prefabricated pediatric halo vests often have the same width bracket on both child-size and full-size adult vests. This additional inappropriate width leads to inadequate contact anteriorly in the deltopectoral groove. Without this contact, there can be decreased posterior directed force by the anterior section of the vest. Posteriorly, an inappropriately wide halo bracket can lead to uneven pressure over the scapular spine, leading to increased pressure and higher incidence of pressure-related skin complications. Contact should be intimate at the interscapular space directed anteriorly, with potential pressure avoided over the spinous processes. With wider attachment of the superstructure caused by inappropriately increased bracket width, the lever arm of the uprights becomes longer and is at a biomechanical disadvantage.

Pediatric halo vest fit can be performed by three methods: 1) modification of a prefabricated vest; 2) fabrication of a custom-molded vest; or 3) application of a halo cast. Halo casts on small children achieve an intimate custom fit and do not require the fabrication time of a plastic vest. However, a rigid cast remains nonadjustable as body shape changes due to decreasing edema or atrophy, resulting in potentially decreased accuracy of fit with time. The cast also prohibits intimate contact in key areas required for appropriate thoracic stabilization when connected to a halo vest. The cast generally also presents the challenge of applying an appropriately sized bracket for thorax width, which can be crucial in maintaining alignment and position. The custom-molded halo vest provides the most intimate fit. However, it requires plaster or fiberglass mold, which can cause a delay in application. Safe stabilization of the patient during molding may also be an issue that precludes custom molding for certain traumatic unstable cases. Prefabricated vests are often applicable with the understanding that significant modification is frequently necessary.


AOI in young children presents a challenging task for the orthotist with regard to halo fit. Guidelines for appropriate halo ring, vest, pin placement, and pin torque must be adhered to for successful clinical results. Meticulous follow-up care is also imperative to reduce and attempt to prevent complications and loss of anatomic reduction. The combination of superior pediatric halo component selection and fit, as well as close monitoring throughout the period of immobilization, can lead to success with conservative and postoperative halo management of AOI in children.


1.Davis D, Bohlman H, Walker E, et al. The pathological findings in fatal craniospinal injuries. J Neurosurg 1971;34:603–613.
2.Browner BD, Jupiter JB, Levine AM, Trafton PG. Skeletal Trauma. Philadelphia: W.B. Saunders; 1992.
3.Blackwood N. Atlo-occipital dislocation: a case of fracture of the atlas and axis, and forward dislocation of the occiput on the spinal column, life being maintained for thirty-four hours and forty minutes by artificial respiration, during which a laminectomy was performed upon the third cervical vertebrae. Ann Surg 1908;47:654–658.
4.Rose L, Muburak O, Saber AA, Plaisier B. Traumatic atlanto-occipital dislocation. Am J Surg 2003;185:376–377.
5.Matava MJ, Whitesides TE, Davis PC. Traumatic atlanto-occipital dislocation with survival. Spine 1993;18(13):1897–1903.
6.Morrissey RT, Weinstein SL. Lovell & Winter's Pediatric Orthopaedics, 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2001.
7.Hosono N, Yonenobu K, Kawagoe K, et al. Traumatic atlanto-occipital dislocation: a case report with survival. Spine 1993;18 (6):786–790.
8.Alker GJ, Oh YS Jr, Leslie EV. High cervical spine and craniocervical junction injuries in fatal traffic accidents: a radiologic study. Orthop Clin North Am 1978;9:1003–1010.
9.Bucholz RW, Burkhead WZ. The pathological anatomy of fatal atlanto-occipital dislocation. J Bone Joint Surg Am 1979;61:248–250.
10.Shamoun JM, Riddick L, Powell RW. Atlanto-occipital subluxation/dislocation: a ‘survivable' injury in children. Am Surg 1999;65(4):317–320.
11.Traynelis V, Marano G, Dunker R, Kaufman H. Traumatic atlanto-occipital dislocation. Case report. J Neurosurg 1986;65:863–870.
12.Cohen A, Hirsch M, Katz M, Sofer S. Traumatic atlanto-occipital dislocation in children: review and report of five cases. Pediatr Emerg Care 1991;7(1):24–27.
13.Powers B, Miller MD, Kramer RS, et al. Traumatic anterior atlanto-occipital dislocation. Neurosurgery 1979;4(1):12–17.
14.Collalto P, DeMuth W, Schwentker E, Boal D. Traumatic atlanto-occipital dislocation. Case report. J Bone Joint Surg Am 1979;68:1106–1109.
15.Dublin A, Marks W, Weinstock D, Newton T. Traumatic dislocation of the atlanto-occipital articulation (AOA) with short-term survival. Neurosurgery 1980;52A:1653–1660.
16.Gabrielson T, Maxwell J. Traumatic atlanto-occipital dislocation. Am J Roentgenol 1966;97:624–629.
17.Van Den Bout AH, Dommisse GF. Traumatic atlanto-occipital dislocation: Case report. Neurosurgery 1985;17:345–347.
18.Shkrum M, Green R, Nowack E. Upper cervical trauma in motor vehicle collisions. J Forensic Sci 1989;34:381–390.
19.Werne S. Studies in spontaneous atlas dislocation. Acta Orthop Scand Suppl 1957;23:1–150.
20.Houle P, McDonnell DE, Vender J. Traumatic atlanto-occipital dislocation in children. Pediatr Neurosurg 2001;34:193–197.
21.Farley FA, Graziano GP, Hensinger RN. Traumatic atlanto-occipital dislocation in a child. Spine 1992;17(12):1539–1541.
22.Johnson R, Hart D, Simmons E, et al. Cervical orthoses: a study comparing their effectiveness in restricting cervical motion in normal subjects. J Bone Joint Surg Am 1977;59:332–339.
23.Steinmetz MP, Verrees M, Anderson J S, Lechner RM. Dual-strap augmentation of a halo orthosis in the treatment of atlantooccipital dislocation in infants and young children. J Neurosurg 2002;96 Suppl 3:346–349.
24.Mubarak SJ, Camp JF, Vuletich W, et al. Halo application in the infant. J Pediatr Orthop 1989;9(5):612–614.
25.Ewald FC. Fracture of the odontoid process in a seventeen-month-old infant treated with a halo. J Bone Joint Surg Am 1971;53(8):1636–1640.
26.Mandabach M, Ruge JR, Hahn YS, McLone DG. Pediatric axis fractures: early halo immobilization, management and outcome. Pediatr Neurosurg 1993;19(5):225–232.
27.Letts M, Kaylor D, Gouw G. A biomechanical analysis of halo fixation in children. J Bone Joint Surg Br 1988;70(2):277–279.
28.Dormans JP, Criscitiello AA, Drummond DS, Davidson RS. Complications in children managed with immobilization in a halo vest. J Bone Joint Surg Am 1995;77(9):1370–1373.
29.Baum JA, Hanley EN, Pullekines J. Comparison of halo complications in adults and children. Spine 1989;14(3):251–252.
30.Garfin SR, Roux R, Botte MJ, et al. Skull osteology as it affects halo pin placement in children. J Pediatr Orthop 1986;6(4):434–436.
31.Wong WB, Haynes RJ. Osteology of the pediatric skull: considerations of halo pin placement. Spine 1994;19(13):1451–1454.
32.Garfin SR, Botte MJ, Centeno RS, Nickel VL. Osteology of the skull as it affects halo pin placement. Spine 1985;10(8):696–698.
33.Copley LA, Pepe MD, Tan V, et al. A comparative evaluation of halo pin designs in an immature skull model. Clin Orthop Relat Res 1998;357:212–218.
34.Mirza SK, Moquin RR, Anderson PA, et al. Stabilizing properties of the halo apparatus. Spine 1997;22(7):727–733.

atlanto-occipital dislocation; cervical; halo; immobilization; pediatric

© 2007 American Academy of Orthotists & Prosthetists