High-yield points
- Suspect it in any child in a high-energy crash, especially with a brain injury, cardiorespiratory arrest or lower cranial nerve or high cord deficits. Children survive it more often than adults .
- Arrest is common: every child with complete AOD in the SickKids series arrived in traumatic cardiac arrest (13 of 14 died), vs 1 of 10 with incomplete AOD (none died) . Arrest predicts death , but resuscitated children can recover .
- CT: the CCI is the key measurement.
- Normal is about 1.3 mm .
- Pang used 4 mm or more , but pediatric values vary with age. A screening threshold of sagittal 2.5 mm or more plus coronal 3.5 mm or more is 95% sensitive .
- Get an MRI when the mechanism fits, even if CT is borderline .
- Plain films: the BAI–BDI rule of 12 applies, but the BDI is unreliable under 13 . Powers' ratio is insensitive .
- Never use traction (10% neurological deterioration) .
- Treat with early occipitocervical fusion . Halo alone is a selected-case option .
- Screen the vessels (CTA): more than half have blunt cerebrovascular injury .
- After fusion: if the child declines, think hydrocephalus .
Also called atlanto-occipital dislocation or, when the injury spans occiput to C2, craniocervical dissociation. It is a ligamentous disruption of the craniocervical junction that was once thought almost uniformly fatal, but children now make up many of the survivors .
Diagrams are original schematics drawn for this text, not to scale.
Etiology
- High-energy trauma: most commonly a child passenger in a motor vehicle collision. In a 14-child series (mean age 5.2 years), 12 had associated injuries, 11 had a brain injury, and nearly half had a spinal cord injury . Another series of 15 children aged 8 months to 8 years were all passengers or pedestrians struck by vehicles .
- Why children: children are more susceptible because of their disproportionately large heads, ligamentous laxity, and injury mechanisms .
- Survival:
- Traynelis's 1986 review of survivors noted that most were children .
- Children and adolescents appear more likely than adults to survive, but with possible long-term neurological deficits .
- In a mostly adult 2026 multicenter series (mean age 41, range 9–86), in-hospital mortality was 36.8% (7 of 19). In a forensic series, AOD was found in 8.1% of 897 road-traffic deaths .
- Vascular injury: in 29 screened patients with craniocervical distraction injuries, more than half had blunt cerebrovascular injury, and 3 of those 15 patients (20%) had major strokes, vs none without vascular injury. The authors recommend screening everyone with C0–C1 or C1–C2 distraction injuries .
Presentation: cardiorespiratory arrest is common
- Arrest is the typical presentation of complete dislocation:
- In a SickKids series of 24 children with AOD, all 14 with complete AOD presented in traumatic cardiac arrest, and only 1 (7%) survived .
- Of the 10 with incomplete AOD, only 1 (10%) presented in arrest. None died, and most were managed without surgery .
- Overall, 15 of 24 (63%) presented in arrest .
- In the peer-reviewed SickKids series (2000–2020): 15 children had complete atlanto-occipital dislocation. All were hemodynamically unstable with signs of herniation, and 14 (93%) died. The 16 with subluxation were usually stable and survived with good outcomes .
- Dead on arrival: in a 16-child trauma-registry series (mean age 7.6 years), 8 were declared dead on arrival in the emergency department, and 11 were intubated in the field .
- Arrest predicts death: in 27 children with craniocervical dissociation, 10 (37%) died. Survivors needed significantly less CPR and had shorter transport times. Cardiac arrest, a high-velocity mechanism, a high Injury Severity Score and a low GCS were associated with mortality .
- A review of pediatric cases describes AOD as commonly presenting with cardiorespiratory arrest, spinal cord injury or cranial nerve deficits. Overall, 34% died, 41% survived with deficits and 25% recovered completely .
- Survival after arrest is possible: a 9-year-old resuscitated from cardiac arrest after a crash underwent occiput–C2 fusion and returned to independent walking .
- In the trauma bay: a lateral cervical radiograph during resuscitation showed the injury in 5 of 7 children with AOD who had one . The authors recommend lateral cervical radiography when resuscitating unstable children .
Classification
- Traynelis types (by direction of displacement) :
- Type I: anterior.
- Type II: longitudinal (distraction). The most common type in Astur's pediatric series (8 of 14) .
- Type III: posterior.
- Harris patterns: among 31 dissociations on lateral radiographs, 65% were combined anterior and distracted, 19% purely distracted, 13% purely anterior and 3% purely posterior .
- Severity: Bellabarba proposed an injury-severity classification with treatment implications .
Imaging
- CT with reformats and the condyle–C1 interval (CCI):
- The normal pediatric CCI is tiny and symmetric: a mean of 1.28 ± 0.26 mm in 89 children, with none above 1.95 mm, and it did not change with age .
- Pang's criterion of CCI 4 mm or more was 100% sensitive (16 of 16) and 100% specific (10 of 10 non-AOD injuries). The other tests performed far worse: Wholey's dens–basion interval 50%, Powers' ratio 37.5%, Harris' basion–axis interval 31% and Sun's interspinous ratio 25% sensitive .
- In 40 children who had occipitocervical fixation, the CCI had 100% sensitivity for AOD .
- But the CCI varies with age in later studies:
- In 124 subjects (2 days to 22 years), the CCI was largest at 2–4 years, coronal values were larger than sagittal, and a single threshold may not fit all ages .
- In 14 injured children, a 4.0 mm cutoff was only 36% sensitive (100% specific), while 2.5 mm was 93% sensitive but 18–100% specific. MRI is needed to make the diagnosis in children and young teens with the right mechanism .
- Screening threshold: in 597 pediatric trauma CTs, a sagittal CCI of 2.5 mm or more plus a coronal CCI of 3.5 mm or more was 95% sensitive and 73% specific (PPV 10.3%, NPV 99%) for AOD needing fusion. Younger children had larger CCIs .
- Plain-film measures (Harris "rule of 12") :
- In 400 adults, the basion–axial interval (BAI) and basion–dens interval (BDI) were both 12 mm or less, and the BAI was 12 mm or less in all 50 children aged 2–13 .
- The BDI is unreliable under age 13 because the dens ossifies at variable ages .
- The BAI–BDI method identified every dissociation, whereas the Powers ratio and X-line could not be applied in 46% of cases .
- The 2002 guidelines recommend a lateral cervical radiograph, with the BAI–BDI method if a measurement is used. Prevertebral soft-tissue swelling on a nondiagnostic film should prompt more imaging, and CT or MRI is recommended when suspicion persists .
- Diagnosis is often delayed: in 17 survivors, the injury was recognized on the initial lateral radiograph in only 12%, although the BDI was abnormal on review in 94%. An average 2-day delay was associated with profound neurological deterioration in 29% .
- MRI of the ligaments:
- Pang's indirect signs: tectorial membrane damage (71% sensitive), perimedullary blood (63%) and C1–C2 extra-axial blood (75%) .
- In 193 children who had both CT and MRI, 28% had ligamentous injury at the craniocervical junction. Single CT measurements were insensitive, but combining any abnormal measurement, fracture or hemorrhage gave 84–96% sensitivity and 78–95% specificity .
- A preserved tectorial membrane may be a favorable prognostic sign (3 patients with excellent outcomes despite alar, apical and cruciate disruption) .
Treatment
- Immobilize without traction:
- In the 2002 guidelines, traction carried a 10% risk of neurological deterioration .
- A pediatric review calls halo immobilization and traction contraindicated because they risk displacing the injured joint .
- Early occipitocervical fusion is standard:
- The guidelines recommend internal fixation and arthrodesis .
- All 14 children in Astur's series had posterior occipitocervical fusion, and all fused by 4–6 months .
- In Pang's series, 15 of 16 had occipitocervical fusion .
- Early stabilization protected against worsening cord injury: patients with useful motor function (ASIA D or E) rose from 41% before surgery to 76% after .
- Instrumentation in small children:
- Rigid screw-and-rod fixation in 15 children aged 8 months to 8 years was stable in 14 at a mean of 31 months, without deformity, growth disturbance or subaxial instability. The authors favor C2 translaminar screws .
- In 40 pediatric occipitocervical fixations (45% for trauma), all fused at 1 year, and complications occurred in 7.5%, including 1 vertebral artery injury .
- Halo alone in selected children: at SickKids, 8 of 10 children were treated with 3 months of halo alone, and none later needed fusion . This conflicts with the view that halo is contraindicated , so it is a selected-case option.
- After fixation: postoperative hydrocephalus occurred in 4 of 14 children. Suspect it if the child declines neurologically after fixation .
Outcomes
- In Pang's 16 children, mortality was 19% and 12% had severe residual high quadriplegia, but 69% had an excellent neurological recovery .
- Half of Astur's 14 children had neurological impairment at follow-up (mean 75 months) . Nearly half of pediatric survivors have residual deficits .
- In Keen's series, 4 of 7 children with weakness improved after stabilization, including one with quadriplegia .
Related topics
- Atlantoaxial (C1–2) Rotatory Subluxation: the other pediatric craniocervical injury.
- Spinal Cord Injury in Children: cord injury classification and prognosis.
- Subaxial Cervical Facet Dislocations: other cervical ligamentous injuries.
- Abusive Head Trauma: craniocervical ligamentous injury in inflicted injury.
- Cervical Spine Fractures in Children: age-specific fracture patterns and the PECARN imaging rule.
References
- Pang D, Nemzek WR, Zovickian J. Atlanto-occipital dislocation: part 1—normal occipital condyle-C1 interval in 89 children. Neurosurgery 2007;61(3):514–521. doi:10.1227/01.NEU.0000290897.77448.1F. PMID 17881963. verified
- Pang D, Nemzek WR, Zovickian J. Atlanto-occipital dislocation—part 2: the clinical use of (occipital) condyle-C1 interval, comparison with other diagnostic methods, and the manifestation, management, and outcome of atlanto-occipital dislocation in children. Neurosurgery 2007;61(5):995–1015. doi:10.1227/01.neu.0000303196.87672.78. PMID 18091277. verified
- Ravindra V, Riva-Cambrin J, Horn K, et al. A 2D threshold of the condylar-C1 interval to maximize identification of patients at high risk for atlantooccipital dislocation using computed tomography. J Neurosurg Pediatr 2017;19(4):458–463. doi:10.3171/2016.10.PEDS16459. PMID 28156214. verified
- Corcoran B, Linscott LL, Leach JL, et al. Application of normative occipital condyle-C1 interval measurements to detect atlanto-occipital injury in children. AJNR Am J Neuroradiol 2016;37(5):958–962. doi:10.3174/ajnr.A4641. PMID 26744446. verified
- Smith P, Linscott LL, Vadivelu S, et al. Normal development and measurements of the occipital condyle-C1 interval in children and young adults. AJNR Am J Neuroradiol 2016;37(5):952–957. doi:10.3174/ajnr.A4543. PMID 26514612. verified
- Harris JH Jr, Carson GC, Wagner LK. Radiologic diagnosis of traumatic occipitovertebral dissociation: 1. Normal occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994;162(4):881–886. doi:10.2214/ajr.162.4.8141012. PMID 8141012. verified
- Harris JH Jr, Carson GC, Wagner LK, et al. Radiologic diagnosis of traumatic occipitovertebral dissociation: 2. Comparison of three methods of detecting occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994;162(4):887–892. doi:10.2214/ajr.162.4.8141013. PMID 8141013. verified
- Traynelis VC, Marano GD, Dunker RO, et al. Traumatic atlanto-occipital dislocation. Case report. J Neurosurg 1986;65(6):863–870. doi:10.3171/jns.1986.65.6.0863. PMID 3772485. verified
- Astur N, Klimo P Jr, Sawyer JR, et al. Traumatic atlanto-occipital dislocation in children: evaluation, treatment, and outcomes. J Bone Joint Surg Am 2013;95(24):e194(1–8). doi:10.2106/JBJS.L.01295. PMID 24352780. verified
- Astur N, Sawyer JR, Klimo P Jr, et al. Traumatic atlanto-occipital dislocation in children. J Am Acad Orthop Surg 2014;22(5):274–282. doi:10.5435/JAAOS-22-05-274. PMID 24788443. verified
- Tubbs RS, Patel C, Loukas M, et al. Traumatic atlanto-occipital dislocation: do children and adolescents have better or worse outcomes than adults? A narrative review. Childs Nerv Syst 2016;32(8):1387–1392. doi:10.1007/s00381-016-3118-y. PMID 27226061. verified
- Bellabarba C, Mirza SK, West GA, et al. Diagnosis and treatment of craniocervical dislocation in a series of 17 consecutive survivors during an 8-year period. J Neurosurg Spine 2006;4(6):429–440. doi:10.3171/spi.2006.4.6.429. PMID 16776353. verified
- Hadley MN, Walters BC, Grabb PA, et al. Diagnosis and management of traumatic atlanto-occipital dislocation injuries. Neurosurgery 2002;50(3 Suppl):S105–S113. doi:10.1097/00006123-200203001-00018. PMID 12431294. verified
- Vilela MD, Kim LJ, Bellabarba C, et al. Blunt cerebrovascular injuries in association with craniocervical distraction injuries: a retrospective review of consecutive cases. Spine J 2015;15(3):499–505. doi:10.1016/j.spinee.2014.10.012. PMID 25463396. verified
- Kralik SF, Akbari Y, Huisman TAGM, et al. Accuracy of cervical spine CT for detection of craniocervical junction ligamentous injury in children. AJNR Am J Neuroradiol 2026 (online ahead of print). doi:10.3174/ajnr.A9576. PMID 42580862. verified
- Kimchi G, Greenberg G, Traynelis VC, et al. Integrity of the tectorial membrane is a favorable prognostic factor in atlanto-occipital dislocation. Br J Neurosurg 2020;34(4):470–474. doi:10.1080/02688697.2020.1761292. PMID 32368931. verified
- Abel TJ, Yan H, Canty M, et al. Traumatic atlanto-occipital dislocation in children: is external immobilization an option? Childs Nerv Syst 2021;37(1):177–183. doi:10.1007/s00381-020-04680-w. PMID 32519132. verified
- Keen JR, Ayer RE, Taha A, et al. Rigid internal fixation for traumatic cranio-cervical dissociation in infants and young children. Spine (Phila Pa 1976) 2019;44(1):17–24. doi:10.1097/BRS.0000000000002741. PMID 29894452. verified
- Martinez-Del-Campo E, Turner JD, Rangel-Castilla L, et al. Pediatric occipitocervical fixation: radiographic criteria, surgical technique, and clinical outcomes based on experience of a single surgeon. J Neurosurg Pediatr 2016;18(4):452–462. doi:10.3171/2016.2.PEDS15544. PMID 27286444. verified
- Ebken J, Alireza S, Küchler JN, et al. Survival and neurological outcomes after traumatic atlanto-occipital dislocation: a combined multicenter clinical and forensic study. Eur Spine J 2026;35(5):2604–2614. doi:10.1007/s00586-026-09821-4. PMID 41746342. verified
- Jeong D, Freire G, Beno S. 131 Characterizing atlanto-occipital dissociation and its association with pediatric traumatic cardiac arrest: a descriptive case series [meeting abstract]. Paediatr Child Health 2021;26(Suppl 1):e92–e93. doi:10.1093/pch/pxab061.105. No PMID. Full text (PMC) verified
- Jeong D, Beno S, Drake JM, et al. Clinical characteristics of upper cervical spine injuries in children: a retrospective case series. J Neurosurg Pediatr 2024;33(5):452–460. doi:10.3171/2023.12.PEDS23197. PMID 38335518. verified
- Hosalkar HS, Cain EL, Horn D, et al. Traumatic atlanto-occipital dislocation in children. J Bone Joint Surg Am 2005;87(11):2480–2488. doi:10.2106/JBJS.D.01897. PMID 16264124. verified
- Hazboun R, Muñoz A, Krafft PR, et al. Craniocervical dissociation in pediatric patients: pearls and pitfalls of diagnosis and management. Pediatr Emerg Care 2021;37(10):e602–e608. doi:10.1097/PEC.0000000000001721. PMID 30624426. verified
- Beez T, Brown J. Traumatic atlanto-occipital dislocation in children—a case-based update on clinical characteristics, management and outcome. Childs Nerv Syst 2017;33(1):27–33. doi:10.1007/s00381-016-3265-1. PMID 27757567. verified
Verification (2026-10-23): each reference's authors, title, journal, volume, pages, DOI and PMID were checked on PubMed, and every figure cited was checked against the PubMed abstract. Full texts were not reviewed. The Traynelis type definitions come from reference 9's abstract, because reference 8's abstract only says there are three types. The details of Bellabarba's severity classification are not in its abstract, so they are not given here. Reference 12's series and reference 14's series do not state ages in their abstracts and likely include adults, and reference 20 is mostly adult. Reference 21 is a meeting abstract without a PMID; its numbers were checked against the abstract text on PMC. It and reference 22 come from the same SickKids cohort, reported at different stages (24 vs 31 children with AOD), so their counts differ slightly. The overall arrest figure of 15 of 24 (63%) is this text's arithmetic from reference 21. No source gives a field-arrest rate for all pediatric AOD, so "presented in arrest" is reported as each source states it, without separating arrest in the field from arrest on arrival. The other "suspect it" cues (cranial nerve or high cord deficits) are this text's clinical gloss. The figures are original anatomical illustrations (redrawn 2026-11-03); the torn alar ligaments shown in the coronal figure are supported by reference 16's abstract (alar, apical and cruciate ligaments severely disrupted).