ORIGINAL ARTICLE
Spine Trauma in Very Young Children: A Retrospective Study of 206 Patients Presenting to a Level 1 Pediatric Trauma Center Jeffrey B. Knox, MD,* John E. Schneider, MD,w Jason M. Cage, DO,* Robert L. Wimberly, MD,z and Anthony I. Riccio, MDz Background: The immature spine has anatomic and biomechanical properties that differ from the adult spine and result in unique characteristics of pediatric spinal trauma. Although distinct patterns of spinal injury have been identified in children younger than 10 years of age, little research has explored the differing characteristics of spinal trauma within this age group, particularly in the very young. The purpose of this study is to identify differences in the epidemiology and characteristics of spinal trauma between children under the age of 4 years and those between 4 and 9 years of age. Methods: A review of all patients treated for spinal injury at a single large level I pediatric trauma center between 2003 and 2011 was conducted. Demographic data, injury mechanism, neurologic status, and details of any associated injuries were compiled. Radiographic studies were used to determine injury location and fracture classification. The patient population was divided into 2 groups: the infantile/toddler (IT) group (ages 0 to 3 y) and the young (Y) group (ages 4 to 9 y). Data were compared between these groups using the w2 test and the Student t test to identify differences in injury characteristics. Results: A total of 206 patients were identified. Fifty-seven patients were between 0 and 3 years of age and 149 were between 4 and 9 years old. Although motor vehicle collision was the most common cause of injury in both the groups, nonaccidental trauma was responsible for 19% of spine trauma among patients aged 0 to 3 years. Cervical spine injuries were much more common in the youngest patients (P < 0.05) with injuries primarily in the upper cervical spine. Children in the IT group were more likely to sustain ligamentous injuries, whereas Y patients had more compression fractures (P < 0.05). Neurologic injury was common in both the groups with IT patients more often presenting with complete loss of function or hemiplegia and Y patients sustaining more spinal cord injuries (P < 0.05). IT pa-
From the *Tripler Army Medical Center, Orthopedic Surgery Service, Honolulu, HI; wDepartment of Pediatrics, University of Texas Southwestern Medical Center and Childrens Medical Center; and zDepartment of Orthopedics, Texas Scottish Rite Hospital for Children and Children’s Medical Center, Dallas, TX. The views expressed in this manuscript are those of the authors and do not reflect the official policy or position of the Department of the Army, Department of Defense, or the US Government. The authors declare no conflicts of interest. Reprints: Anthony I. Riccio, MD, 1935 Medical District Dr, Dallas, TX 75235. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins
698 | www.pedorthopaedics.com
tients had a 25% mortality rate, which was significantly higher than that of the Y group (P = 0.005). Conclusions: This study shows many significant differences in characteristics of spinal injury in infants/toddlers when compared with older children. These differences can help guide diagnostic evaluation and initial management, as well as future prevention efforts. Level of Evidence: Level III. Key Words: spine, trauma, cervical, fracture, infant, pediatric (J Pediatr Orthop 2014;34:698–702)
S
pine trauma in children represents a rare and potentially devastating diagnosis. The immature spine has anatomic and biomechanical properties that differ from that of the adult1–4 and, subsequently, differences in injury patterns are expected.5,6 The spine completes maturation by 8 to 10 years of age after which injury patterns mirror those of an adult spine.7 There is a paucity of literature focused upon the characteristics of spinal trauma in young children. Ruge et al6 suggested that children under 36 months of age are a distinct population with differences in location of injury, sex distribution, and treatment expectations. Despite being commonly cited, this study was limited as it included only 9 patients under 4 years of age. The purpose of our study was to compare the epidemiology and characteristics of spinal column trauma between children under the age of 4 years and children between 4 and 9 years of age.
METHODS We performed an IRB approved, retrospective review of all patients treated for spinal injury at a large level I pediatric trauma center. The medical records and trauma registry were queried for patients with spinal trauma sustained between the years of 2003 and 2011. Spinal injury was defined as injury to the vertebral column or neural elements between the occipitocervical and lumbosacral junctions. Minor injuries such as contusions or strains were not included for analysis. Pathologic fractures and patients with preexisting vertebral anomalies or spinal deformities were excluded. Only patients younger than 10 years of age at the time of presentation were included.
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
Demographic data to include age and sex was documented. Levels of injury and injury patterns were determined from the medical records and radiographic imaging reports. If the specific pattern could not be determined from the existing record or if conflicting information was identified, radiographs were reviewed by the senior author. We recorded the mechanism of injury, neurologic status on arrival, and changes in neurologic status noted during an inpatient stay. Associated extremity and visceral and head trauma were also documented. Minor injuries such as abrasions, sprains, and minor lacerations were not included as an associated injury. For those patients who were triaged at an outside hospital, medical records from the initial institution were reviewed for additional information including injury data, neurologic status, and treatment rendered before transfer. The Revised Pediatric Trauma Score (RTS) was also obtained from the medical records as documented by the initial treating provider. This scoring system includes the Pediatric Trauma Score plus the Glasgow Coma Score (GCS) converted to a 4-point scale. The Pediatric Trauma Score represents a well-validated system that incorporates patient size, airway, blood pressure, level of consciousness, fractures, and soft tissue injury. This system has been correlated with morbidity and mortality in pediatric trauma patients.8,9 Addition of the converted GCS to the Pediatric Trauma Score generates the RTS with a range from 6 to 16. All patients presenting to our trauma center are assigned an RTS as a matter of routine during the initial trauma evaluation. If the scoring worksheet was incomplete, available data from the initial presentation was used to calculate the appropriate score if possible. The patient population was divided into 2 groups, the infantile/toddler (IT) group (ages 0 to 3 y) and the young children (Y) group (ages 4 to 9 y). Statistical analysis was performed to compare the 2 groups using the w2 test and the Student t test with the exception of the RTS. As age is a factor in this scoring system, comparisons would be inherently biased and not particularly useful. Significance was defined as a P value of 0.05).
Volume 34, Number 7, October/November 2014
juries were identified, as the absent function could not be verified to be the result of head trauma, spinal cord trauma, or a combination of both. Incomplete spinal cord injury was present in 3/57 (5%) of IT and 24/149 (16%) of Y patients (P < 0.04). Seven of the 57 (14%) of IT and 4/149 (3%) of Y patients presented with hemiplegia (P = 0.01) and 2 (1%) Y patients had focal cranial nerve deficits on initial presentation (P = 1.0).
Treatment The majority of spinal injuries in both the age groups were managed nonsurgically with treatments ranging from observation and activity modification to bracing and casting. Five percent of IT and 2% of Y patients were placed into Halo vest immobilization. Surgical intervention with decompression and/or fusion and instrumentation was selected for only 9% of IT and 15% of Y patients (P > 0.05).
Associated Injuries
The majority of patients appeared neurologically intact; however, evaluation of neurologic injuries was difficult in many cases because of the frequency of associated head trauma. Sixteen of 57 (27%) of IT and 6/149 (4%) of Y patients had no demonstrable neurologic function on initial evaluation. Such patients were often unconscious on arrival with severe neurologic impairment due to either head trauma or high cervical spinal cord injury. It is the protocol of our trauma service to obtain both head and cervical MRI on all children admitted with a suspected head injury and associated neurologic abnormality. Despite the use of such imaging, the exact etiology of a neurologic deficit was frequently indeterminate when concomitant head and spinal cord in-
Patients in this study demonstrated a high rate of associated injuries consistent with high-energy trauma. Among all patients, the average revised trauma score was 9 (range, 3 to16). The most common associated injury in both the age groups was head trauma, which was present in 36/57 (63%) of IT and 57/149 (43%) of Y patients (P > 0.5). These injuries were often severe with 56% of IT and 45% of Y head injury patients presenting with a GCS of 3. Facial trauma was common in the Y group occurring in 23/149 (16%) of patients, compared with only 3/57 (5%) in the IT group (P < 0.05). Pelvic/extremity fractures were also more common in the Y group (37%) versus 18% in the IT group (P = 0.01) as were hollow viscus injuries (16% Y vs. 2% IT) (P < 0.01). Solid viscus injury rates, however, were similar between the groups (11% IT vs. 17% Y group) (P = 0.3). Thoracic trauma was also commonly present in both the groups, occurring in 14/57 (25%) of IT and 45/149 (32%) of Y patients (P = 0.3). Vascular trauma was uncommon in both groups, occurring in 2% of both the IT and Y group. The presence of abdominal wall injury (seatbelt sign) was significantly more common in the Y group with 32/149 (20%) demonstrating that injury versus 4/57 (7%) in the IT group (P = 0.03).
TABLE 2. Injury Type by Age
Mortality
Neurologic Status
n (%)
Isolated ligamentous Compression Chance Burst SCIWORA Dens Hangman Spinous process Other
Infantile Group (N = 57)
Young Group (N = 149)
P
29 (51)
51 (34)
< 0.04
51 23 3 5 5 2 9
< 0.005 NS NS NS NS NS NS NS
7 4 1 2 3 1 1 9
(12) (7) (2) (4) (5) (2) (2) (15)
(34) (15) (2) (3) (3) (1) (6) 0
SCIWORA indicates Spinal Cord Injury Without Radiographic Abnormality.
700 | www.pedorthopaedics.com
Although both groups demonstrated high mortality rates, death was significantly more likely in the IT group. Patients in this group had a 25% in-hospital mortality compared with 9% in the young age group (P = 0.005).
Hospital Stay Patients in both the groups often required prolonged hospital stays with an average inpatient stay of 10 days in the IT group (range, 0 to 82) and 12.4 days in the Y group (range, 0 to 365).
DISCUSSION In this series, we identified age-related differences among children with spinal column injuries when comparing r
2014 Lippincott Williams & Wilkins
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
children under 4 years of age with an older cohort of 4 to 9 years of age. Mechanism of injury was similar between the age groups studied, with MVC being the dominant mechanism followed by falls and motor vehicle versus pedestrian injuries. Prior series of pediatric spine trauma have demonstrated similar injury mechanisms.5,7,10–16 In these series, MVC and falls were the primary mechanism in the younger children, whereas in older children/adolescents sports injuries begin to take a dominant role.11,15,17 Nonaccidental trauma caused spinal injury in 19% of all patients younger than 4 years of age in our study. The characteristics of spinal injuries in this subset of patients will be examined in a subsequent study. As expected, there was a high rate of MRI use in the evaluation of acute spine trauma in our patients. The younger cohort was statistically more likely to have had this study performed. This likely is related to the difficulty in obtaining reliable neurologic examinations in infants and toddlers. In addition, these patients demonstrated a high rate of closed head injuries, which further obscured neurologic examination. Frank et al18 showed MRI to be an effective and cost-effective method of cervical spine clearance in the unconscious pediatric trauma patient allowing for early cervical spine clearance. In this age group, a significant portion of the spinal column is composed of cartilage, which is better imaged with MRI. Sledge et al19 demonstrated MRI to be an effective tool in the diagnosis and classification of pediatric thoracolumbar injuries. Anderson et al20 recommended the use of MRI as a part of a cervical spine clearance protocol in children between 0 and 3 years of age who either demonstrate evidence of neurologic injury or a high-energy mechanism. Although injuries at all levels were seen, injuries at the craniovertebral junction were significantly more common in the infant/toddler age group. The youngest patients are anatomically at increased risk of upper cervical spine trauma. Because of an increased cranium to body weight ratio, the fulcrum of rotation during neck flexion occurs at about C2-3. This fulcrum descends to the lower cervical spine with growth. Younger children also have more horizontally oriented facet joints. These anatomic differences gradually diminish as spine anatomy nears adult morphology. Injuries to the cervical spine have been shown to represent up to 79% of spinal injuries in children under 8 years of age,16 with up to 83% located in the upper cervical spine.13 Patel et al21 also showed young children to be at higher risk of upper cervical spine injuries with a bimodal distribution with peaks at both ages of 3 and 16 years. Bilston and Brown10 compared the location of serious spine injuries between children aged 0 and 4 years and 5 and 8 years and demonstrated the youngest group to have predominantly cervical spine injuries (83%) with thoracic and lumbar injuries each comprising 8%. In the 4- to 8-year-old age group, cervical spine injuries were still the most common (62%) and lumbar spine was the second most common location (24%) followed by the thoracic spine (10%). Although our study found similarly high rates of cervical spine injury in the infant/toddler r
2014 Lippincott Williams & Wilkins
Spine Trauma in Very Young Children
group, we found a relatively equal distribution of injury between cervical, thoracic, and lumbar spine injuries in the older group. This study identified higher rates of ligamentous injury in the younger age groups. Conversely, there was an increased incidence of compression fractures in older patients. Nitecki and Moir15 also found young children to have higher rates of ligamentous injuries and subluxations, whereas older children were more likely to sustain fracture-dislocation injuries. The etiology behind this age-dependent difference is likely multifactorial with a contribution from both biomechanical differences in the spinal column as well as differences in injury mechanisms. Mahan et al13 found multi-level spinal injuries to be much more common in older children with a 35% incidence after 8 years of age compared with 19% in those younger than 8 years. The majority of these injuries were contiguous with only 6% of multi-level injures occurring at noncontiguous levels. Firth et al,22 however, demonstrated noncontiguous spinal injuries in 11.8% of their series of pediatric spine trauma. Multi-level spine trauma is also associated with as greater likelihood of SCI, suggesting a higher energy mechanism of injury. In our series, approximately 40% of patients presented with multi-level injuries and, similar to prior studies, the majority of these injuries were contiguous. Our series demonstrated a high rate of neurologic injury with 46% of IT and 26% of Y patients demonstrating some neurologic dysfunction. Full characterization of type and degree of neurologic injury was complicated by the high rate of associated head injury in our patient population. Due, in part, to this high incidence of head trauma, 27% of children younger than 4 years of age and 5% of children in the older age group had no demonstrable neurologic function on arrival. Prior studies have shown variability in the rate of neurologic injury. Mahan et al13 found neurologic symptoms to be present in only 12% of patients, whereas Osenbach and Menezes16 found 52% of patients to have some degree of neurologic injury despite similar injury patterns between the 2 studies. Carreon et al7 also demonstrated that children under 10 years of age were at an increased risk of neurologic injury compared with older children. It should be noted that the current definition of SCIWORA has become increasingly ambiguous with the common use of MRI for evaluation of ligamentous injuries in children with neurologic deficits in the setting of normal plain radiographs. To avoid confusion with the definition of SCIWORA, the newer literature has recommended reserving this term for patients with no abnormalities on plain radiographs or MRI.23 We chose to use this stricter definition for our study, thereby preventing the classification of isolated ligamentous injuries as SCIWORA and allowing us to more accurately classify injury patterns. Spine trauma in children represents high-energy trauma and is often associated with multiple, often severe, associated injuries. The most common associated injury in this population is head trauma. This is most often associated with cervical spine trauma with head injury rates www.pedorthopaedics.com |
701
J Pediatr Orthop
Knox et al
ranging from 30%24 to 66%.5,17,25-27 Although head injuries are also commonly present in patients with thoracolumbar trauma, such patients are more likely to sustain abdominal injuries.28,29 Patients in our study sustained multiple associated injuries with head injury representing the most common type in both age groups. These patients often sustained severe head trauma with GCS of 3. Significant differences in injury rates were seen in other injury locations. Infants/toddlers had significantly lower rates of facial trauma, associated pelvis and extremity fractures, as well as hollow viscus injuries. These children were also significantly less likely to present with a seatbelt sign. These differences likely relate to the differences in restraint use, as these patients were more likely to be properly restrained, thus potentially providing protection against these injuries. High mortality rate after spine injury has been demonstrated in prior studies with the highest mortality rates associated with cervical spinal cord injuries as well as upper cervical spine fractures.15,30 This increased mortality is likely related to the different injury patterns with a predilection for upper cervical spine injuries. AO injuries, in particular, have been associated with very high mortality rates with rates up to 100% reported.15 Our study confirmed these high rates of mortality, especially in the younger cohort of patients who were significantly more likely to expire during initial resuscitation or subsequent hospital admission. Children in the youngest age group had a 25% in-hospital mortality compared with 9% in our older cohort. Only 14% of IT and 18% of Y patients were treated with surgical intervention. Although some studies have shown similar rates of 18%,7,13 this need for operative intervention approached 30% in multiple series.5,16,30 The reason behind this decreased incidence of surgical intervention may be related, in part, to differences in practice patterns. In addition, our high rate of MRI use may likely have led to an increased rate of detection of more minor stable injuries for which surgery was not required. In addition, Ruge et al6 found very young children to need surgical intervention more often than children aged 4 to 12 years. Our series failed to identify any significant differences between the 2 age groups. In summary, although infants/toddlers demonstrated many similarities compared with young children, many important differences were identified. The infants/ toddlers were more likely to sustain injury to the upper cervical spine including the AO junction, were more likely to sustain purely ligamentous injuries versus compression or chance injuries, and were more often evaluated using MRI. These patients were also less likely to sustain concomitant injuries to the face, hollow viscus, or have associated fractures and demonstrated a much higher mortality rate compared with older patients. Infants/ toddlers were also found to have a high rate of spinal injury secondary to nonaccidental trauma in our series. REFERENCES 1. Kumaresan S, Yoganandan N, Pintar FA, et al. Biomechanical study of pediatric human cervical spine: a finite element approach. J Biomech Eng. 2000;122:60–71.
702 | www.pedorthopaedics.com
Volume 34, Number 7, October/November 2014
2. Lustrin ES, Karakas SP, Ortiz AO, et al. Pediatric cervical spine: normal anatomy, variants, and trauma. Radiographics. 2003;23:539–560. 3. Nuckley DJ, Eck MP, Carter JW, et al. Spinal maturation affects vertebral compressive mechanics and vBMD with sex dependence. Bone. 2004;35:720–728. 4. Ouyang J, Zhu Q, Zhao W, et al. Biomechanical assessment of the pediatric cervical spine under bending and tensile loading. Spine. 2005;30:E716–E723. 5. Eleraky MA, Theodore N, Adams M, et al. Pediatric cervical spine injuries: report of 102 cases and review of the literature. J Neurosurg. 2000;92:12–17. 6. Ruge JR, Sinson GP, McLone DG, et al. Pediatric spinal injury: the very young. J Neurosurg. 1988;68:25–30. 7. Carreon LY, Glassman SD, Campbell MJ. Pediatric spine fractures: a review of 137 hospital admissions. J Spinal Disord Tech. 2004;17:477–482. 8. Narci A, Solak O, Turhan-Haktanir N, et al. The prognostic importance of trauma scoring systems in pediatric patients. Pediatr Surg Int. 2009;25:25–30. 9. Tepas JJ IIIrd, Mollitt DL, Talbert JL, et al. The pediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg. 1987;22:14–18. 10. Bilston LE, Brown J. Pediatric spinal injury type and severity are age and mechanism dependent. Spine. 2007;32:2339–2347. 11. Cirak B, Ziegfeld S, Knight VM, et al. Spinal injuries in children. J Pediatr Surg. 2004;39:607–612. 12. Kokoska ER, Keller MS, Rallo MC, et al. Characteristics of pediatric cervical spine injuries. J Pediatr Surg. 2001;36:100–105. 13. Mahan ST, Mooney DP, Karlin LI, et al. Multiple level injuries in pediatric spinal trauma. J Trauma. 2009;67:537–542. 14. McGrory BJ, Klassen RA, Chao EY, et al. Acute fractures and dislocations of the cervical spine in children and adolescents. J Bone Joint Surg. 1993;75:988–995. 15. Nitecki S, Moir CR. Predictive factors of the outcome of traumatic cervical spine fracture in children. J Pediatr Surg. 1994;29:1409–1411. 16. Osenbach RK, Menezes AH. Pediatric spinal cord and vertebral column injury. Neurosurg. 1992;30:385–390. 17. Platzer P, Jaindl M, Thalhammer G, et al. Cervical spine injuries in pediatric patients. J Trauma. 2007;62:389–396; discussion 394-386. 18. Frank JB, Lim CK, Flynn JM, et al. The efficacy of magnetic resonance imaging in pediatric cervical spine clearance. Spine. 2002;27:1176–1179. 19. Sledge JB, Allred D, Hyman J. Use of magnetic resonance imaging in evaluating injuries to the pediatric thoracolumbar spine. J Pediatr Orthop. 2001;21:288–293. 20. Anderson RC, Kan P, Vanaman M, et al. Utility of a cervical spine clearance protocol after trauma in children between 0 and 3 years of age. J Neurosurg Pediatr. 2010;5:292–296. 21. Patel JC, Tepas JJ IIIrd, Mollitt DL, et al. Pediatric cervical spine injuries: defining the disease. J Pediatr Surg. 2001;36:373–376. 22. Firth GB, Kingwell S, Moroz P. Pediatric non-contiguous spinal injuries: the 15 year experience at a level 1 trauma center. Spine. 2012;37:E599–E608. 23. Yucesoy K, Yuksel KZ. SCIWORA in MRI era. Clin Neurol Neurosurg. 2008;110:429–433. 24. Orenstein JB, Klein BL, Gotschall CS, et al. Age and outcome in pediatric cervical spine injury: 11-year experience. Pediatr Emerg Care. 1994;10:132–137. 25. Birney TJ, Hanley EN Jr. Traumatic cervical spine injuries in childhood and adolescence. Spine. 1989;14:1277–1282. 26. Henrys P, Lyne ED, Lifton C, et al. Clinical review of cervical spine injuries in children. Clin Orthop Related Res. 1977;129:172–176. 27. Michael DB, Guyot DR, Darmody WR. Coincidence of head and cervical spine injury. J Neurotrauma. 1989;6:177–189. 28. Arkader A, Warner WC Jr, Tolo VT, et al. Pediatric Chance fractures: a multicenter perspective. J Pediatr Orthop. 2011;31:741–744. 29. Beaunoyer M, St-Vil D, Lallier M, et al. Abdominal injuries associated with thoraco-lumbar fractures after motor vehicle collision. J Pediatr Surg. 2001;36:760–762. 30. Puisto V, Kaariainen S, Impinen A, et al. Incidence of spinal and spinal cord injuries and their surgical treatment in children and adolescents: a population-based study. Spine. 2010;35:104–107.
r
2014 Lippincott Williams & Wilkins
Spine Trauma in Very Young Children: A Retrospective Study of 206 Patients Presenting to a Level 1 Pediatric Trauma Center Jeffrey B. Knox, MD,* John E. Schneider, MD,w Jason M. Cage, DO,* Robert L. Wimberly, MD,z and Anthony I. Riccio, MDz Background: The immature spine has anatomic and biomechanical properties that differ from the adult spine and result in unique characteristics of pediatric spinal trauma. Although distinct patterns of spinal injury have been identified in children younger than 10 years of age, little research has explored the differing characteristics of spinal trauma within this age group, particularly in the very young. The purpose of this study is to identify differences in the epidemiology and characteristics of spinal trauma between children under the age of 4 years and those between 4 and 9 years of age. Methods: A review of all patients treated for spinal injury at a single large level I pediatric trauma center between 2003 and 2011 was conducted. Demographic data, injury mechanism, neurologic status, and details of any associated injuries were compiled. Radiographic studies were used to determine injury location and fracture classification. The patient population was divided into 2 groups: the infantile/toddler (IT) group (ages 0 to 3 y) and the young (Y) group (ages 4 to 9 y). Data were compared between these groups using the w2 test and the Student t test to identify differences in injury characteristics. Results: A total of 206 patients were identified. Fifty-seven patients were between 0 and 3 years of age and 149 were between 4 and 9 years old. Although motor vehicle collision was the most common cause of injury in both the groups, nonaccidental trauma was responsible for 19% of spine trauma among patients aged 0 to 3 years. Cervical spine injuries were much more common in the youngest patients (P < 0.05) with injuries primarily in the upper cervical spine. Children in the IT group were more likely to sustain ligamentous injuries, whereas Y patients had more compression fractures (P < 0.05). Neurologic injury was common in both the groups with IT patients more often presenting with complete loss of function or hemiplegia and Y patients sustaining more spinal cord injuries (P < 0.05). IT pa-
From the *Tripler Army Medical Center, Orthopedic Surgery Service, Honolulu, HI; wDepartment of Pediatrics, University of Texas Southwestern Medical Center and Childrens Medical Center; and zDepartment of Orthopedics, Texas Scottish Rite Hospital for Children and Children’s Medical Center, Dallas, TX. The views expressed in this manuscript are those of the authors and do not reflect the official policy or position of the Department of the Army, Department of Defense, or the US Government. The authors declare no conflicts of interest. Reprints: Anthony I. Riccio, MD, 1935 Medical District Dr, Dallas, TX 75235. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins
698 | www.pedorthopaedics.com
tients had a 25% mortality rate, which was significantly higher than that of the Y group (P = 0.005). Conclusions: This study shows many significant differences in characteristics of spinal injury in infants/toddlers when compared with older children. These differences can help guide diagnostic evaluation and initial management, as well as future prevention efforts. Level of Evidence: Level III. Key Words: spine, trauma, cervical, fracture, infant, pediatric (J Pediatr Orthop 2014;34:698–702)
S
pine trauma in children represents a rare and potentially devastating diagnosis. The immature spine has anatomic and biomechanical properties that differ from that of the adult1–4 and, subsequently, differences in injury patterns are expected.5,6 The spine completes maturation by 8 to 10 years of age after which injury patterns mirror those of an adult spine.7 There is a paucity of literature focused upon the characteristics of spinal trauma in young children. Ruge et al6 suggested that children under 36 months of age are a distinct population with differences in location of injury, sex distribution, and treatment expectations. Despite being commonly cited, this study was limited as it included only 9 patients under 4 years of age. The purpose of our study was to compare the epidemiology and characteristics of spinal column trauma between children under the age of 4 years and children between 4 and 9 years of age.
METHODS We performed an IRB approved, retrospective review of all patients treated for spinal injury at a large level I pediatric trauma center. The medical records and trauma registry were queried for patients with spinal trauma sustained between the years of 2003 and 2011. Spinal injury was defined as injury to the vertebral column or neural elements between the occipitocervical and lumbosacral junctions. Minor injuries such as contusions or strains were not included for analysis. Pathologic fractures and patients with preexisting vertebral anomalies or spinal deformities were excluded. Only patients younger than 10 years of age at the time of presentation were included.
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
Demographic data to include age and sex was documented. Levels of injury and injury patterns were determined from the medical records and radiographic imaging reports. If the specific pattern could not be determined from the existing record or if conflicting information was identified, radiographs were reviewed by the senior author. We recorded the mechanism of injury, neurologic status on arrival, and changes in neurologic status noted during an inpatient stay. Associated extremity and visceral and head trauma were also documented. Minor injuries such as abrasions, sprains, and minor lacerations were not included as an associated injury. For those patients who were triaged at an outside hospital, medical records from the initial institution were reviewed for additional information including injury data, neurologic status, and treatment rendered before transfer. The Revised Pediatric Trauma Score (RTS) was also obtained from the medical records as documented by the initial treating provider. This scoring system includes the Pediatric Trauma Score plus the Glasgow Coma Score (GCS) converted to a 4-point scale. The Pediatric Trauma Score represents a well-validated system that incorporates patient size, airway, blood pressure, level of consciousness, fractures, and soft tissue injury. This system has been correlated with morbidity and mortality in pediatric trauma patients.8,9 Addition of the converted GCS to the Pediatric Trauma Score generates the RTS with a range from 6 to 16. All patients presenting to our trauma center are assigned an RTS as a matter of routine during the initial trauma evaluation. If the scoring worksheet was incomplete, available data from the initial presentation was used to calculate the appropriate score if possible. The patient population was divided into 2 groups, the infantile/toddler (IT) group (ages 0 to 3 y) and the young children (Y) group (ages 4 to 9 y). Statistical analysis was performed to compare the 2 groups using the w2 test and the Student t test with the exception of the RTS. As age is a factor in this scoring system, comparisons would be inherently biased and not particularly useful. Significance was defined as a P value of 0.05).
Volume 34, Number 7, October/November 2014
juries were identified, as the absent function could not be verified to be the result of head trauma, spinal cord trauma, or a combination of both. Incomplete spinal cord injury was present in 3/57 (5%) of IT and 24/149 (16%) of Y patients (P < 0.04). Seven of the 57 (14%) of IT and 4/149 (3%) of Y patients presented with hemiplegia (P = 0.01) and 2 (1%) Y patients had focal cranial nerve deficits on initial presentation (P = 1.0).
Treatment The majority of spinal injuries in both the age groups were managed nonsurgically with treatments ranging from observation and activity modification to bracing and casting. Five percent of IT and 2% of Y patients were placed into Halo vest immobilization. Surgical intervention with decompression and/or fusion and instrumentation was selected for only 9% of IT and 15% of Y patients (P > 0.05).
Associated Injuries
The majority of patients appeared neurologically intact; however, evaluation of neurologic injuries was difficult in many cases because of the frequency of associated head trauma. Sixteen of 57 (27%) of IT and 6/149 (4%) of Y patients had no demonstrable neurologic function on initial evaluation. Such patients were often unconscious on arrival with severe neurologic impairment due to either head trauma or high cervical spinal cord injury. It is the protocol of our trauma service to obtain both head and cervical MRI on all children admitted with a suspected head injury and associated neurologic abnormality. Despite the use of such imaging, the exact etiology of a neurologic deficit was frequently indeterminate when concomitant head and spinal cord in-
Patients in this study demonstrated a high rate of associated injuries consistent with high-energy trauma. Among all patients, the average revised trauma score was 9 (range, 3 to16). The most common associated injury in both the age groups was head trauma, which was present in 36/57 (63%) of IT and 57/149 (43%) of Y patients (P > 0.5). These injuries were often severe with 56% of IT and 45% of Y head injury patients presenting with a GCS of 3. Facial trauma was common in the Y group occurring in 23/149 (16%) of patients, compared with only 3/57 (5%) in the IT group (P < 0.05). Pelvic/extremity fractures were also more common in the Y group (37%) versus 18% in the IT group (P = 0.01) as were hollow viscus injuries (16% Y vs. 2% IT) (P < 0.01). Solid viscus injury rates, however, were similar between the groups (11% IT vs. 17% Y group) (P = 0.3). Thoracic trauma was also commonly present in both the groups, occurring in 14/57 (25%) of IT and 45/149 (32%) of Y patients (P = 0.3). Vascular trauma was uncommon in both groups, occurring in 2% of both the IT and Y group. The presence of abdominal wall injury (seatbelt sign) was significantly more common in the Y group with 32/149 (20%) demonstrating that injury versus 4/57 (7%) in the IT group (P = 0.03).
TABLE 2. Injury Type by Age
Mortality
Neurologic Status
n (%)
Isolated ligamentous Compression Chance Burst SCIWORA Dens Hangman Spinous process Other
Infantile Group (N = 57)
Young Group (N = 149)
P
29 (51)
51 (34)
< 0.04
51 23 3 5 5 2 9
< 0.005 NS NS NS NS NS NS NS
7 4 1 2 3 1 1 9
(12) (7) (2) (4) (5) (2) (2) (15)
(34) (15) (2) (3) (3) (1) (6) 0
SCIWORA indicates Spinal Cord Injury Without Radiographic Abnormality.
700 | www.pedorthopaedics.com
Although both groups demonstrated high mortality rates, death was significantly more likely in the IT group. Patients in this group had a 25% in-hospital mortality compared with 9% in the young age group (P = 0.005).
Hospital Stay Patients in both the groups often required prolonged hospital stays with an average inpatient stay of 10 days in the IT group (range, 0 to 82) and 12.4 days in the Y group (range, 0 to 365).
DISCUSSION In this series, we identified age-related differences among children with spinal column injuries when comparing r
2014 Lippincott Williams & Wilkins
J Pediatr Orthop
Volume 34, Number 7, October/November 2014
children under 4 years of age with an older cohort of 4 to 9 years of age. Mechanism of injury was similar between the age groups studied, with MVC being the dominant mechanism followed by falls and motor vehicle versus pedestrian injuries. Prior series of pediatric spine trauma have demonstrated similar injury mechanisms.5,7,10–16 In these series, MVC and falls were the primary mechanism in the younger children, whereas in older children/adolescents sports injuries begin to take a dominant role.11,15,17 Nonaccidental trauma caused spinal injury in 19% of all patients younger than 4 years of age in our study. The characteristics of spinal injuries in this subset of patients will be examined in a subsequent study. As expected, there was a high rate of MRI use in the evaluation of acute spine trauma in our patients. The younger cohort was statistically more likely to have had this study performed. This likely is related to the difficulty in obtaining reliable neurologic examinations in infants and toddlers. In addition, these patients demonstrated a high rate of closed head injuries, which further obscured neurologic examination. Frank et al18 showed MRI to be an effective and cost-effective method of cervical spine clearance in the unconscious pediatric trauma patient allowing for early cervical spine clearance. In this age group, a significant portion of the spinal column is composed of cartilage, which is better imaged with MRI. Sledge et al19 demonstrated MRI to be an effective tool in the diagnosis and classification of pediatric thoracolumbar injuries. Anderson et al20 recommended the use of MRI as a part of a cervical spine clearance protocol in children between 0 and 3 years of age who either demonstrate evidence of neurologic injury or a high-energy mechanism. Although injuries at all levels were seen, injuries at the craniovertebral junction were significantly more common in the infant/toddler age group. The youngest patients are anatomically at increased risk of upper cervical spine trauma. Because of an increased cranium to body weight ratio, the fulcrum of rotation during neck flexion occurs at about C2-3. This fulcrum descends to the lower cervical spine with growth. Younger children also have more horizontally oriented facet joints. These anatomic differences gradually diminish as spine anatomy nears adult morphology. Injuries to the cervical spine have been shown to represent up to 79% of spinal injuries in children under 8 years of age,16 with up to 83% located in the upper cervical spine.13 Patel et al21 also showed young children to be at higher risk of upper cervical spine injuries with a bimodal distribution with peaks at both ages of 3 and 16 years. Bilston and Brown10 compared the location of serious spine injuries between children aged 0 and 4 years and 5 and 8 years and demonstrated the youngest group to have predominantly cervical spine injuries (83%) with thoracic and lumbar injuries each comprising 8%. In the 4- to 8-year-old age group, cervical spine injuries were still the most common (62%) and lumbar spine was the second most common location (24%) followed by the thoracic spine (10%). Although our study found similarly high rates of cervical spine injury in the infant/toddler r
2014 Lippincott Williams & Wilkins
Spine Trauma in Very Young Children
group, we found a relatively equal distribution of injury between cervical, thoracic, and lumbar spine injuries in the older group. This study identified higher rates of ligamentous injury in the younger age groups. Conversely, there was an increased incidence of compression fractures in older patients. Nitecki and Moir15 also found young children to have higher rates of ligamentous injuries and subluxations, whereas older children were more likely to sustain fracture-dislocation injuries. The etiology behind this age-dependent difference is likely multifactorial with a contribution from both biomechanical differences in the spinal column as well as differences in injury mechanisms. Mahan et al13 found multi-level spinal injuries to be much more common in older children with a 35% incidence after 8 years of age compared with 19% in those younger than 8 years. The majority of these injuries were contiguous with only 6% of multi-level injures occurring at noncontiguous levels. Firth et al,22 however, demonstrated noncontiguous spinal injuries in 11.8% of their series of pediatric spine trauma. Multi-level spine trauma is also associated with as greater likelihood of SCI, suggesting a higher energy mechanism of injury. In our series, approximately 40% of patients presented with multi-level injuries and, similar to prior studies, the majority of these injuries were contiguous. Our series demonstrated a high rate of neurologic injury with 46% of IT and 26% of Y patients demonstrating some neurologic dysfunction. Full characterization of type and degree of neurologic injury was complicated by the high rate of associated head injury in our patient population. Due, in part, to this high incidence of head trauma, 27% of children younger than 4 years of age and 5% of children in the older age group had no demonstrable neurologic function on arrival. Prior studies have shown variability in the rate of neurologic injury. Mahan et al13 found neurologic symptoms to be present in only 12% of patients, whereas Osenbach and Menezes16 found 52% of patients to have some degree of neurologic injury despite similar injury patterns between the 2 studies. Carreon et al7 also demonstrated that children under 10 years of age were at an increased risk of neurologic injury compared with older children. It should be noted that the current definition of SCIWORA has become increasingly ambiguous with the common use of MRI for evaluation of ligamentous injuries in children with neurologic deficits in the setting of normal plain radiographs. To avoid confusion with the definition of SCIWORA, the newer literature has recommended reserving this term for patients with no abnormalities on plain radiographs or MRI.23 We chose to use this stricter definition for our study, thereby preventing the classification of isolated ligamentous injuries as SCIWORA and allowing us to more accurately classify injury patterns. Spine trauma in children represents high-energy trauma and is often associated with multiple, often severe, associated injuries. The most common associated injury in this population is head trauma. This is most often associated with cervical spine trauma with head injury rates www.pedorthopaedics.com |
701
J Pediatr Orthop
Knox et al
ranging from 30%24 to 66%.5,17,25-27 Although head injuries are also commonly present in patients with thoracolumbar trauma, such patients are more likely to sustain abdominal injuries.28,29 Patients in our study sustained multiple associated injuries with head injury representing the most common type in both age groups. These patients often sustained severe head trauma with GCS of 3. Significant differences in injury rates were seen in other injury locations. Infants/toddlers had significantly lower rates of facial trauma, associated pelvis and extremity fractures, as well as hollow viscus injuries. These children were also significantly less likely to present with a seatbelt sign. These differences likely relate to the differences in restraint use, as these patients were more likely to be properly restrained, thus potentially providing protection against these injuries. High mortality rate after spine injury has been demonstrated in prior studies with the highest mortality rates associated with cervical spinal cord injuries as well as upper cervical spine fractures.15,30 This increased mortality is likely related to the different injury patterns with a predilection for upper cervical spine injuries. AO injuries, in particular, have been associated with very high mortality rates with rates up to 100% reported.15 Our study confirmed these high rates of mortality, especially in the younger cohort of patients who were significantly more likely to expire during initial resuscitation or subsequent hospital admission. Children in the youngest age group had a 25% in-hospital mortality compared with 9% in our older cohort. Only 14% of IT and 18% of Y patients were treated with surgical intervention. Although some studies have shown similar rates of 18%,7,13 this need for operative intervention approached 30% in multiple series.5,16,30 The reason behind this decreased incidence of surgical intervention may be related, in part, to differences in practice patterns. In addition, our high rate of MRI use may likely have led to an increased rate of detection of more minor stable injuries for which surgery was not required. In addition, Ruge et al6 found very young children to need surgical intervention more often than children aged 4 to 12 years. Our series failed to identify any significant differences between the 2 age groups. In summary, although infants/toddlers demonstrated many similarities compared with young children, many important differences were identified. The infants/ toddlers were more likely to sustain injury to the upper cervical spine including the AO junction, were more likely to sustain purely ligamentous injuries versus compression or chance injuries, and were more often evaluated using MRI. These patients were also less likely to sustain concomitant injuries to the face, hollow viscus, or have associated fractures and demonstrated a much higher mortality rate compared with older patients. Infants/ toddlers were also found to have a high rate of spinal injury secondary to nonaccidental trauma in our series. REFERENCES 1. Kumaresan S, Yoganandan N, Pintar FA, et al. Biomechanical study of pediatric human cervical spine: a finite element approach. J Biomech Eng. 2000;122:60–71.
702 | www.pedorthopaedics.com
Volume 34, Number 7, October/November 2014
2. Lustrin ES, Karakas SP, Ortiz AO, et al. Pediatric cervical spine: normal anatomy, variants, and trauma. Radiographics. 2003;23:539–560. 3. Nuckley DJ, Eck MP, Carter JW, et al. Spinal maturation affects vertebral compressive mechanics and vBMD with sex dependence. Bone. 2004;35:720–728. 4. Ouyang J, Zhu Q, Zhao W, et al. Biomechanical assessment of the pediatric cervical spine under bending and tensile loading. Spine. 2005;30:E716–E723. 5. Eleraky MA, Theodore N, Adams M, et al. Pediatric cervical spine injuries: report of 102 cases and review of the literature. J Neurosurg. 2000;92:12–17. 6. Ruge JR, Sinson GP, McLone DG, et al. Pediatric spinal injury: the very young. J Neurosurg. 1988;68:25–30. 7. Carreon LY, Glassman SD, Campbell MJ. Pediatric spine fractures: a review of 137 hospital admissions. J Spinal Disord Tech. 2004;17:477–482. 8. Narci A, Solak O, Turhan-Haktanir N, et al. The prognostic importance of trauma scoring systems in pediatric patients. Pediatr Surg Int. 2009;25:25–30. 9. Tepas JJ IIIrd, Mollitt DL, Talbert JL, et al. The pediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg. 1987;22:14–18. 10. Bilston LE, Brown J. Pediatric spinal injury type and severity are age and mechanism dependent. Spine. 2007;32:2339–2347. 11. Cirak B, Ziegfeld S, Knight VM, et al. Spinal injuries in children. J Pediatr Surg. 2004;39:607–612. 12. Kokoska ER, Keller MS, Rallo MC, et al. Characteristics of pediatric cervical spine injuries. J Pediatr Surg. 2001;36:100–105. 13. Mahan ST, Mooney DP, Karlin LI, et al. Multiple level injuries in pediatric spinal trauma. J Trauma. 2009;67:537–542. 14. McGrory BJ, Klassen RA, Chao EY, et al. Acute fractures and dislocations of the cervical spine in children and adolescents. J Bone Joint Surg. 1993;75:988–995. 15. Nitecki S, Moir CR. Predictive factors of the outcome of traumatic cervical spine fracture in children. J Pediatr Surg. 1994;29:1409–1411. 16. Osenbach RK, Menezes AH. Pediatric spinal cord and vertebral column injury. Neurosurg. 1992;30:385–390. 17. Platzer P, Jaindl M, Thalhammer G, et al. Cervical spine injuries in pediatric patients. J Trauma. 2007;62:389–396; discussion 394-386. 18. Frank JB, Lim CK, Flynn JM, et al. The efficacy of magnetic resonance imaging in pediatric cervical spine clearance. Spine. 2002;27:1176–1179. 19. Sledge JB, Allred D, Hyman J. Use of magnetic resonance imaging in evaluating injuries to the pediatric thoracolumbar spine. J Pediatr Orthop. 2001;21:288–293. 20. Anderson RC, Kan P, Vanaman M, et al. Utility of a cervical spine clearance protocol after trauma in children between 0 and 3 years of age. J Neurosurg Pediatr. 2010;5:292–296. 21. Patel JC, Tepas JJ IIIrd, Mollitt DL, et al. Pediatric cervical spine injuries: defining the disease. J Pediatr Surg. 2001;36:373–376. 22. Firth GB, Kingwell S, Moroz P. Pediatric non-contiguous spinal injuries: the 15 year experience at a level 1 trauma center. Spine. 2012;37:E599–E608. 23. Yucesoy K, Yuksel KZ. SCIWORA in MRI era. Clin Neurol Neurosurg. 2008;110:429–433. 24. Orenstein JB, Klein BL, Gotschall CS, et al. Age and outcome in pediatric cervical spine injury: 11-year experience. Pediatr Emerg Care. 1994;10:132–137. 25. Birney TJ, Hanley EN Jr. Traumatic cervical spine injuries in childhood and adolescence. Spine. 1989;14:1277–1282. 26. Henrys P, Lyne ED, Lifton C, et al. Clinical review of cervical spine injuries in children. Clin Orthop Related Res. 1977;129:172–176. 27. Michael DB, Guyot DR, Darmody WR. Coincidence of head and cervical spine injury. J Neurotrauma. 1989;6:177–189. 28. Arkader A, Warner WC Jr, Tolo VT, et al. Pediatric Chance fractures: a multicenter perspective. J Pediatr Orthop. 2011;31:741–744. 29. Beaunoyer M, St-Vil D, Lallier M, et al. Abdominal injuries associated with thoraco-lumbar fractures after motor vehicle collision. J Pediatr Surg. 2001;36:760–762. 30. Puisto V, Kaariainen S, Impinen A, et al. Incidence of spinal and spinal cord injuries and their surgical treatment in children and adolescents: a population-based study. Spine. 2010;35:104–107.
r
2014 Lippincott Williams & Wilkins
