Spine Fractures in Children and Adolescents—Frequency, Causes, Diagnostics, Therapy and Outcome—A STROBE-Compliant Retrospective Study at a Level 1 Trauma Centre in Central Europe

The aim of this study was to present the frequencies and characteristics of paediatric spine fractures, focusing on injury mechanisms, diagnostics, management, and outcomes. This retrospective, epidemiological study evaluated all patients aged 0 to 18 years with spine fractures that were treated at a level 1 trauma centre between January 2002 and December 2019. The study population included 144 patients (mean age 14.5 ± 3.7 years; 40.3% female and 59.7% male), with a total of 269 fractures. Common injury mechanisms included fall from height injuries (45.8%), with an increasing prevalence of sport incidents (29.9%) and a decreasing prevalence of road incidents (20.8%). The most common localisation was the thoracic spine (43.1%), followed by the lumbar spine (38.2%), and the cervical spine (11.8%). Initially, 5.6% of patients had neurological deficits, which remained postoperatively in 4.2% of patients. Most (75.0%) of the patients were treated conservatively, although 25.0% were treated surgically. A small proportion, 3.5%, of patients presented postoperative complications. The present study emphasises the rarity of spinal fractures in children and adolescents and shows that cervical spine fractures are more frequent in older children, occurring with a higher rate in sport incidents. Over the last few years, a decrease in road incidents and an increase in sport incidents in paediatric spine fractures has been observed.


Introduction
Paediatric spine fractures are relatively rare, with an incidence ranging from 1% to 4% [1]. Currently, cervical spine fractures constitute about 1% of all spine fractures in children and adolescents, whereas thoracic and lumbar fractures make up 2-3% [2][3][4][5][6]. The localisation of spine fractures varies with age: thoracic and lumbar spine fractures occur more often in older children (>10 years), whereas cervical spine fractures occur more frequently in younger children [6][7][8]. The upper cervical spine (C0-C2) is especially at risk of fracturing in children younger than 8-10 years due to the different fulcrum and relatively large head compared to adults [9]. In addition, the paediatric spine is generally more unstable due to ligamentous laxity, weak paravertebral muscles, and the horizontal orientation of the facet joints [9]. At ages between 8 and 10 years, the anatomy and biomechanics of the paediatric spine are comparable with the adult spine [10].
Common causes for paediatric spine fractures are falls, sport and road incidents, as well as child abuse [2]; most epidemiological studies have identified road incidents are the most frequent cause of spine fractures [11,12]. Currently, epidemiological studies assessing the total spine (cervical, thoracic, and lumbar spine) are relatively rare; most studies either focus only on a special region of the spine, or present heterogenous injuries by also including ligamentous injuries. This makes current studies difficult to compare.
Due to the rarity of up-to-date epidemiological studies and their heterogeneity at present, the aim of this study was to describe the frequency and characteristics of paediatric spine fractures of the entire vertebral column, focusing on injury mechanisms, diagnostic procedures, management, and outcomes at a level 1 trauma centre.

Methods
This study was performed as a retrospective, epidemiological data analysis at a level I trauma centre and was approved by the Ethics Committee of the Medical University of Vienna (Code 1816/2020).
This study was performed following the STROBE guidelines in Appendix A. Initially, 211 children and adolescents aged from 0 to 18 years with spine fractures were treated at the Department of Trauma and Orthopaedic Surgery at the Medical University of Vienna, during an observation period from January 2002 to December 2019. Finally, 144 children were included after applying the inclusion and exclusion criteria ( Figure 1).

Figure 1.
Flow chart of the overall study population (A0 = minor non-structural fractures; MRI, morphologically detectable "bone bruise" as well as spinous and transverse process fractures).
All patients aged from 0 to 18 years with a fracture of the cervical, thoracic, and/or lumbar vertebra during the observation period were included. Exclusion criteria were age over 18 years, spinal injuries such as contusion or distortion, exclusive ligamentous injuries, sacral fractures, and fractures of the spinous or transverse processes, and the vertebral arch. Furthermore, healed or questionable fractures were excluded, as were patients who were initially treated at an external hospital. If radiological documentation was incomplete, patients were also excluded from the study.
The data were collected retrospectively from the patient's charts, and included age, sex, injury mechanism, fracture localisation (cervical, thoracic, lumbar, or multiple regions), diagnostics using plain radiography (X-ray), computer tomography (CT scans), magnetic resonance imaging (MRI), management (operative or non-operative), as well as the exact surgical procedure or conservative treatment. Fractures were classified using the Gehweiler classification for C1 fractures, the Anderson and D'Alonzo classification for odontoid fractures, the Effendi classification for C2 fractures, and the AO Spine classification for lower cervical spine, thoracic, and lumbar spine fractures [13][14][15][16]. The clinical outcome, complications, and mobility (walking, crutches, bedridden, etc.) were extracted after treatment. Furthermore, the Frankel Score [17] was used to describe neurological deficits at the time of presentation and at the last follow-up.

Data Analysis
Descriptive data are reported for the entire patient cohort, including the mean, range, and standard deviation (SD). In order to develop an epidemiological overview, the following parameters were evaluated: age, sex, fracture classification and localisation, injury mechanism, diagnostic imaging methods applied, management (operative or nonoperative), surgical procedure, conservative treatment, neurological examination, complications, and mobility. Nominal and ordinal variables are presented as absolute and relative frequencies. Metric variables are reported as the mean, range, and standard deviation. The confidence interval for relative frequencies was 95%. Statistical analysis was performed using Microsoft Excel (Version 16.50., Microsoft Corp., Redmond, WA, USA) and SPSS software (Version 27.0.0., SPSS Inc., Chicago, IL, USA).
In 44 patients, fractures were located in the thoracolumbar (Th12 and/or L1); L1 was the most frequently observed fractured vertebra.
Surgical treatment was mostly indicated because of compression of the spinal canal (25/144, 17.4%) and was necessary in 25.0% (36/144) of patients. Anterior stabilisation was used in all patients with cervical spine fractures (6/144, 4.2%), with posterior stabilisation more frequently applied in the thoracic (7/144, 4.9%) and lumbar spine (10/144, 6.9%) ( Figure 5).  1 and 2 show the radiographs of an A1.2 L1 fracture in a 12-year-old girl sustained after falling off a trampoline. The patient was treated conservatively by receiving adequate analgesia and sports abstinence. The radiographs at the last follow-up after three months of therapy show no further dynamics and a healed fracture (Images 3 and  4).

Neurological Deficits and Outcome
In total, 5.6% (8/144; mean age 15.6 ± 4.7 years) of patients presented with neurological deficits after trauma and appeared most frequently in adolescents (7/8) and those who sustained sport incidents (3/8). Neurological involvement improved in two patients after treatment: one patient improved from Frankel A to Frankel C categorisation after posterior stabilisation, and one patient showed improvements after bracing from Frankel D to E. The other six patients (4.2%; mean age 15 ± 5.4 years) showed consistent neurological deficits after treatment.
Postoperative complications were classified according to the Clavien-Dindo Classification and occurred in 3.5% of patients (5/144), including wound infections, insufficient implants, breakage of screws, and screws invading the spinal canal. One complication (cage loosening, treated conservatively) was classified as group 1, and the other four complications requiring revisions under general anaesthesia were accordingly classified as group 3b. Patients who underwent a conservative treatment had no complications.

Discussion
The current study shows that paediatric spine fractures are relatively rare with a peak in middle aged children with a mortality rate of 1.4%. Similar data can be seen in many other retrospective studies [11,[18][19][20]. In the current study, injury patterns changed from road incidents to sport incidents during the observation period. Accordingly, a higher frequency of cervical spine fractures was noted in adolescents. These findings are comparable to Poorman et al. and Shin et al., but are contrary to the findings of Compagnon et al. and Mahan et al., who reported a tendency of cervical spine injuries in younger children [18,[21][22][23]. This discrepancy may be explained by differences in the study population and inclusion criteria. Poorman et al. only referred to cervical spine fractures in their study reporting a higher prevalence of cervical spine fractures in adolescents (ages 11-18 years) and young adults (ages 19-20 years) [21]. The inclusion criteria of Compagnon et al. and Mahan et al. contained not only vertebral fractures, but also ligamentous injuries of the paediatric cervical spine [18,23]. In addition, no differentiation was made between ligamentous injuries and solely bony fractures when it was stated that spinal injuries occurred, especially in young children (ages 0-8 years). It was not possible to determine the percentage of actual cervical spine fractures; hence, the findings cannot be compared directly to the current study. The reason for the higher frequency of younger children in these studies might be attributable to the inclusion of ligamentous injuries, because this is the main difference between the present study and that of Poorman et al. [21]. Therefore, cervical spine fractures were more frequently observed in high schoolers and adolescents. Furthermore, the higher rate of cervical spine fractures in high schoolers and adolescents in the present study may be related to the high frequency of sport incidents in this patient cohort. The results in the current study present sport incidents as the most frequent injury mechanism leading to cervical spine fractures (40% of all cervical patients; mean age 14.6 ± 3.8 years). High-schoolers (6) and adolescents (5), out of a total of 12 patients (91.7%), accounted for the majority of these patients, which is comparable to reports in the literature; the incidence of cervical spine injuries in sports increases with age [24]. Furthermore, the current data reveal that road incidents have decreased over the years: from 2002 to 2007, 16/44 (36.4%) patients suffered spinal fractures in a road incident, whereas from 2014 to 2019, the number decreased to 7/56 (12.5%) patients, which is in contrast to the extant literature [12,20,25,26]. However, the observation period of these studies is more comparable to the period from 2002 to 2007 than to the latest data (2014 to 2019). The decrease in road incidents may be related to the increased safety features of cars in the last few years, considering similar findings in a recent study from Compagnon et al. published in 2020 [18]. The posting of more speed limits may have influenced this trend as well. Although road incidents have decreased, we reported an increase in sport incidents, from 10 to 20 patients over the years, constituting 30% of all injury mechanisms. Similar findings have been presented in other retrospective studies [11,18,27,28]. Sport incidents might have become more frequent due to increases in at-risk sports such as horse riding [18] or skiing, the latter being the most frequently implicated sport in our study. A limitation of the present study is the retrospective design of our investigation and the relatively low number of patients, partly due to the fact that this was a single-centre study. However, the number of patients was also a result of the strict inclusion criteria necessary in order to generate a homogenous study population. We only included spine fractures classified from A1 to C (AO Spine Classification) and further excluded isolated injuries of the ligaments, because these are defined as minor injuries if still, stable conditions are present [29]. Furthermore, the injury pattern of SCIWORA (spinal cord injury without radiographic abnormality) was not included, because this study only focused on bony injuries. The lack of differentiation between ligamentous and osseous injuries in many studies resulted in a heterogeneity of injuries. The main advantage of the present study is the overview of paediatric spine fractures of a level one trauma centre and the strict inclusion criteria only including bony injuries to the spine. Additionally, the long observation period of almost 2 decades even enabled us to present changes in the frequencies of certain injury mechanisms.

Conclusions
The present study emphasises the rarity of spine fractures in children and adolescents and shows that cervical spine fractures are more frequent in older children, occurring with a higher rate in sport incidents. Over the last few years, a decrease in road incidents and an increase in sport incidents in paediatric spine fractures has been observed. Informed Consent Statement: All included patients in this study gave informed consent.

Data Availability Statement:
The datasets generated and/or analysed during the current study are not publicly available due to data privacy, but are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare that they have no competing interest.

Introduction
Background/rationale (applied) 2 Explain the scientific background and rationale for the investigation being reported Objectives (applied) 3 State specific objectives, including any prespecified hypotheses

Study design (applied) 4
Present key elements of study design early in the paper Setting (applied) 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection Participants (not applicable) 6 (a) Cohort study-Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Case-control study-Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls Cross-sectional study-Give the eligibility criteria, and the sources and methods of selection of participants

Other Information
Funding (not applicable) 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based Note: An Explanation and Elaboration article discusses each checklist item and gives the methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org. * Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies.