Pediatric scapular fractures and associated injuries following blunt chest trauma

a b s t r a c t

Background: Scapular fractures in the pediatric population are rare, and medical literature is lacking regarding these specific injuries in the pediatric population. Prior studies have shown that scapular fractures resulting from blunt chest trauma have been associated with significant morbidities in adults, and that a majority of scap- ular fractures are missed on Chest x-ray and seen on computerized tomography only (SOCTO). Further guidance is needed regarding the prevalence of coinciding injuries in the pediatric population and the modality for diagnosis.

Objectives: The primary objectives of this study were to assess 1) the frequency of scapular fractures following blunt trauma in the pediatric cohort, 2) the frequency of other associated thoracic injuries, 3) the proportion on scapular fractures SOCTO.

Methods: We conducted a retrospective cohort study with data obtained from our study site’s Trauma Registry. Patients under 18 years receiving both a CXR and chest CT following blunt trauma or any patient diagnosed with a scapular fracture by any modality from January 2009 to December 2019 were included. Primary outcome variables were the presence of a scapular fracture diagnosed by any modality, absence of scapular fracture, and scapular fractures SOCTO. Charts were also reviewed for the following concurring injuries: 1) contusion/atelec- tasis, 2) pneumothorax, 3) hemothorax, 4) rib fracture, 5) other fracture, 6) Vascular injury, 7) mediastinal injury,

8) diaphragm rupture, 9) foreign body, 10) incidental finding.

Results: Of 12,826 charts of pediatric patients with blunt chest trauma, 1405 obtained both CXR and chest CT. Sixty (0.47%) were diagnosed with scapular fracture, and 48 (73.3%) of the fractures were SOCTO. The most com- monly Associated injuries were other fracture (88.3%), lung contusion/atelectasis (78.3%), pneumothorax (58.3%) and rib fracture (58.3%). Patients with scapular fractures had higher injury severity scores (ISS) and more fre- quently required surgery for other intrathoracic injuries. Only five patients required surgical management of the scapular fracture with the rest managed conservatively.

Conclusion: Pediatric scapular fractures are rare and are often associated with other intrathoracic injury. A major- ity of scapula fractures are missed on CXR, but identification of the injury did not change management as most were treated conservatively.

Published by Elsevier Inc.

  1. Background

Scapular fractures in the pediatric population are rare, accounting for approximately 1% of fractures sustained in children [1]. As a result, medical literature is lacking regarding these specific injuries in the pedi- atric population. A literature review starting from 1839 through 2021 only describes 70 scapular fractures in children and adolescents, with a majority being single case reports [2].

Historically, scapular fractures caused by blunt chest trauma have been associated with significant morbidities in adults. Prior

* Corresponding author.

E-mail address: [email protected] (F.S. Fonacier).

to the widespread availability of Computerized tomography , patients found to have scapular fractures also had an increased like- lihood of pulmonary contusion, rib fracture, and clavicular fracture [3]. With the increased availability of chest CT, more scapular frac- tures can be identified. In adults, the majority of scapular fractures are diagnosed by CT only with the most common coinciding injuries being rib fracture, pulmonary contusion, pneumothorax and thoracic spine fracture [4].

The primary objective of this study is to assess the incidence of scap- ular fractures and associated thoracic injuries following blunt trauma in a pediatric population. We hypothesize that similar to the adult popula- tion, scapular fractures are rare in children and that most of these will be seen on CT only (SOCTO).

https://doi.org/10.1016/j.ajem.2021.12.014 0735-6757/Published by Elsevier Inc.

  1. Methods
    1. Study design, setting

The Committee for the Protection of Human Subjects and the Institu- tional Review Board approved this retrospective cohort study. Most of the data was extracted from this study site’s Trauma Registry. Patients are included in the trauma registry if they have an ICD-9 or ICD-10 cod- able injury and meet one of the following criteria: 1) admitted to the hospital, 2) died in the emergency department (ED), 3) transferred out of the ED for a higher level of care, 4) transferred to the ED from an- other acute care ED or hospital. Patients are excluded from the trauma registry if their codable injury is considered isolated and superficial such as abrasion or soft tissue contusion. All patients are assigned a reg- istry number and medical records are abstracted for 65 data points in- cluding dates and times, vitals, prehospital information, procedures, comorbidities and outcomes. Validation is done with audits of up to 10% of each trauma registrar by a trauma registry manager, physician investigator, nurses, and trauma program managers.

Our study site is a tertiary care pediatric hospital within one of the largest Hospital systems in a large metropolitan city. The study site is the only American College of Surgeons-verified level I pediatric and adult trauma center in the city.

We conformed to the methods of proper medical review studies [5]. All data abstractors used a standardized instrument to collect injury data and clinical and surgical procedures from the electronic medical re- cord. Abstractors received training on extraction of data. Inclusion and exclusion criteria were defined, and categorization of injury and types of procedures were determined in advance. We monitored research ab- stractors for accuracy by double checking data entry for all patients who met inclusion criteria for our study.

    1. Study population

Patients under 18 years receiving both a CXR and chest CT following blunt trauma or any patient diagnosed with a scapular fracture by any modality from January 2009 to December 2019 were included. Exclu- sion criteria were patients with a non-acute scapular fracture as deter- mined by a radiologist, trauma occurring >24 h prior to presentation to ED, patients who did not receive both CXR and chest CT, or penetrat- ing chest traumas.

    1. Measurements

Primary outcome variables were the presence of a scapular fracture diagnosed by any modality, absence of scapular fracture, and scapular fractures SOCTO. Secondary outcome variables were the specific con- curring intrathoracic injuries: 1) pulmonary contusion/atelectasis,

2) pneumothorax, 3) hemothorax, 4) rib fracture, 5) other fracture,

6) vascular injury, 7) mediastinal injury, 8) diaphragm rupture, 9) for- eign body, 10) incidental finding. The presence of the above injuries was determined based on final read by board-certified pediatric radiol- ogist. If an official report was considered equivocal on CXR, it was con- sidered positive. Management of the injuries was determined by physician notes in the medical record and trauma registry.

Independent variables included basic demographics (age, sex, race), mechanism of injury, injury severity score (ISS) [6], and ED disposition. ED disposition included admission to the hospital, admission to the in- tensive care unit, disposition to the operating room for surgery, dis- charge to home and death in the ED.

    1. Data analysis

Patients were stratified into three groups: 1) those with scapular fracture, 2) those without scapular fracture, and 3) those with scapular fracture SOCTO. We assessed differences between these three groups in

terms of age, sex, race, mechanism of injury, associated injuries, ISS and ED disposition. Descriptive statistical analysis was performed on the co- variates including age, sex, race, mechanism of injury, ISS and ED dispo- sition, stratified by the three above mentioned groups. Then, adjusted and unadjusted odds ratio of associated injuries were estimated with simple logistic regression models to estimate the odds of having another specific injury.

  1. Results

A total of 12,826 charts of pediatric patients with blunt injury were identified from the trauma registry. Seventy-eight charts were excluded for incomplete imaging (either missing CXR or chest CT read). Of the 78 excluded charts, however, eight patients were diagnosed with scapular fractures. Five of the eight were identified on CT only with no CXR done, three were identified on CXR only as no CT was done. Of the 1405 pa- tients who had both chest CT and CXR, 60 (4.3%) were diagnosed with a scapular fracture. Of these 60 patients, 48 (73.3%) were diagnosed with scapular fracture SOCTO (Fig. 1). Of the SOCTO group, scapular fracture was found to be the isolated injury in only two (4.5%) patients. Radiology reports described 26 of the scapular fractures as displaced, 11 as non-displaced, and 23 made no mention of displacement. Reports described 25 of the scapular fractures as comminuted, with 35 not fur- ther specified. A majority of the scapular fractures occurred in the scap- ular body (37), followed by the coracoid process [12], scapular spine [8] and acromion process [4]. Some fractures contained more than one an- atomical area. Two of the patients suffered bilateral scapular fractures. Of the 60 total patients with scapular fractures, five required surgical management. Three of these scapular fractures were SOCTO and two were seen on both CXR and CT. Three of the five patients had open scap- ula fractures requiring irrigation and debridement in the operating room. Two patients required open reduction and internal fixation. Among the SOCTO fractures, two were for Open fractures and the third patient had a GCS of 3 on ED presentation. While the third patient was initially treated conservatively, open reduction and internal fixation oc- curred two weeks after initial presentation to the ED while still hospital- ized due to Functional limitation of the shoulder. All other patients had scapular fractures managed conservatively with non-weight bearing in- structions and sling. Of the eight patients excluded from our analysis for incomplete imaging, seven were also managed conservatively and one


Demographic differences between patients with scapular fractures seen on both CXR and CT, patients without scapula fractures, and those with scapular fracture SOCTO are compared in Table 1. Notably, patients with scapular fractures were more likely to require surgical management of other injuries than those without. Patients without scapular fractures appear to have a higher admission rate to the hospi- tal, however, this is likely due to a greater percentage of patients with scapular fractures receiving surgery for other coinciding injuries. The median ISS was greater for patients with scapular fracture (23.5) com- pared to those without (16), although median ISS for scapular fracture seen on CXR and CT (23.5) was largely the same for those SOCTO (21.5). The unadjusted and adjusted odds ratio for associated injuries in pa- tients with scapular fractures versus patients without scapular fractures are shown in Table 2. Overall, patients with scapular fractures had a higher percentage of all listed concurrent injuries with the exception of foreign body and incidental findings. The percentages of concurrent injuries between the scapular fracture cohort and SOCTO cohort were

largely the same.

  1. Discussion

Scapular fractures are exceedingly rare and their clinical significance in blunt chest trauma in the pediatric population is not well understood. Our study is the largest sample of pediatric patients with scapular frac- tures to date. The only known prior study of pediatric scapular fractures

Image of Fig. 1

Fig. 1. Flow chart of Medical records reviewed.

Table 1

Characteristics of scapular fracture versus non-scapular fracture patients.

Scapula fracture

Non-scapula fracture


N = 45

N = 60

N = 1345

Age, median (IQR)

14.0 (12.0-17.0)

12.0 (6.0-16.0)

16.0 (12.0-17.0)

Sex, n (%)


18 (30.0)

534 (39.7)

13 (28.9)


42 (70.0)

811 (60.3)

32 (71.1)

Race, n (%) White

20 (33.3)

545 (40.5)

15 (33.3)


10 (16.7)

252 (18.7)

7 (15.6)

reported a frequency of approximately 2% among patients less than 18 years of age who presented after MVC [1]. Our findings showed scap- ular fractures were seen in 4.3% of our sample screened with both CT and CXR but only 0.5% of all blunt Pediatric trauma patients during the study period. Patients with scapular fracture in our cohort had a higher likelihood of also having other fracture (88.3%), lung contusion/atelec- tasis (78.3%), pneumothorax (58.3%) and rib fracture (58.3%). A study in the adult population by Brown et al. found concurrent injuries to be similar, with the most common being rib fracture, pneumothorax, and pulmonary contusion [4]. Shannon et al.’s pediatric study noted the same three concurrent thoracic injuries as Brown’s adult study, and found other injuries in general were 2.5x more likely to occur with


1 (1.7)

25 (1.9)

0 (0.0)

a scapular fracture after blunt trauma [1,4]. In addition, Brown also


23 (38.3)

474 (35.2)

18 (40.0)

compared the percentage of concurrent injury between their scapular


6 (10.0)

48 (3.6)

5 (11.1)

fracture cohort and scapular fracture SOCTO cohort, demonstrating

Injury mechanism, n (%)

Assault/non-accidental trauma

Pedestrian struck

15 (25.0)

269 (20.0)

11 (24.4)


0 (0.0)

4 (0.3)

0 (0.0)

Crush injury

0 (0.0)

24 (1.8)

0 (0.0)


2 (3.3)

90 (6.7)

1 (2.2)


43 (71.7)

895 (66.5)

33 (73.3)

Other accidents

0 (0.0)

5 (0.4)

0 (0.0)

Sporting injury

0 (0.0)

25 (1.9)

0 (0.0)

ED Disposition, n (%) Admitted to ICU

25 (41.7)

532 (39.6)

19 (42.2)

Admitted to Hospital

16 (26.7)

588 (43.7)

13 (28.9)


0 (0.0)

3 (0.2)

0 (0.0)


0 (0.0)

7 (0.5)

0 (0.0)


19 (31.7)

215 (16.0)

13 (28.9)

Mortality, n (%)

0 (0.0)

3 (0.2)

0 (0.0)

Injury Severity Score (ISS), median (IQR)

23.5 (15.5-34.0)

16.0 (9.0-25.0)

22.0 (14.0-33.0)

0 (0.0) 33 (2.5) 0 (0.0)

that the presence of concurrent intrathoracic injury was very similar between the two groups. Our study demonstrated similar findings in the pediatric population.

Shannon’s pediatric study demonstrated a higher ISS and increased mortality for patients with scapular fracture [1]. While our study found an increased ISS in patients with scapular fractures (ISS of 23.5 in our scapular fracture cohort, 22.0 in our SOCTO cohort versus 16.0 in our non-scapular fracture cohort), there were no fatalities in any patients with scapular fractures and only three fatalities in our non- scapular fracture cohort due to other injuries. A higher percentage of patients with scapular fracture (31.7% in scapular fracture cohort, 28.9% in SOCTO cohort) required surgery for other intra-thoracic inju- ries during hospitalization compared to those without scapular fracture (16.0%). The most common surgical intervention for other intrathoracic injury associated with scapular fracture was thoracostomy tube place- ment, consistent with the high prevalence of concurrent pneumothorax seen in our data. Brown et al. found that while the scapula group had higher Hospital admission rates, ISS and hospital length of stay, when

Table 2

Associated injuries in scapular fracture versus non-scapular fracture patients.

Associated injury, n (%)

Scapula fracture

Non-scapula fracture


Unadjusted odds ratio a

Adjusted odds ratio b

N = 60

N = 1345

N = 45

(95% confidence interval)

(95% confidence interval)


47 (78.3)

841 (62.5)

34 (75.6)

2.2 (1.2-4.0)

1.7 (0.8-3.3)


35 (58.3)

320 (23.8)

23 (51.1)

4.5 (2.6-7.6)

3.1 (1.8-5.5)


9 (15.0)

51 (3.8)

6 (13.3)

4.5 (2.1-9.6)

2.4 (1.0-5.6)

Rib Fracture

35 (58.3)

233 (17.3)

26 (57.8)

6.7 (3.9-11.4)

5.0 (2.9-8.9)

Other Fracture

53 (88.3)

256 (19.0)

38 (84.4)

32.2 (14.5-71.7)

28.5 (12.6-64.6)

Vascular Injury

5 (8.3)

11 (0.8)

4 (8.9)

11.0 (3.7-32.8)

4.9 (1.5-16.3)

Mediastinal Injury

18 (30.0)

185 (13.8)

14 (31.1)

2.7 (1.5-4.8)

2.0 (1.1-3.7)

Diaphragm Rupture

1 (1.7)

1 (0.1)

1 (2.2)

22.8 (1.4-368.7)

24.9 (0.9-666.5)

Foreign Body

1 (1.7)

43 (3.2)

0 (0.0)

0.5 (0.1-3.8)

0.4 (0.1-3.1)

Incidental Finding

3 (5.0)

105 (7.8)

3 (6.7)

0.6 (0.2-2.0)

0.6 (0.2-2.0)

a The unadjusted odds ratio is comparing scapular fractures to non-scapular fracture subjects.

b The adjusted odds ratio is comparing scapular fractures to non-scapular fracture subjects, adjusted with age, sex, injury mechanism, ED disposition, and ISS.

compared to the non-scapula group, their overall mortality was not sig- nificantly different in their adult population [4].

The basic Imaging study to diagnose scapular fractures is an AP and

Y-axis radiograph. A majority of fractures, especially those of the glenoid, scapular neck and coracoid, require CT imaging with 3D recon- struction [2]. Brown et al. confirmed the importance of chest CT in the diagnosis of scapular fractures in the adult population, with a finding of 60.3% to be SOCTO. Our study found 73.3% of scapular fractures were SOCTO. As demonstrated in the literature, chest CT is more sensi- tive in diagnosing a number of thoracic injuries in children including rib fractures, pulmonary contusions, pneumothoraces, hemothoraces and thoracic spine fracture but rarely result in surgical changes in manage- ment [7-9]. Most of the scapular fractures in our present cohort were treated conservatively and only five required surgical management, two diagnosed on both CXR and CT and three were SOCTO. The three pa- tients from the SOCTO group that went on to surgery had clinical find- ings that would have prompted further imaging and surgery.

We published previously on the increased sensitivity of chest CT in comparison to chest x-ray to diagnose pulmonary contusions, rib fractures, hemothoraces and pneumothoraces. These injuries rarely changed surgical management for patients. In many instances, when other intrathoracic injuries found on chest CT and coincidentally missed on CXR resulted in surgical Changes in management, there was a clinical suspicion on exam or the patient had an exam precluding an assessment such as a GCS of 3 [7].

The decision to obtain chest CT must consider charges and costs to patient, insurers, and society while balancing trauma outcomes. Medical imaging is one of the costliest diagnostic techniques and the most widely used. A new higher slice CT scan machine can cost $2.5 million before adding the recurring maintenance and equipment fees. Radiation increases the carcinogenic risk in children compared to adults so careful consideration regarding the use of CT in the pediatric population is par- amount [10-12]. Additionally, the overutilization of CT has dramatically contributed to national Healthcare costs [13,14]. In addition, obtaining these studies adds to the length of stay in the department. We would recommend the use of chest CT in the setting of a suspected scapular fracture when surgical management is warranted clinically.

    1. Limitations

This was a retrospective study conducted at a single center. While our study initially enrolled nearly 13,000 pediatric patients, approxi- mately only one-tenth of these patients had a chest CT obtained, limit- ing our sample significantly. This is likely due to physicians’ judicious practice of limiting radiation in the pediatric population as there is no definitive protocol when determining need for chest CT following blunt trauma. Since this study only included patients who received both chest X-ray and CT, we narrowed the population to those with higher suspicion of thoracic injury. The even smaller number of patients

with scapular fractures may have limited our ability to detect significant differences between groups. Outcomes such as ISS, disposition, surgical management, and mortality are not based on the injury of scapular frac- ture alone. While a good marker for severity of associated injuries, other specific injuries could be confounding variables not considered.

  1. Conclusion

Overall, CT was superior in diagnosing scapular fractures as CXR alone missed these fractures 73.3% of the time. Patients with scapular fractures were also more likely to sustain additional thoracic injury par- ticularly pulmonary contusion/atelectasis, pneumothorax, hemothorax and rib fracture. This cohort was noted to have had higher ISS, and more frequently required surgical management for the other injuries.

This research did not receive any specific grant from funding agen- cies in the public, commercial or not-for-profit sectors. This article has not been presented elsewhere and is not under for consideration for publication elsewhere.


  1. Shannon SF, Hernandez NM, Sems SA, Larson AN, Milbrandt TA. High-energy pediat- ric scapula fractures and their associated injuries. J Pediatr Orthop. 2019;39(7): 377-81. https://doi.org/10.1097/bpo.0000000000000969.
  2. Bartonicek Jan, Nanka Ondrej. History of diagnostics and treatment of scapular frac- tures in children and adolescents and its clinical importance. Arch Orthop Trauma Surg. 2021. https://doi.org/10.1007/s00402-021-03800-8.
  3. Thompson David A, et al. The significance of scapular fractures. J Trauma. 1985; 25.10:974-7 Web.
  4. Brown C, Elmobdy K, Raja AS, Rodriguez RM. Scapular fractures in the pan-scan era. Acad Emerg Med. 2018;25(7):738-43. https://doi.org/10.1111/acem.13377.
  5. Worster A, Bledsoe RD, Cleve P, et al. Reassessing the methods of medical record re- view studies in emergency medicine research. Ann Emerg Med. 2005;45(4):448-51.
  6. Brunello LFS, Abreu P, Schulz GJ, Saavedra Tomasich F, Meizoso J, Ginzburg E. Injury severity scoring. In: Nasr A, Saavedra Tomasich F, Collaco I, Abreu P, Namias N, Marttos A, editors. The trauma golden hour. Cham: Springer; 2020. https://doi.org/ 10.1007/978-3-030-26443-7_4.
  7. Ugalde Irma T, et al. Chest X-ray vs. computed tomography of the chest in pediatric blunt trauma. J Pediatr Surg. 2021;56.5:1039-46 (Web).
  8. Golden J, Isani M, Bowling J, et al. Limiting chest computed tomography in the eval- uation of pediatric Thoracic trauma. J Trauma Acute Care Surg. 2016;81(2):271-7.
  9. Hershkovitz Y, Zoarets I, Stepansky A, et al. Computed tomography is not justified in every pediatric blunt trauma patient with a Suspicious mechanism of injury. Am J Emerg Med. 2014;32(7):697-9.
  10. Kutanzi KR, Lumen A, Koturbash I, Miousse IR. Pediatric exposures to ionizing radi- ation: carcinogenic considerations. Int J Environ Res Public Health. 2016;13(11): 1057 Published 2016 Oct 28. https://doi.org/10.3390/ijerph13111057.
  11. Brenner David J, Hall Eric J. Computed tomography — an increasing source of radia- tion exposure. N Engl J Med. 2007;357.22:2277-84 (Web).
  12. Pearce Mark S, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. The Lancet (Br Edn). 2012;380.9840:499-505 (Web).
  13. Armao Diane, Semelka Richard C, Elias Jr Jorge. Radiology’s ethical responsibility for healthcare reform: tempering the overutilization of medical imaging and trimming down a heavyweight. J Magn Reson Imaging. 2012;35.3:512-7 (Web).
  14. Hendee William R, et al. Addressing overutilization in medical imaging. Radiology. 2010;257.1:240-5 (Web).