Do not waste your time: straight to magnetic resonance imaging for pediatric burners and stingers

a b s t r a c t

Background: Permanent neurologic injury in pediatric patients with burner and stinger syndrome (BSS) is unlike- ly. This study aims to assess the feasibility of clinical observation without extensive radiologic workup in this se- lective population.

Methods: A retrospective study was conducted of patients aged younger than 18 years evaluated at a level I trau- ma center from 2012 to 2014. Patients were grouped according to positive deficit (PD) or negative deficit (ND) upon physical examination. Demographics, clinical findings, and outcomes were analyzed.

Results: Thirty patients (ND, n = 14; PD, n = 16) were evaluated for BSS, most often as a result of injurious foot- ball tackle. Age and length of stay were similar between groups. Injury Severity Score was lower in the ND group than the PD group (1.6 +- 1.2 vs 3.8 +- 3.1, respectively; Pb .05). Cervical computed tomography was performed on 11 patients (78.6%) in the ND group and 15 patients (93.8%) in the PD group at considerable added cost, with only 1 positive result in the ND group and none in the PD group. Magnetic resonance imaging (MRI) revealed 2 positive findings in each group, and no surgical interventions were indicated. Ten ND (71.4%) and 12 PD (75%) patients reported complete resolution of symptoms at discharge (PN .05).

Conclusions: Children presenting with BSS experience temporary symptoms that resolve without surgical inter- vention. Magnetic resonance imaging identified more injuries than computed tomographic imaging; therefore, we suggest that management for BSS should include observation, serial Neurologic examinations, and MRI eval- uation as appropriate.

Introduction

Burner and stinger syndrome (BSS), also known as transient brachial plexopathy, is a constellation of symptoms of pain, weakness, numb- ness, and tingling that affects a unilateral upper extremity after a high-energy collision. Burner and stinger syndrome most commonly oc- curs during contact sports such as football, wrestling, basketball, or rugby. As many as 65% of college football players have experienced an episode of BSS [1,2]. The symptoms of BSS arise from trauma to the

? Sources of support: None to declare.

?? The results described were presented at the 2015 Annual Meeting of the American As- sociation for the Surgery of Trauma in Las Vegas, Nevada.

* Corresponding author at: Carolinas Medical Center, 1000 Blythe Boulevard, MEB Suite 601, Charlotte, NC 28203. Tel.: +704 355 3176; fax: +704 355 5619.

E-mail addresses: [email protected] (R. Sola), [email protected] (A.B. Christmas), [email protected] (B.W. Thomas), [email protected] (P.E. Fischer), [email protected] (G.C. Eubanks), [email protected] (N.E. Raynor), [email protected] (R.F. Sing).

brachial plexus or nerve roots [3]. Although most symptoms improve quickly with no long-term sequelae, 5% to 10% of cases can have persis- tent symptoms, which could indicate more serious injury requiring medical or surgical management [2].

Currently, no consensus has been reached in the trauma literature regarding the clinical management and radiographic evaluation of pedi- atric patients with sports-related BSS. Computed tomography (CT) is currently one of the most widely used radiologic modalities to evaluate for pediatric blunt trauma injuries; however, only 4.3% of cervical spine CT scans have been found to be positive in pediatric blunt trauma pa- tients [4]. Further evidence suggests that the radiation harm for routine cervical CT scans for low-risk cases equals or exceeds its benefit [5]. As one diagnostic test for BSS, magnetic resonance imaging (MRI) is better to evaluate soft tissue abnormalities (ligamentous injury, soft tissue edema, and herniated disks) and uses less radiation than CT [6]. Histor- ically, CT has been the imaging of choice for Osseous injuries, although recent evidence suggests that MRI has a comparable sensitivity and specificity for evaluating osseous injuries [7].

The objective of this study was to review the demographics, symptoms, diagnostic imaging, and clinical outcomes of patients presenting with BSS. We hypothesized that a conservative approach

http://dx.doi.org/10.1016/j.ajem.2016.04.020

R. Sola Jr et al. / American Journal of Emergency Medicine 34 (2016) 1442–1445 1443

of clinical observation without extensive radiologic workup in this patient population will not increase the low likelihood of permanent neurologic damage.

Methods

The protocol for this study was approved by the Institutional Review Board of Carolinas HealthCare System. We reviewed patients who were evaluated at the FH “Sammy” Ross Jr Trauma Center of Carolinas Medi- cal Center over a 24-month period beginning in January 2012. We retro- spectively queried, from our institutional trauma registry, data from patients who were evaluated for a concern of BSS. Patients younger than the age of 18 years who were involved in a contact sports injury were included. Patients involved in head-to-head collisions with no motor or sensory deficits and those patients with Delayed presentation after injury were excluded from the study. Patients were grouped based on physical examination findings found by the trauma surgery team. The positive deficit (PD) group was found to have motor and/or sensory deficits that persisted on physical examination. The negative deficit (ND) group had no motor or sensory deficits by the time of evaluation by the trauma surgery team.

Registry data collected included demographics, Injury Severity Score (ISS), length of stay, and emergency department (ED) disposition. Further data detailing diagnostic imaging, ED and surgery physical findings, neuroSurgical consultation management, and mechanism of injury were obtained from patient electronic medical records. Radio- logic charges were approximated based on the specified imaging International Classification of Diseases, Ninth Revision, code and the hypo- thetical assumption of patient self-pay status. Charges did not include radiologist fees.

Descriptive statistics including means, SD, counts, and percentages were calculated. Fisher exact tests were used to compare categorical variables. The Wilcoxon rank sum test and the Student t test were used for comparisons of continuous and ordinal variables, and frequen- cies of continuous variables were reported as mean +- SD. Statistical sig- nificance was set at Pb .05, and all reported P values are 2 tailed.

Results

Thirty patients were evaluated for BSS during the study period; of these, 14 and 16 patients were classified as ND and PD, respectively. No differences were found between the groups in age, race, and length

than or equal to 2 (Pb .05). Fewer ND patients were admitted to the hos- pital compared with PD patients (57.1% vs 93.8%, respectively; Pb .05). Table 2 demonstrates the subjective clinical findings at initial pre- sentation in the ED. Those in the PD group presented with a higher inci- dence of reported decreased motor symptoms compared with the ND group (75.0% vs 30.8%; Pb .05). Although the PD group was more likely to have loss of consciousness, headaches, neck pain, back pain, and de- creased sensation compared with the ND group, the differences were not significantly different. Although the ND group had no motor or sen- sory physical findings, 30.8% and 71.4% of patients in this group subjec- tively reported motor and sensory symptoms either at the time of the accident or in the ED before evaluation by the trauma surgery service, respectively. Patients reported cervical spine tenderness at a high rate (85.7%) that is uncharacteristic of BSS. On the other hand, 100% and 75% of patients in the PD group had decreased sensation and decreased motor symptoms, respectively, on presentation. Some of these patients were found to have lower extremity motor deficits, which are, by defi- nition, not BSS. However, those patients with only upper extremity motor deficits were found to have positive cervical MRI findings com-

pared with those having upper and lower extremity deficits.

Table 3 shows the physical examination findings by the trauma surgery team. Although 85.7% of ND patients presented with cervical spine tenderness compared with 56.3% of PD patients, this difference was not statistically significant (PN .05). The most common motor/ sensory deficit finding for the PD group was left upper extremity motor weakness. Within 24 hours, 71.4% of the PD patients reported resolution of their symptoms. Most patients demonstrated complete resolution of symptoms on discharge; however, this difference was not statistically significant (PN .05). No patients reported worsening of symptoms on discharge.

Table 4 shows the comparison of imaging evaluation between both groups. Positive deficit patients had no positive cervical spine CT find- ings; however, 1 ND patient exhibited asymmetry on CT scan, and sub- sequent MRI revealed a ligamentous injury. In addition, 3 (1 other ND, 2 PD) patients had ligamentous injury, spinal cord edema, and disc pro- trusion as shown by cervical spine MRI. The PD patients with cervical spine MRI findings demonstrated only upper extremity motor deficits on physical examination. Of the 4 PD patients with positive findings on Thoracic and lumbar MRI films, 3 demonstrated no thoracic or lum- bar tenderness with the upper and lower extremity motor/sensory

Table 2

Clinical findings upon initial presentation to the ED

of stay (Table 1). All included patients were male, and the most com-

mon mechanism of injury was football tackle. Ten percent of patients						ND (n = 14)	PD (n = 16)	P
in the ND group compared with 69.2% in the PD group had an ISS greater					Loss of consciousness, n (%)			N.05
					Yes	4 (28.6)	8 (50.0)
					No	9 (64.3)	4 (25.0)
Table 1					Questionable GCS score, n (%)	1 (7.1)	4 (25.0)	N.05
Patient demographics					15	13 (92.9)	13 (81.3)
	ND (n = 14)	PD (n = 16)	P	14 13		1 (7.1) 0 (0.0)	2 (12.5) 1 (6.3)
Age (y), mean +- SD	14.5 +- 1.9	14.1 +- 2.7	N.05	b13		0 (0.0)	0 (0.0)
Race, n (%)			N.05	Headache, n (%)				N.05
White	10 (71.4)	11 (68.8)		Yes		6 (50.0)	9 (81.8)
African American	4 (28.6)	4 (25.0)		No		6 (50.0)	2 (18.2)
Unknown	0 (0.0)	1 (6.3)		Neck pain, n (%)				N.05
Men, n (%)	14 (100.0)	16 (100.0)	–	Yes		9 (69.2)	11 (84.6)
ISS, mean +- SD	1.6 +- 1.2	3.8 +- 3.1	b.05	No		4 (30.8)	2 (15.4)
Mechanism of injury, n (%)			N.05	Back pain, n (%)				N.05
Football tackle	9 (64.3)	8 (50.0)		Yes		5 (71.4)	7 (77.8)
Football helmet to helmet	4 (28.6)	4 (25.0)		No		2 (28.6)	2 (22.2)
Other^a	1 (7.1)	4 (25.0)		Decreased motor symptoms, n (%)				b.05
ED disposition, n (%)			b.05	Yes		4 (30.8)	9 (75.0)
Discharge	6 (42.9)	1 (6.3)		No		9 (69.2)	3 (25.0)
Admission	8 (57.1)	15 (93.8)		Decreased sensation, n (%)				N.05
Length of stay (h), mean +- SD	17.8 +- 9.4	28.3 +- 25.9	N.05	Yes		10 (71.4)	14 (100)

^a Hockey check to boards, baseball collision at home plate, wrestling tackle, and rugby tackle.

No 4 (28.6) 0 (0.0)

Abbreviation: GCS, Glasgow Coma Scale.

1444 R. Sola Jr et al. / American Journal of Emergency Medicine 34 (2016) 1442–1445

Table 3

Physical examination findings on trauma surgery evaluation

Physical examination	ND (n = 14)	PD (n = 16)	P
Cervical spine tenderness, n (%)			N.05
Positive	12 (85.7)	9 (56.3)
Negative	2 (14.3)	7 (43.8)
thoracic spine tenderness, n (%)			N.05
Positive	3 (21.4)	8 (53.3)
Negative	11 (78.6)	7 (46.7)
lumbar spine tenderness, n (%)			N.05
Positive	4 (28.6)	3 (21.4)
Negative	10 (71.4)	11 (78.6)
RUE motor weakness, n (%)			b.05
Positive	0 (0.0)	9 (56.3)
Negative	14 (100)	7 (43.8)
RLE motor weakness, n (%)			b.05
Positive	0 (0.0)	6 (37.5)
Negative	14 (100)	10 (62.5)
LUE motor weakness, n (%)			b.05
Positive	0 (0.0)	13 (81.3)
Negative	14 (100)	3 (18.8)
LLE motor weakness n (%)			b.05
Positive	0 (0.0)	9 (56.3)
Negative	14 (100)	7 (43.8)
RUE sensory weakness, n (%)			b.05
Positive	0 (0.0)	8 (50)
Negative	14 (100)	8 (50)
RLE sensory weakness, n (%)			N.05
Positive	0 (0.0)	3 (18.8)
Negative	14 (100)	13 (81.3)
LUE sensory weakness, n (%)			b.05
Positive	0 (0.0)	8 (50)
Negative	14 (100)	8 (50)
LLE sensory weakness, n (%)			b.05
Positive	0 (0.0)	6 (37.5)
Negative	14 (100)	10 (62.5)
Symptoms at discharge, n (%)			N.05
Resolved	10 (71.4)	12 (75)
Improved	3 (21.4)	4 (25)
No change	1 (7.1)	0 (0.0)

Abbreviation: RUE, right upper extremity; RLE, right lower extremity, LUE, left upper ex- tremity; LLE, Left lower extremity.

deficits. No surgical interventions were required for the MRI findings. Charges incurred by CT imaging totaled approximately US $89 944 and charges for MRI scans totaled approximately US $104 803.

Table 4

Imaging evaluation results
	ND	PD	P
CT cervical spine, n (%)	n = 11	n = 15	N.05
Positive	1 (9.1)^a	0 (0.0)
Negative	10 (90.9)	15 (100)
MRI cervical spine, n (%)	n = 7	n = 10	N.05
Positive^b	2 (28.6)	2 (20.0)
Negative	5 (71.4)	8 (80.0)
CT thoracic spine, n (%)	n = 1	n = 5	–
Positive	0 (0.0)	0 (0.0)
Negative	1 (100.0)	5 (100.0)
MRI thoracic spine, n (%)	n = 2	n = 9	–
Positive	0 (0.0)	2 (22.2)^c
Negative	2 (100.0)	7 (77.8)
CT lumbar spine, n (%)	n = 1	n = 5	–
Positive	0 (0.0)	0 (0.0)
Negative	1 (100.0)	5 (100.0)
MRI lumbar spine, n (%)	n = 2	n = 7	–
Positive	0 (0.0)	2 (28.6)^d
Negative	2 (100.0)	5 (71.4)

^a Positive findings included asymmetry.

^b Positive findings included disc protrusion, ligament injury, and spinal cord edema.

^c Positive findings included disc protrusion.

^d Positive findings included fracture and disc bulge.

Discussion

Burner and stinger syndrome is defined as motor and sensory symp- toms of a unilateral upper extremity without lower extremity symp- toms. Patients with BSS rarely experience neck pain and most often have transient symptoms lasting for seconds to minutes [1]. The Seddon’s criteria classify BSS by grades, which vary from transient motor and sensory deficits that resolve within 2 weeks to total axonal disruption with symptoms persisting for at least 1 year [3,8].

Our patient population demonstrates the difficulty and variability of identifying those with BSS after a sports collision. In the absence of a standardized protocol, variances in self-reported symptoms at presen- tation can often lead to imaging evaluation according to ED and surgeon preference. This is reflected in the results of our study, in which nearly all of the patients in the PD group and more than 75% of those in the ND group underwent cervical spine CT scan.

A prospective, multicenter study by Viccellio et al [9] found Cervical spine injuries in only 30 (0.98%) of more than 3000 children involved in blunt trauma. Despite a relatively low incidence of cervical spine inju- ries, published reports have shown a 4- to 9-fold increase in cervical spine scans for children from 1996 to 2010. The risk of radiation- induced solid cancer is projected to be 1 case in every 270 to 800 spine CT scans, depending on age [10]. Furthermore, liberal use of CT does not increase yield of injury identification and management, reduce mortality and morbidity, or increase the cost-effectiveness of care [11]. Published reports have demonstrated increased incidence of cervical spine CT and higher doses of radiation in children transferred from out- side facilities to level I trauma centers and pediatric trauma centers [4,12]. Moore et al [13] have shown that children undergoing CT scans for mechanism alone are unlikely to have a clinically significant positive CT finding and are subjected to an increased risk of malignancy. In addi- tion, the literature suggests that the potential benefit of performing rou- tine CT scans on low-risk cases may not outweigh the potential harms to children in terms of cancer risk [5].

Based on the literature, MRI may be the modality of choice to evalu- ate for nerve root injuries as well as other structural injuries such as her- niated disks, ligament injury, and even nondisplaced fracture [1,3,14]. Henry et al [7] demonstrated that MRI had a sensitivity and specificity of 100% and 97%, respectively, for identifying osseous injury, whereas CT had a poor sensitivity of 23% for identifying soft tissue injuries. Our data suggest that MRI alone may be the ideal radiographic imaging tool to evaluate BSS if concerned for neurologic injury.

The sample size of our study somewhat limits the significance of our findings; however, this study is the largest sample to date that evaluates this patient population. Given the retrospective design of this study re- view, subjective and objective clinical findings were obtained from the medical record and not from standardized questions. The imaging mo- dality was determined by preference of the attending physicians in the ED and attending trauma surgeons. Therefore, whether more posi- tive MRI findings would have been obtained in the presence of standard criteria for MRI indicated by motor or sensory deficits is unknown. Liter- ature has shown that a cervical spine protocol has reduced the radiation exposure for children [15]. Lastly, we are limited in understanding why our serious injury rate was 0%. It is unclear from our results whether permanent injuries are truly rare in this patient population or if that they were not identified due to our lack of long-term follow-up. Further studies with standardized history, physical, imaging guidelines, and long-term follow-up would strengthen the identification of children with possible BSS.

To our knowledge, this is the first study to investigate the manage- ment of BSS in the pediatric trauma population. Our study found that children with clinical findings concerning with BSS showed improve- ment in more than 70% of their symptoms within 24 hours with no need for surgical intervention. Our small study suggests that most BSS cases involve temporary symptoms. Computed tomographic imaging did not provide Clinically useful information for diagnosis, whereas

R. Sola Jr et al. / American Journal of Emergency Medicine 34 (2016) 1442–1445 1445

MRI did identify several injuries. Based on these limited data, we sug- gest that management of BSS should include observation, serial neuro- logic examination, and MRI evaluation as appropriate.

Conflicts of interest

All authors have no conflicts of interests to disclose regarding this study.

Authors’ contributions

Study conception and design: Sing. Acquisition of data: Sola, Eubanks, Raynor.

Analysis and interpretation of data: Sola, Eubanks, Christmas, Thomas, Fischer, Sing.

Drafting of manuscript: Sola, Christmas, Thomas, Fischer, Sing. Critical revision: Sola, Christmas, Thomas, Fischer, Eubanks, Sing, Raynor. Approval of final manuscript: All authors.

Acknowledgments

The authors acknowledge Jennifer C. Barnes, PhD, ELS, CMPP for critical review and editing of the manuscript.

References

AJEM