Pediatric cervical spine injuries can occur in both bones and ligaments. These injuries are uncommon and the incidence of cervical spine injuries in children is reported anywhere from 1% to 4% of pediatric trauma patients admitted to major trauma centers.1 However, when blunt force trauma occurs in children, the proportion of the child's head to the body exerts excessive force not only on the bones of the cervical spine but also on the ligaments. Ligamentous injury is more common in children, especially younger children, than in adults.1 Early detection of ligamentous injury in children is important because missed injury may result in the dislocation of the cervical spine, spinal cord edema, hemorrhage within the spinal cord, and vertebral injury-any of these could result in neurologic deficit if undetected.2-4 Therefore, it is of particular importance to assess for injury to both cervical bones and ligaments in children.
Traditional adult guidelines to detect injury of the cervical spine include assessing for the presence of tenderness and obtaining routine radiographic images to look for bony and ligamentous injury. Alert patients can offer subjective information related to pain, which may indicate ligamentous injury. However, nonalert, noncommunicative, or unreliable patients pose an additional challenge to healthcare providers to detect ligamentous injury to the cervical spine. Traditional radiographs and computed tomographic (CT) scans do not show ligamentous injury; therefore, the use of magnetic resonance imaging (MRI) is often indicated. In the pediatric population in which ligamentous injury is more common, MRI may be even more important.5
This descriptive study sought to identify current practices and elicit expert data from a national population of trauma experts regarding pediatric cervical spine clearance in nonalert, noncommunicative, or unreliable injured children.
Currently, there is no known nationally recognized evidence-based guideline or protocol for clearing the cervical spine of injury in nonalert, noncommunicative, or unreliable injured children.2,5-7 The National Emergency X-Radiography Utilization Study (NEXUS) criteria have been developed for clearing the cervical spine injury in adults but have been found to lack sensitivity in young children and injuries in children are missed.8
There is significant variation in the practice, recommendations, and diagnostics for clearing the cervical spine in nonalert, noncommunicative, or unreliable children. Institutions vary in diagnostics and timing of their implementation of radiographs, CT scans, and MRI for clearing the cervical spine injury in children. Controversy exists as to which methods are most reliable and at what point to use them during blunt trauma injury management.2,4,9-11 In both alert and nonalert children, traditional radiographs are commonly taken, including anterior-posterior (A-P), lateral, flexion-extension, swimmer's, and/or open-mouth radiographs.2,4,12,13
In most institutions, if a cervical spine injury is discovered in a child, the cervical collar is continued and treatment begins along with the evaluation for additional injuries. However, if the child is nonalert, noncommunicative, or unreliable, some institutions obtain a CT scan of the cervical spine, other institutions obtain an MRI scan of the cervical spine, and some do both diagnostic tests.10,13 Variation also exists in the time frame for performing diagnostic tests. Some institutions and experts suggest obtaining an MRI scan after 24 hours,2,7 others after 48 hours,10,14,15 and still others after 72 hours or even days later.13
Uncertainty also exists as to when cervical spine injuries in nonalert, noncommunicative, or unreliable children can be treated using "adult" guidelines for diagnostic studies and time frames.12 On the basis of anatomical development, research studies and expert opinions have suggested that the use of "adult" guidelines should occur at ages 8, 12, 14, or 15 years.14-18 However, other studies have found determinations by age to be less reliable. In their study utilizing the NEXUS criteria in children, other researchers found that the sensitivity of plain radiographs in detecting injury in children younger than 8 years was 75%, and in children older than 8 years, it was 93%.8 Studies evaluating the use of a modified adult protocol for detecting cervical spine injury in children found that adult protocols are too rigid and lack sensitivity for use in children.8,19 Therefore, diagnostics used in children younger than 8 years may still need to be used for those older than 8 years.
Misdiagnosis and missed cervical spine injuries resulting in further injury are a significant problem. Traditional radiographs have been shown to be unreliable in detecting ligamentous injury in the pediatric patient.15,20 In addition, CT scans may miss ligamentous injury and fractures in children.21 Avellino et al21 reviewed medical records and radiologic data from 37 pediatric cervical spine injury cases for possible misdiagnosis. Overall, 7 cases were misdiagnosed. It was concluded that the use of routine radiographs and CT scans contributed to misdiagnosis in four children <=8 years and in three children between 9 and 14 years.21 One of the 7 misdiagnosed cases resulted in neurologic sequelae from a missed injury.
Reducing variation in how cervical spine injuries are diagnosed in the nonalert, noncommunicative, or unreliable pediatric patient can reduce costs. Clearing the cervical spine injury earlier has been found to decrease costs for intensive care unit (ICU) length of stay and decrease the risk of pressure ulcers from the cervical collar. A study by Frank et al22 found that the use of an MRI protocol for clearing the cervical spine injury in children resulted in the following: (1) decreasing time to clearance by almost 2 days; (2) decreasing the ICU stay by about 2 days; and (3) decreasing the hospital stay from 20.1 to 15.5 days, resulting in a $7,700 per patient savings.
The evidence from peer-reviewed literature indicates that a variety of diagnostics and time frames are used for clearing the pediatric cervical spine injury in blunt trauma; thus, an evidence-based guideline is clearly indicated. Reducing costs and reducing the number of missed injuries are just two of the advantages of creating a guideline for clearing the cervical spine in nonalert, noncommunicative, or unreliable pediatric patients. Literature supports the need for this study aiming to examine current effective practices for positive outcomes and begin the process for later development and testing of an evidence-based guideline for clearing the cervical spine injury in pediatric patients.
The primary objective of this study was to determine whether there is a standard practice for clearing the cervical spine injury of nonalert, noncommunicative, or unreliable pediatric blunt trauma patients presenting to trauma centers in which the Society of Trauma Nurses (STN) members practice. Two specific aims related to the primary objective were (1) to determine the age at which trauma programs treat pediatric cervical spine injuries using an adult guideline and (2) to determine which outcomes STN members would value to evaluate the usage of a cervical spine clearance guideline for nonalert, noncommunicative, or unreliable pediatric blunt trauma patients.
METHODS
We conducted a descriptive survey study to gather data on pediatric cervical spine clearance practices in trauma centers around the United States. The Pediatric Cervical Spine Clearance Survey, a 93-item electronic survey instrument, was developed with the input of various clinicians. Following an extensive literature review, an initial draft of the survey was developed by a graduate nursing student, a pediatric trauma clinical nurse specialist, and a senior nurse researcher. The survey was reviewed and edited by a pediatric trauma medical director and then sent to 10 STN Pediatric Special Interest Group (SIG) members, who reviewed and validated the survey items. After feedback was received, the survey was modified to reflect suggested changes. The research protocol including the final survey instrument was then reviewed and approved by the Spectrum Health Institutional Review Board.
Survey questions centered on the study objectives. Basic demographic information such as the size of the institution, volume of pediatric trauma patients admitted per year, and trauma verification agency were collected first. Respondents were asked whether their institution had a guideline for pediatric cervical spine clearance. Respondents were then asked to consider a description of a pediatric trauma patient presenting with a mechanism of injury suggestive of a possible cervical spine injury. The patient description was adapted from Lee et al, as cited by Eubanks et al.2
Questions then centered on the sequence and time frames of diagnostic studies used for cervical spine clearance of patients of different age groups presenting with this injury profile. Age groups were divided into the following: younger than 3 years, 3 to 8 years, older than 8 to 12 years, and older than 12 to 17 years. Responders were asked which diagnostic tests were used (lateral cervical spine radiograph, A-P cervical spine radiograph, flexion-extension cervical spine radiograph, odontoid cervical spine radiograph, swimmer's view cervical spine radiograph, CT scan, MRI scan, or other) and the target time frame for each test (within: first hour of arrival, emergency department [ED] length of stay, 24, 48, 72, after 72 hours, or other). Respondents were then asked to identify diagnostic tests that were never used by their hospital for cervical spine clearance. Finally, through 2 open-ended questions, responders were asked what supports or what barriers existed in their hospital systems to meet their target time frames for diagnostic tests.
Respondents were then asked to indicate which medical service(s) (trauma, orthopedic spine, neurosurgery, critical care, pediatric, or other) most frequently clear cervical spine injuries for pediatric patients of each age group. The final item asked which outcomes would be most valued in evaluating the use of a cervical spine protocol (missed cervical spine injury, amount of x-ray exposure, pressure ulcer from cervical collar, time with cervical collar on, length of stay, cost, patient comfort, patient satisfaction, and parent/guardian satisfaction).
Using SurveyMonkey, an e-mail link to the survey was sent to 309 RNs who were members of the STN Pediatric SIG. The survey instructions identified the voluntary nature of the research study and the presumed consent if respondents answered and submitted the survey. Respondents were given 2 weeks to complete the questionnaire.
Categorical and ordinal data were summarized using SPSS, Version 13.0. Responses to open-ended questions were analyzed for themes by the researchers. Responders were also given an opportunity when completing the survey to provide an e-mail address for future contact for a copy of the cervical spine guideline used in their institution. These protocols/guidelines are currently being received and analyzed for similarities and differences.
RESULTS
Forty-four responses were received during the survey time frame, which is a 14.2% return rate. Approximately 50% of the respondents were from a verified level I or level II pediatric trauma center. Another 30% of respondents were from a verified level I or level II trauma center. Of the 44 responders, 70% of institutions were verified by the American College of Surgeons and 54% had state or regional designation. Most respondents were from high-volume trauma centers treating more than 200 pediatric patients in all age groups. Most respondents noted that the trauma service was responsible for cervical spine clearance in all age groups.
Adult guidelines for cervical spine clearance were reported to be never used for pediatric patients by 42% of the respondents. A hospital-specific pediatric cervical spine clearance guideline was reported by 69% of the respondents, and many of the respondents offered to be contacted for a copy.
The number of respondents for each item varied throughout the survey; therefore, the number of respondents is provided throughout this section. In the age group younger than 3 years, 46% of respondents obtained a CT scan whereas 41% obtained a lateral cervical spine radiograph (n = 23). Of the respondents, 73% completed all first-line diagnostics within 1 hour of arrival (Fig 1). The second diagnostic test was reported as either a CT scan or an MRI scan 89% of the time and completed within 72 hours of arriving in the ED 90% of the time (n = 21) (Fig 2). The third diagnostic test was reported as either a CT scan (25%) or an MRI scan (42%) and completed within 72 hours of arriving in the ED 80% of the time (n = 12) (Fig 3).
For the age group 3 to 8 years, 41% obtained a lateral cervical spine radiograph first and 35% a CT scan; this was done within 1 hour of arrival 73% of the time (n = 18) (Fig 4). The second diagnostic test (n = 16) was a CT scan 32% of the time and an MRI scan 38% of the time, with 40% completed within the first hour of the ED visit (Fig 5). The third reported diagnostic in the sequence (n = 11) was a CT scan or an MRI scan 74% of the time, completed within 72 hours of arrival, as reported by 90% of respondents (Fig 6).
In the age group older than 8 to 12 years, 47% reported that they obtained a lateral cervical spine radiograph first and 30% a CT scan, which was done within 1 hour of arrival 100% of the time (n = 17) (Fig 7). For the second diagnostic (n = 16), 31% used a CT scan, 25% used A-P cervical spine radiograph, and 25% used an MRI scan, which were completed within 72 hours 100% of the time (Fig 8). The third sequence (n = 12) was a CT scan 16% of the time, odontoid radiograph 17% of the time, flexion-extension radiograph 8% of the time, and MRI 50% of the time, and 92% completed these diagnostics within 72 hours of arrival to the ED (Fig 9).
In the age group older than 12 to 17 years, 33% reported that they obtained a lateral cervical spine radiograph first, 13% A-P radiograph, and 40% a CT scan; this was done within 1 hour of arrival 100% of the time (n = 15) (Fig 10). The second diagnostic (n = 11) was a CT scan 27% of the time, MRI 37% of the time, and flexion-extension radiograph 18% of the time, with 100% completed within 72 hours (Fig 11). The third diagnostic in the sequence (n = 6) was odontoid radiograph 20% of the time, swimmer's view radiograph 20% of the time, and MRI 60% of the time, with 100% completed within 72 hours (Fig 12).
In all age groups, the odontoid radiograph was reported to be never used by more than 60% and the swimmer's view radiograph by more than 50% of the respondents. Another interesting finding was that no respondent reported never using a CT scan as a diagnostic tool.
The leading supports (mentioned more than once) that helped institutions meet their target time frames for diagnostic testing included the following:
* Trauma alert or activation and quick response of the team
* Immediate and accessible radiology
* Trauma team members who specialize in trauma and act as advocates for the patient
* Use of an algorithm or pathway to clear the cervical spine
The leading barriers to meeting the target time frames included the following:
* None
* Busy (in ED or with multiple-trauma admissions)
* Unstable patient
* Immediate surgery needed
* Waiting for personnel (surgeon, radiology, trauma team, anesthesiology, transport nurse)
* Unaware of guideline
* Poor coordination between services (unaware that the patient has priority)
Respondents reported that they would most value measuring outcomes related to missed cervical spine injury, time with cervical spine collar on, pressure ulcers from cervical spine collars, and the amount of x-ray exposure (Fig 13).
DISCUSSION
Our survey provides a preliminary snapshot of the current clinical practice of pediatric cervical spine clearance in a specialized population of children seen in trauma centers in the United States. A high percentage (70%, n = 38) of level I and level II (pediatric and adult) verified centers had a cervical spine guideline available, although it was reported in the comments section that a number of them did not follow the guideline or their trauma teams were unaware of its existence. In the survey, we should have asked the question "What percentage of the time is your guideline followed?"
Our study reveals that institutions in which STN members practice use a variety of diagnostic tests within a variety of time frames to clear the cervical spine of injury in the nonalert, noncommunicative, or unreliable pediatric blunt trauma patient. These findings were consistent whether or not the trauma center was verified and regardless of patient age groupings. Both the literature2,4,9-11 and our findings support the need for a nationally accepted standardized guideline for clearing the cervical spine in this specialized population.
In the literature, we found that an adult guideline may not be sensitive enough for use in the pediatric population. Our findings support this because respondents to the survey indicated that there is no standard age at which trauma programs treat pediatric cervical spine injuries using an adult guideline. These findings support the notion that the clearance of the cervical spine differs between adults and children and reinforces the need for a nationally recognized evidence-based guideline for pediatric cervical spine clearance.
Another specific aim of the study was to determine what outcomes would be most valued to measure the effectiveness of a pediatric blunt trauma cervical spine clearance guideline. From the literature, we found that time spent with the collar on, cost, missed injuries, and length of stay in the ICU and hospital were valued outcomes.21,22 The current study similarly demonstrated that time spent with the collar on and missed injuries are important outcomes, along with both reducing pressure ulcers from the collar and reducing the amount of x-ray exposure to the child. Therefore, a future guideline should be tested using these identified valued outcomes.
Valuable data were obtained from this survey, yet the conclusions we can draw are limited because of several factors. First, the survey did not account for the possibility of duplicate reports from a single institution. Also, the opinion of a single respondent may not adequately reflect the true practice of the overall institution. Second, sequential age groupings on the survey seemed to be a deterrent in the participants' responses. For each subsequent age grouping, the number of respondents decreased (eg, n = 23 for <3 years and n = 11 for >12-17 years). Age groupings required respondents to answer similar questions for different age groups, which may have been perceived as a long, confusing survey. Some respondents reported that the age groupings in the survey differed from their usual trauma center report groupings. This may have required more time and effort to calculate their data to complete the survey. Third, some centers may do a complete c-spine series (lateral, A-P, odontoid) as 1 test on all pediatric blunt trauma patients before proceeding to CT or MRI. In this case, the data may be skewed because the survey asked for the first, second, and third diagnostic tests and listed each radiograph separately. Finally, the survey was sent only to STN members who are involved in the Pediatric SIG. Therefore, our results do not reflect the practices of institutions in which healthcare professionals are not a part of STN and the Pediatric SIG.
CONCLUSION
Our results demonstrate that current practices for clearing the cervical spine in a special pediatric population are inconsistent. On the basis of the literature and the findings of this study, evidence-based recommendations are warranted and can result in a nationally recognized pediatric cervical spine clearance guideline for this special population. Clinicians who are expected to regularly evaluate a nonalert, noncommunicative, or unreliable pediatric blunt trauma patient for cervical spine injury would be greatly assisted by a nationally accepted guideline. Outcome measures of the effectiveness of the guideline should evaluate process improvements such as decreased variance in diagnostic testing, time frames for testing, and decreased cost. Patient outcomes may include reduced exposure to extra radiological procedures, improved comfort, and reduction in missed injuries. All of these outcomes are likely to significantly impact the child, family, and institution. We recommend that this project continue by obtaining guidelines from major trauma centers, analyzing them for similarities and differences, and creating an evidence-based guideline for clearing the cervical spine of injury in this special population of children. This could prompt a prospective, multicenter, randomized, controlled trial comparing outcomes of using the guideline versus usual practice.
Acknowledgments
The authors thank the membership of the STN and the STN Pediatric SIG leadership for their interest and support of this study. They also thank Deb Brown, chairperson for STN Pediatric SIG, for her leadership in this project. Finally, they thank Gretchen Schumacher, Assistant Professor, Grand Valley State University, for her assistance in writing the literature review.
REFERENCES