a b s t r a c t

Study objectives: We sought to validate National Emergency X-Radiography Utilizations Study low-risk cervical spine (C spine) criteria in a geriatric trauma population. We sought to determine whether patients’ own baseline mental status (MS) could substitute for Glasgow Coma Scale (GCS) to meet the criteria “normal alertness.” We further sought to refine the definition of “distracting injury.” Methods: This is a retrospective review of geriatric fall patients presenting to a level 1 trauma center and triaged to the trauma bay. We queried our database from 2008 to 2013. Abstractors recorded GCS, deviation from base- line MS, midline Neck tenderness, intoxication, focal deficit, signs of trauma, and presence of other injury. Patients were considered at baseline MS if specific documentation was present on the chart, or if their GCS was 15.

Results: Six hundred sixty elderly fall patients were trauma alerts during the study period. Seventeen were excluded for incomplete records/death before imaging, leaving 647. The median age was 81 (interquartile range, 74-87). Fifty patients (8.0%) had C spine or cord injury. Two hundred ninety-four (44.5%) had baseline MS (including GCS 13-15), no spine tenderness, no intoxication, and no focal neurologic deficit. Of these, 18 had C-spine injury. Using Physical findings of head trauma as the only “distracting injury,” no injury would have been missed (sensitivity, 100% [confidence interval, 91.1-100]; specificity, 14.2%).

Conclusions: Our study suggests that National Emergency X-Radiography Utilizations Study can be safely applied in elderly fall patients who are at their personal baseline MS. Furthermore, our data support a more narrow definition of distracting injury to include only patients with signs of trauma to the head.

(C) 2015

Introduction

The National Emergency X-Radiography Utilizations Study (NEXUS) criteria, developed in the early 90s, is a clinical decision tool emergency medicine physicians use to safely rule out cervical spinal injuries (CSI) in the blunt traumatic patient without radiographic imaging [1]. The Nexus criteria are met if a patient denies posterior cervical tenderness on palpation, is not intoxicated, has normal alertness, has no focal neu- rologic deficits, and has no painful distracting injuries. The NEXUS has demonstrated effectiveness at finding Cervical spine injuries, with a sen- sitivity of 99.6% in a large prospective trial [2,3].

Since the introduction of NEXUS, several studies have questioned its reliability in detecting CSI in elderly (>= 65 years) patients [4,5]. Elderly patients often have CSI from lower mechanisms of injury, such as falls, lacking the more obvious, and comorbid distracting injuries that would otherwise make them NEXUS positive [5]. Older age is an inde- pendent risk factor for CSI in blunt trauma and is a commonly cited rea- son for failing to apply NEXUS in clinical care [3,6].

? Conflict of interest statement: The authors have no conflicts of interest to report.

?? Presentations: Society for Academic Emergency Medicine, Dallas, TX, May 2014.

* Corresponding author. Tel.: +1 610 838 6147.

E-mail address: rebeccajeanmonod@yahoo.com (R. Jeanmonod).

The application of NEXUS clinically is subject to variability due to the subjective nature of “altered alertness [7].” In fact, some studies on NEXUS include “evaluable patients” with Glasgow Coma Scale (GCS) as low as 13 as NEXUS negative [8], whereas the original criteria de- scribe altered alertness as patients having a GCS of 14 or less [1]. This has direct implications on geriatric trauma care. A large proportion of elderly patients have baseline cognitive impairment, and it is unclear whether NEXUS can be safely applied in the elderly patient with a GCS less than 15 who is at their personal baseline mental status (MS) [9-11].

The interpretation of distracting injury is also subjective and is intentionally vaguely defined in the original NEXUS criteria [7,2]. Some studies have attempted to narrow the subjective definition of distracting injuries to limit them to upper torso trauma, but, currently, there is no standard agreement for what denotes distracting injury [8,12,13].

Our current study aims to validate NEXUS criteria in geriatric fall patients triaged to the trauma bay at a level 1 trauma center. We sought to identify whether GCS less than 15 in a patient described at their personal baseline MS altered the sensitivity/specificity of the NEXUS criteria. We additionally sought to determine if narrowing the definition of “distracting painful injury” to “any observable trauma to the head or face” had any impact on the sensitivity/specificity of NEXUS in this population.

http://dx.doi.org/10.1016/j.ajem.2015.05.031

0735-6757/(C) 2015

Table 1

Adult trauma alert criteria for falls

Vital signs	Anatomy of injury	Mechanism	Head trauma in anticoagulation Physician/EMS judgment
GCS b14 or combative SBP b 90 mm Hg RR b10 or N 29	flail chest Suspected pneumothorax >=2 long bone fractures	Fall N 20 ft	Abnormal neurologic examination Loss of consciousness Amnesia
Intubated Physiologic deterioration en route	Amputation proximal to wrist/ankle Mangled extremity Pelvis fracture Open/depressed skull fx Paralysis		Headache Nausea/vomiting

Abbreviations: EMS, emergency medical services; SBP, systolic blood pressure; RR, respiratory rate; fx, fracture.

Materials and methods

Study design

This study is a retrospective cohort of elderly fall patients triaged to the trauma bay. The research protocol was reviewed by the institutional review board at the study facility and found to be exempt.

Study setting and population

The study site is a level 1 community trauma center that hosts a trau- ma/critical care fellowship. Patients were eligible for enrollment if they were Age 65 years or older and were triaged to the trauma bay for fall. This included falls from standing, falls down stairs, falls from a height, and patients “found down.” Determination for trauma bay triage was made by medical command physicians at the receiving facility based on standard criteria posted in a prominent location next to the com- mand radio station (Table 1). Patients were excluded from the study if they were transferred from an outside facility for trauma evaluation be- cause radiographic studies were not always repeated at our facility, and NEXUS criteria on patient arrival at the outside facility were not reliably documented. Patients were also excluded if they were triaged to the main emergency department (ED) and then were later made trauma consults, as trauma consultation generally occurred after diagnosis of injuries, which introduces bias in the assessment of NEXUS criteria by the trauma team. For both these groups, there was concern that the knowledge of an injury might influence the recorded NEXUS criteria. In patients triaged directly to the trauma bay, NEXUS criteria are docu- mented before imaging.

Study protocol and measurements

Patients eligible for this study were identified by query of the elec- tronic trauma registry at the study site. The registry was queried from May 2008 to May 2013. These dates were chosen because, before May 2008, standard trauma documentation forms at the study institution did not contain specific documentation of NEXUS criteria.

Data obtained from the trauma registry included age, mechanism, injury severity score (ISS), GCS, and disposition. Three trained research associates reviewed each patient’s electronic medical record to populate other data fields. The physician trauma documentation form was reviewed to determine presence or absence of altered MS beyond GCS (for instance, some patients with GCS of 14 were specifically described as being at baseline MS, and some patients with GCS 15 were neverthe- less described as being altered), presence or absence of neck tenderness, signs of trauma to the head and face, and focal neurologic deficits. All ra- diology reports from the trauma evaluation were reviewed to deter- mine the presence of other injuries, and discharge/death summaries were reviewed to determine additional injuries that may have been missed in the trauma bay but found during hospitalization. Significant traumatic injuries such as visceral injuries or bony injuries were noted. soft tissue injuries such as abrasions, contusions, skin tears, and lacerations were not recorded, unless they occurred to the head and

face. Procedure reports were also reviewed in cases of patients requiring surgical intervention.

A patient was determined to have no significant neck injury if (1) he/ she had a negative computed tomography or magnetic resonance imag- ing of the neck, or (2) the patient did not undergo spine imaging but was admitted to the hospital and was observed and had no sequelae referable to the neck at discharge. All acute fractures or fractures of undetermined age, dislocations, and subluxations from C1 to T1 were recorded as positive for neck injury. All patients with magnetic reso- nance imaging findings were recorded as positive. Patients with meta- static spine lesions with no fracture or other traumatic injury were recorded as negative for neck injury.

Patients were deemed NEXUS negative if they had a GCS of 15, no midline neck tenderness, no clinical intoxication, no focal deficit, and no painful distracting injury including long bone fracture, signs of head trauma, or visceral injury. Sensitivity and specificity were calculat- ed. The same analyses were then carried out using the patients’ personal baseline MS as the standard for normal alertness, with all other NEXUS items remaining the same. We then carried out sensitivity and specific- ity calculations for NEXUS using GCS 15, no midline tenderness, no in- toxication, no focal deficit, and no “observable trauma to the head and face” as the only distracting injury. Finally, we applied both “patient is at personal baseline MS” for normal alertness and “no observable trauma to the head and face” as the only painful distracting injury, and sensitivity and specificity were calculated. Absent data fields were conservatively estimated to be positive/present. Data were entered by trained research associates into a standardized Microsoft Excel 2007 spreadsheet (Microsoft Corporation, Redmond, WA).

Data analysis

Descriptive statistics were used to assess demographic factors and presence and absence of specific NEXUS criteria. ?² was used to assess NEXUS sensitivity and specificity. Logistic regression was used to con- trol for ISS in patients with and without cervical spine injury. Data were analyzed using MedCalc (1993-2013; MedCalc, Ostend, Belgium) and VassarStats: Website for Statistical Computation (1998-2013; vassarstats.net, author Richard Lowry, PhD, Professor of Psychology Emeritus, Vassar College, Poughkeepsie, NY).

Results

Demographics

Six hundred sixty elderly patients with fall events were triaged to the trauma bay during the 5-year period. Medical records could not be located for 10 patients, and 7 patients died before completion of their radiographic evaluation, leaving 643 for analysis (Figure). Fourteen pa- tients did not undergo cervical spine imaging but were admitted and observed in the hospital with no sequelae at discharge. The demo- graphics of the enrolled patients are shown in Table 2.

Table 2

No intervention

n = 45

Surgical intervention

n = 5

Injured

n = 50

Un Injured

n = 570

Neck CT

n = 629

Inpatient obs without CT n = 14

No record or died

n = 17

660 Patients

Baseline characteristics

Total (n = 650)

Median age (IQR) 81 (74-87)

Sex (%)

Male 312 (48)

Female 338 (52)

Living environment (%)

Home 539 (82.9)

Assisted living and nursing home 110 (16.9)

Hospice 1 (0.15)

Position before fall (%)

Standing/sitting/lying	375 (57.7)
Down stairs	162 (24.9)
From height	36 (5.5)
Found down	77 (11.8)
Unknown	30 (3.8)

their noncervical spine-injured counterparts (median, 6; IQR, 4-14; P

b .0001).

Figure. Baseline characteristics of elderly fall patients triaged to the trauma bay.

Altered MS/GCS/intoxication

Three hundred ninety-nine patients (61.4%) did not have altered MS. Of these, 328 had a GCS of 15, and the remainder had a lower GCS with specific documentation on the chart that the patient was at baseline MS (2 with a GCS of 10, 4 with a GCS of 11, 6 with a GCS of 12, 10 with a GCS of 13, and 49 with a GCS of 14). Twenty-two patients with a GCS of 15 were described as having altered MS. Eleven patients had documenta- tion of clinical intoxication, and all of these had laboratory values consis- tent with intoxication. Other patients with laboratory detection of alcohol but clinical sobriety were not considered intoxicated.

Midline tenderness

Four hundred twelve patients (63.4%) had no midline tenderness. Eighty-seven (13.4%) had midline tenderness, and the remaining 151(23.2%) either had absent data fields or the patient could not be assessed due to altered MS or acuity.

New focal deficit

Five hundred twenty-eight patients (81.2%) had no new focal deficit on examination. Fifty-three patients (8.2%) had a new focal deficit. The remaining 69 (10.6%) either had absent data fields or the patient could not be assessed due to MS or acuity.

Distracting injury

One hundred thirty-four patients (20.6%) had distracting injuries discreet from head and face trauma. One hundred twenty-two of these were orthopedic injuries, with the remainder being visceral inju- ry. Four hundred fifty-seven patients (70.3%) had physical signs of face or head trauma.

Injuries

Fifty patients (7.8%) had cervical spine injuries. Five injuries re- quired surgery. Of these injuries, 29 involved C1 to C3. Five injuries only involved the spinous processes or transverse processes. Of the pa- tients with cervical spine injuries, 19 injured patients had altered MS, 28 either had tenderness or were unreliable on examination, and 6 had a focal deficit. Seventeen patients with injury had normal MS, no tenderness, and no focal deficit. Patients with cervical spine injuries had higher ISS (median, 11; interquartile range [IQR], 9-17.5) than

The NEXUS application

The criterion standard for NEXUS using GCS 15 to define normal alertness and broadly including distracting injuries had a sensitivity of 100% (confidence interval [CI], 91.1-100) and a specificity of 7.4% (CI, 5.5-9.8). Using change from baseline MS rather than GCS less than 15 as a determinant for abnormal alertness improved specificity to 9.5% (CI, 7.4-12.3) without reducing sensitivity. Using signs of head trauma as the only distracting injury improved specificity to 12.6% (CI, 10.1- 15.6) without reducing sensitivity (100%; CI, 91.1-100). Using change from baseline MS as a determinant for abnormal alertness and signs of head trauma as the only distracting injury, there were a total of 84 NEXUS negative patients (12.9%) seen in the trauma bay. The NEXUS had a sensitivity of 100% (CI, 91.1-100%) and a specificity of 14.1% (CI, 11.4-17.2%) in detecting cervical spine injury.

Controlling for ISS, patients with signs of head trauma were 3.6 times more likely to have a cervical spine injury (CI, 1.4-9.5) as com- pared with their nonhead-injured counterparts, and of all NEXUS criteria, this was the strongest predictor of injury. The NEXUS per- formed well in this elderly fall population, with no missed injuries.

Discussion

Trauma, particularly falls, is a significant cause of morbidity and mortality in elderly individuals. Up to a third of elders living indepen- dently fall annually, and 10% will have a significant injury [14,15]. Al- though the risk of radiation exposure is less consequential in this population of advanced age, there are good reasons to avoid unneces- sary immobilization and imaging. Immobilization is uncomfortable, and immobilization time is the most significant contributing factor in collar-related Skin breakdown [16]. In addition, medical imaging is cost- ly. Between 2000 and 2006, Medicare spending for advanced medical imaging more than doubled, and medical imaging has the potential to prolong patients’ ED stays or delay other interventions [17].

Missed cervical spine injury has tremendous potential for morbidity and necessitates that any decision rule used to determine need for med- ical imaging has near 100% sensitivity. The NEXUS has been shown to be just such a rule, but its individual components introduce substantial subjectivity to the rule. In a recent study, it was demonstrated that there was only fair agreement on level of alertness (? = 0.22) and slight agreement on distracting painful injury (? = 0.13) between residents and faculty in application of NEXUS [7]. Although individual provider practice styles and experience will always play a role in clinical medi- cine, this degree of disagreement in a widely used decision rule war- rants further clarification.

Normal alertness can be difficult to determine in the geriatric popu- lation. This particular cohort was largely composed of individuals living at home, but one cannot assume that those individuals have no cogni- tive impairment. In a voluntary survey, 12.7% of elderly individuals at home report memory impairment and intermittent confusion and that percentage is higher in ED population-based studies and nursing home studies [10,18,19]. Altered alertness implies a change from base- line, and, unfortunately, physicians often do not have access to informa- tion regarding a patient’s baseline. Furthermore, a patient may be at his or her personal baseline but may not be evaluable due to disability (for instance, in the case of severe dysarthria, hearing impairment, or prior stroke). When in doubt, it is always safer to perform advanced imaging in the elderly trauma patient. That said, there are patients with GCS less than 15 who are evaluable. Although it is beyond the scope of this article to describe in detail, most patients determined to be at baseline alert- ness but with a GCS less than 15 had lower GCS for baseline demen- tia/confusion or lost points for eye opening. Of the patients believed to be baseline MS, only 82% had GCS of 15. Our data support that patients with baseline unchanged cognitive impairment may have their cervical spines cleared clinically.

It makes intuitive sense that signs of trauma to the head would cor- respond to an increased risk of cervical spine injury, as it signifies force in close proximity to the neck. Other injuries requiring substantial force, such as Long bone fractures, did not provide additional sensitivity to NEXUS. This may be because these other injuries absorb some of the force of the trauma at a site distant from the neck (eg, in forearm frac- tures when the patient has assumed a defensive position during the fall to protect the head and neck) or because, as these injuries are geographically discreet from the neck, the patient is still able to localize neck tenderness and remains NEXUS positive. Patients with head injury generally have higher ISS, but even when controlling for ISS, any sign of head trauma was a significant predictor of cervical injury in this population. Our data support deferring cervical spine imaging in the elderly fall patient with long bone fracture or other injury below the neck provided the patient has no signs of trauma to the head and is oth- erwise NEXUS negative.

Limitations

All elderly patients enrolled in this retrospective cohort were triaged to the trauma bay. Because this typically represents a more injured co- hort, the incidence of disease is likely to be higher than an elderly pop- ulation presenting to an ED. That said, the higher incidence of disease should not affect sensitivity or specificity calculations but would affect positive and negative predictive value. At our institution, the trauma evaluation of each elderly patient was conducted by a trauma surgeon and not an emergency medicine physician, which questions the exter- nal validity of applying our study’s findings to patients being evaluated within the ED by ED staff.

In addition, our study focused specifically on elderly fall patients and did not assess for other mechanisms of injury. Therefore, our data should not be extrapolated to other forms of blunt trauma, such as motor vehicle crashes and assault, without further study.

Our study is a retrospective review and therefore relies heavily on the documentation within the medical record. Our trauma intake

sheet has a specified area devoted to NEXUS criteria, which we believe likely improve the reliability of the medical record. However, this study warrants prospective validation.

Finally, this study was performed at a single tertiary care trauma center, and results may not be able to be generalized to other facilities.

Conclusions

Our study validates NEXUS criteria in elderly patients presenting with fall. Furthermore, it supports expanding “normal alertness” to in- clude patients with GCS less than 15 who are at their personal baseline. In cases of blunt trauma in elderly patients due to fall, the definition of “painful distracting injury” may be narrowed to include signs of trauma to the head and face.

References

1. Hoffman JR, Wolfson AB, Todd K, et al. Selective Cervical spine radiography in blunt trauma: methodology of the National Emergency X-Radiography Utilization Study (NEXUS). Ann Emerg Med 1998;32(4):461-9.
2. Hoffman JR, Mower WR, Wolfson AB, et al. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency X- Radiography Utilization Study Group. N Engl J Med 2000;343(2):94-9.
3. Touger M, Gennis P, Nathanson N, et al. Validity of a decision rule to reduce cervical spine radiography in elderly patients with blunt trauma. Ann Emerg Med 2002; 40(3):287-93.
4. Barry TB, McNamara RM. Clinical decision rules and cervical spine injury in an elder- ly patient: a word of caution. J Emerg Med 2005;29(4):433-6.
5. Goode T, Young A, Wilson SP, et al. Evaluation of cervical spine fracture in the elder- ly: can we trust our physical examination? Am Surg 2014;80(2):182-4.
6. Morrison J, Jeanmonod R. Imaging in the NEXUS-negative patient: when we break the rule. Am J Emerg Med 2014;32(1):67-70.
7. Matteucci MJ, Moszyk D, Migliore SA. Agreement between resident and faculty emergency physicians in the application of NEXUS criteria for suspected cervical spine injuries. J Emerg Med 2015;48(4):445-9.
8. Konstantinidis A, Plurad D, Barmparas G, et al. The presence of nonthoracic distracting injuries does not affect the initial clinical examination of the cervical spine in evaluable blunt trauma patients: a prospective observational study. J Trau- ma 2011;71(3):528-32.
9. Hustey FM, Meldon SW. The prevalence and documentation of impaired men- tal status in elderly emergency department patients. Ann Emerg Med 2002; 39(3):248-53.
10. Nursing Home Compendium. [Internet] http://www.cms.gov/Medicare/ Provider-Enrollment-and-Certification/CertificationandComplianc/Down- loads/nursinghomedatacompendium_508. pdf; 2010. [Accessed 10/21/12].
11. Hustey FM, Meldon SW, Smith D, et al. The effect of mental status screening on the care of elderly emergency department patients. Ann Emerg Med 2003;41(5): 678-84.
12. Heffernan DS, Schermer CR, Lu SW. What defines a distracting injury in cervical spine assessment? J Trauma 2005;59(6):1396-9.
13. Kamenetsky E, Esposito TJ, Schermer CR. Evaluation of distracting pain and clinical judgment in cervical spine clearance of trauma patients. World J Surg 2013;37(1): 127-35.
14. Tinetti ME, Williams CS. Falls, injuries due to falls, and the risk of admission to a nursing home. N Engl J Med 1997;337:1279-84.
15. Stevens JA, Corso PS, Finkelstein EA, Miller TR. The costs of fatal and non-fatal falls among older adults. Inj Prev 2006;12:290-5.
16. Ackland HM, Cooper DJ, Malham GM, et al. Factors predicting cervical collar-related decubitus ulceration in major trauma patients. Spine 2007;32(4):423-8.
17. Government Accountability Office (GAO). Medicare part B imaging services: rapid spending growth and shift to physician offices indicate need for CMS to consider ad- ditional management practices. Washington DC: U.S. Government Accountability Office; 2008.
18. Centers for Disease Control and Prevention (CDC). Self-reported increased confusion or Memory loss and associated functional difficulties among adults aged >=60 years–21 states, 2011. MMWR 2013;62:347-50.
19. Gray LC, Peel NM, Costa AP, et al. Profiles of older patients in the emergency depart- ment: findings from the interRAI Multinational Emergency Department Study. Ann Emerg Med 2013;62:467-74.