Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

A Prospective Multicenter Study of Cervical Spine Injury in Children.

Viccellio P, Simon H, Pressman BD, et al.

Pediatrics 2001;108 e20 [abstract][full-text]

Reviewed by Kathleen Moynihan Hardart MD, Children's Hospital, Boston

Review posted August 19, 2002

N.B. This template is based upon McGinn et al [7], with additional guidance from Laupacis et al.[8] and Stiell et al.[9].

I. What is being studied?

Study objective:

The purpose of this study was to prospectively examine the incidence and spectrum of cervical spine injury in patients younger than 18 years and to evaluate the efficacy of the National Emergency X-Radiography Utilization Study (NEXUS) decision instrument for obtaining cervical spine radiography in pediatric trauma patients. Only fractures and not soft tissue injuries of the cervical spine are included as outcomes.

Study design

This was a prospective, observational study involving those medical centers participating in the National Emergency X-radiography Utilization Study (NEXUS)[1,2]. NEXUS is a multi-center study organized to validate 5 criteria suggestive of low probability of cervical spine injury (CSI) and to test the hypothesis that patients with blunt trauma who meet these 5 criteria have low likelihood of CSI.

Twenty-one centers including academic institutions, community hospitals, tertiary care centers, trauma centers and children's hospitals participated. Emergency Department physicians determined which blunt trauma patients required radiography of the cervical spine. Those selected patients underwent a minimum of X-table lateral, AP and odontoid views. Oblique views, flexion-extension views or cervical CT or MRI were ordered at the discretion of the physician. CT or MRI was obtained if plain film imaging was not feasible. For those patients undergoing radiographic investigation the physicians filled out a data form that elicited information regarding demographics, clinical stability and presence of absence of the following factors: 1) tenderness at posterior midline c-spine, 2) focal neurologic defect, 3) alteration in alertness, 4) evidence of intoxication, 5) distracting pain. Once the data was entered, a voucher was issued in order to obtain radiographs. Study radiologists, blinded to the results of the study questions interpreted all films. If the results were ambiguous, study investigators, also blinded to the findings on the NEXUS forms, evaluated the radiographs to determine the presence of a CSI. This study limited their evaluation to those trauma patients < 18 years old who were included in the NEXUS project. Results were evaluated and reported for several subgroups of children divided by age [0-2 yr (no verbal ability), 2-8 yr (immature c-spine), 8-17 yr (fully developed spinal anatomy), > 18 yr (adult)].

What level of evidence (from the list below) supports this clinical prediction rule?

LEVEL II

Hierarchy of Evidence for Clinical Prediction Rules

Level I: Rules that can be used in a wide variety of settings with confidence that they can change clinician behavior and improve patient outcomes

• At least one prospective validation in a different population and one impact analysis, demonstrating change in clinician behavior with beneficial consequences.

Level II: Rules that can be used in various settings with confidence in their accuracy

• Demonstrated accuracy in either one large prospective study including a broad spectrum of patients and clinicians, or validated in several smaller settings who differ from one another.

Level III: Rules that clinicians may consider using with caution and only if patients in the study are similar to those in your clinical setting

• These rules have been validated in only one narrow prospective sample.

Level IV: Rules that need further evaluation before they can be applied clinically

• These CPRs have been derived but not validated or have only been validated in split samples, large retrospective databases, or by statistical techniques.

II. Are the results of the study valid?

Clinical prediction rule derivation

1. How were predictors chosen and defined?

The authors published a study in the Annals of Emergency Medicine in entitled. "Low-Risk Criteria for Cervical Spine Radiography in Blunt Trauma: A Prospective Study"[3]. This study chose several clinical factors identified in previously conducted retrospective studies as being associated with low-risk of CSI and attempted to demonstrate that these criteria could prospectively predict those patients at low risk of CSI. The authors examined the following criteria: 1) normal alertness 2) no evidence of intoxication 3) no neck tenderness or pain and 4) no distracting injuries/pain. The 5th criteria ultimately used in the development of the Clinical Decision Rule ie. no neurologic deficit, was chosen based on its association with low-risk of CSI in another published trial [4].

2. How was the selection of study subjects performed?

The study described in "Low-Risk Criteria for Cervical Spine Radiography in Blunt Trauma: A Prospective Study" included 1000 consecutive patients presenting to the Emergency Department at UCLA Medical Center with a chief complaint of blunt trauma for whom c-spine films were ordered and for whom prospective data questionnaires were completed. l these predictors seem reasonable.

3. Was the sample size adequate (including adequate number of outcome events)?

The authors did not justify the sample size of the subjects enrolled in the 1992 study as there was no indication that a power estimate was prospectively determined. Indeed, the small number of fractures diagnosed and the resultant wide confidence interval for the sensitivity determination suggest that a larger study might have demonstrated a more powerful and sensitive tool. In the discussion they mentioned that almost 7000 pts would have been necessary to be 95% certain that the true sensitivity is no less than 99%. The authors also conducted a Bayesian meta-analysis combining data from this and 2 other prospective studies that suggested there is a 97.5% chance that the false negative rate is at most less than 4% and a 74% chance that the false negative rate is less than 1. Nevertheless, this derivation study remains somewhat underpowered.

4. Were all important predictors present in the study population?

At least one of the four predictors 1) normal alertness 2) no evidence of intoxication 3) no neck tenderness or pain and 4) no distracting injuries/pain was present in all patients without CSI. Many of these criteria were also present in those with CSI. Only the presence of midline neck tenderness was significantly more common in those with CSI (P < 0.001). No other known predictors were excluded.

5. Does the rule make clinical sense?

The rule does make clinical sense. It is an easily remembered rule that requires no calculators or mathematics. The five clinical criteria all reflect less serious neck injury and thus logically are associated with low-risk for CSI.

Clinical prediction rule validation

1. Has the rule been validated?

The authors derived the clinical decision rule combining what they learned in the above mentioned paper testing 4 clinical criteria associated with low-risk for CSI with a fifth criteria derived from other clinical studies and designed a validation study published in 2000 and entitled, "Validity of a Set of Clinical Criteria to Rule Out Injury to the Cervical Spine in Patients with Blunt Trauma"[1]. This study. a component of the NEXUS project, looked at 34,069 patients from 21 medical centers who underwent radiography of the c-spine after blunt trauma. The study instrument detected all but 8 of the 818 patients with CSI.

From this same dataset, the authors evaluated those patients < 18 years who were at risk for CSI and attempted in the current paper (Vicellio et al) to validate the clinical decision rule for this population. The authors adequately demonstrate the validity of this clinical decision rule for the detection of c-spine fracture.

2. Was the validation population sufficiently different than the derivation population (to allow for a level II recommendation)?

The patients in the derivation paper were all cared for in the Emergency Department at the UCLA medical center. 59.3% of this group were male and the average age was 25y (range 6 mo-98 yr). The patients in the large NEXUS validation study represented patients from academic institutions, community hospitals, tertiary care centers, trauma centers and children's hospitals. 58.7% of these patients were male and the average age of this population was 37 yr (range < 1 yr - 101 yr). The population evaluated in the study of interest are those patients in the NEXUS study < 18 years. Thus the validation population, while representing a more narrow age range, reflects patients from a much broader patients pool.

3. Were the patients chosen in an unbiased fashion and do they represent a wide spectrum of severity of disease?

Emergency Department physicians determined which blunt trauma patients required radiography of the cervical spine. All patients who were to receive a radiographic c-spine evaluation were involved in the study. There was a clear potential selection bias in that only those who were chosen for radiography were entered into the study. Those blunt trauma patients for whom suspicion of injury was so low as to not consider ordering films were not entered into the study.

4. Was there a blinded assessment of the criterion standard for all patients?

All radiographs of the cervical spine were interpreted by study radiologists only. A diagnosis of c-spine fracture was based on final interpretation of all studies. Study investigators reviewed the report and films of any ambiguous interpretation. Both the radiologist and study investigator were blinded to the NEXUS data form information.

5. Was there an explicit and accurate interpretation of the predictor variables and the actual rule without knowledge of the outcome?

All ED physicians had a brief training in NEXUS criteria. All data forms were completed prior to obtaining the radiographs. Physicians were asked to note the demographics, clinical stability and presence or absence of the 5 clinical criteria. The radiographic findings were not known at the time the forms were completed.

6. Was there 100% follow up of those enrolled?

Follow-up was 100%.

III. What are the results?

1. What is the clinical prediction rule and how predictive is it? (Sensitivity, specificity, positive and negative predictive values at least should be available; ideally, likelihood ratios should also be provided

The clinical prediction rule is that patients who sustain blunt trauma and demonstrate ABSENCE OF ALL FIVE of the following criteria - 1) tenderness at posterior midline c-spine, 2) focal neurologic defect, 3) alteration in alertness. 4) evidence of intoxication, 5) distracting pain - are at low risk for CSI and do not warrant radiographic evaluation.

Sensitivity, specificity, positive and negative predictive values were determined using the following data and defining a negative test as one where all five criteria were fulfilled and a positive test as one in which at least one criteria was not fulfilled:

	Positive Test	Negative Test	totals
Presence of Disease	30	0	3
Absence of Disease	2432	603	3035
totals	2462	603

Characteristic	Pediatric Participants
Sensitivity	100% (87.8-100%)
Negative Predictive Value	100% (99.2-100%)
Specificity	19.9% (18.5-21.3%)
Positive Predictive Value	1.2% (0.8%-1.8%)

The negative likelihood ratio (the likelihood of a negative test in a patient with disease compared to a patient without disease) is 0.08 (95% CI 0.005, 1.27). This is quite small, and lends strong support that it is extremely unlikely that cervical spine injury would be present in an individual with a negative test (none of the 5 criteria positive).

2. Has an impact analysis been performed?

No impact analysis has been performed.

IV. Will the results help me in caring for my patients?

1.Will the reproducibility of the test result and its interpretation be satisfactory in my setting?

In light of the broad range of institution settings represented in the validation study, this clinical decision rule could be useful in my setting which is a tertiary care pediatric hospital. A larger study designed specifically to validate the rule in pediatric patients might be more reliable.

2. Are the results applicable to my patient?

This clinical decision rule helps physicians determine which patients are at high risk for "cervical spine injury" - specifically cervical spine fracture. This limits the utility of this rule for the prediction of CSI in pediatrics as children are much more likely to have ligamentous and soft tissue injury that are often not appreciated on plain radiographs. The cervical spine is quite different in children than in adults and for several reasons more prone to soft tissue injury. Children have larger head size in relation to the spine leading to greater flexion and extension injuries. Their smaller neck muscle mass also predisposes them to ligamentous injuries more commonly than fractures. They are also known to have increased flexibility of their intraspinous ligaments. Consequently c-spine fractures are less common than ligamentous injury. Interestingly, in children < 8y, only 30% of fractures are below C3 and SCIWORA (spinal cord injury without radiographic abnormality) represent 35-50% of SCI. For these reasons, children often require additional radiographic imaging such as flexion, extension films, CT scan or what is becoming the test of choice for soft tissue injury, MRI [5, 6]. Interestingly, SCIWORA was not observed (or appreciated) in this large study.

In addition to not testing for soft tissue injury, the clinical decision rule requires a reliable patient able to acknowledge pain and report neurologic symptoms suggesting focal injury. As children < 3-4 are usually unreliable, this rule cannot be safely applied.

3. Will the results change my management?

The clinical decision rule has been adapted into an algorithm for c-spine clearance by our institution. If a patient has no neurologic deficit and is conscious, is cooperative with exam, pain free, not tender on palpation and has no pain or tenderness on active range of motion, a patient can be cleared with no radiographic evaluation. Any patient who is unreliable or has any concerning symptoms is evaluated radiographically, often including images to rule out soft tissue injury.

4. Will patients be better off as a result of the test?

Patients will benefit from the lack of radiation exposure, cost saving and time saved in the E.D. evaluation process. The consequences of a missed CSI however are devastating and thus, strict adherence to the rule, appreciating the shortcomings of its predictive ability is essential.

References

1. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI, for the National Emergency X-Radiography Utilization Study Group. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. N Engl J Med 2000; 343:94-99. [abstract]
2. Hoffman JR, Wolfson AB, Todd KH, Mower WR. Selective cervical spine radiography in blunt trauma: methodology of the National Emergency X-Radiography Utilization Study (NEXUS). Ann Emerg Med 1998;32:461-469. [abstract]
3. Hoffman JR, Schriger DL, Mower WR, Luo JS, Zucker M. Low-risk criteria for cervical-spine radiography in blunt trauma: a prospective study. Ann Emerg Med 1992;12:1454-1460. [abstract]
4. Roberge RJ, Wears RC. Evaluation of neck discomfort, neck tenderness and neurologic deficits as indicators for radiography in blunt trauma victims. J Emerg Med 1992;10:539-44. [abstract]
5. Cantor RM, Leamey JM. "Pediatric Trauma" in Marx: Rosen's Emergency Medicine: Concepts and Clinical Practice, 5th ed. Mosby 2002; 275-276.
6. Rekate HL, Theodore N, Sonntag VKH, Dickman CA. Pediatric spine and spinal cord trauma: state of the art for the Third Millennium. Child Nerv Sys 1999;15:743-750. [abstract]
7. McGinn TG, Guyatt GH, Wyer PC, Naylor CD, Stiell IG et al. Users' Guides to the Medical Literature XXII: How to Use Articles About Clinical Decision Rules. JAMA 2000;284:79-84. [abstract; full-text]
8. Laupacis A, Sekar N, Stiell IG. Clinical Prediction Rules: A Review and Suggested Modifications of Methodological Standards. JAMA 1997;277(6)488-494. [abstract]
9. Stiell IG, Wells GA. Methodologic standards for the development of clinical decision rules in emergency medicine. Ann Emerg Med 1999;33:437-447. [abstract]

 

Comments

Back to the EB Journal Club Index