The Canadian C-spine rule performs better than ...

3/13/13 MD Consult - Print Previewer

Use of this content is subject to the Terms and Conditions

The Canadian C-spine rule performs better than unstructured physician judgment

Annals of Emergency Medicine - Volume 42, Issue 3 (September 2003) - Copyright © 2003 American
College of Emergency Physicians

DOI: 10.1067/mem.2003.293

The Canadian C-spine rule performs better than unstructured physician judgment

Glen Bandiera MD
Ian G. Stiell MD, MSc
George A. Wells PhD
Catherine Clement RN
Valerie De Maio MD, MSc
Katherine L. Vandemheen BScN
Gary H. Greenberg MD
Howard Lesiuk MD
Robert Brison MD, MPH
Daniel Cass MD
Jonathan Dreyer MD
Mary A. Eisenhauer MD
Iain MacPhail MD, MHSc
R. Douglas McKnight MD
Laurie Morrison MD
Mark Reardon MD
Michael Schull MD, MSc
James Worthington MBBS

From the Division of Emergency Medicine (Bandiera, Cass, Morrison, Schull), University of Toronto, Toronto, Ontario; the
Department of Emergency Medicine (Stiell, Greenberg, Reardon, Worthington), Department of Epidemiology and Community
Medicine (Wells), Division of Neurosurgery (Lesiuk), and Clinical Epidemiology Unit (Clement, DeMaio, Vandemheen),
University of Ottawa, Ottawa, Ontario; the Department of Emergency Medicine (Brison), Queen's University, Kingston, Ontario;
the Division of Emergency Medicine (Dreyer, Eisenhauer), University of Western Ontario, London, Ontario; and the Division of
Emergency Medicine (MacPhail, McKnight), University of British Columbia, Vancouver, British Columbia, Canada.

For the Canadian C-Spine and CT Head Study Group
Received for publication October 18, 2002.
Revisions received March 5, 2003, and April 9, 2003.
Accepted for publication April 15, 2003.

Dr. Stiell holds a Distinguished Investigator Award and Dr. Schull a New Investigator Award, both from the Canadian Institutes
of Health Research.

Supported by peer-reviewed grants from the Medical Research Council of Canada (MT-13699) and the Ontario Ministry of
Health Emergency Health Services Committee (11896N).

Reprints not available from the authors.

Address for correspondence: Glen Bandiera, MD, Sunnybrook and Women's College Health Sciences Center, Suite BG-13,

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 1/12

3/13/13 MD Consult - Print Previewer

2075 Bayview Avenue, North York, Ontario, Canada M4N 3M5.

Copyright © 2003 by the American College of Emergency Physicians.

0196-0644/2003/$30.00 + 0

Study objectives: We compare the predictive accuracy of emergency physicians' unstructured clinical judgment to the
Canadian C-Spine rule.
Methods: This prospective multicenter cohort study was conducted at 10 Canadian urban academic emergency
departments. Included in the study were alert, stable, adult patients with a Glasgow Coma Scale score of 15 and trauma to
the head or neck. This was a substudy of the Canadian C-Spine and CT Head Study. Eligible patients were prospectively
evaluated before radiography. Physicians estimated the probability of unstable cervical spine injury from 0% to 100%
according to clinical judgment alone and filled out a data form. Interobserver assessments were done when feasible. Patients
underwent cervical spine radiography or follow-up to determine clinically important cervical spine injuries. Analyses
included comparison of areas under the receiver operating characteristic (ROC) curve with 95% confidence intervals (CIs)
and the κ coefficient.
Results: During 18 months, 6,265 patients were enrolled. The mean age was 36.6 years (range 16 to 97 years), and 50.1%
were men. Sixty-four (1%) patients had a clinically important injury. The physicians' κ for a 0% predicted probability of
injury was 0.46 (95% CI 0.28 to 0.65). The respective areas under the ROC curve for predicting cervical spine injury were
0.85 (95% CI 0.80 to 0.89) for physician judgment and 0.91 (95% CI 0.89 to 0.92) for the Canadian C-Spine rule (P<.05).
With a threshold of 0% predicted probability of injury, the respective indices of accuracy for physicians and the Canadian
C-Spine rule were sensitivity 92.2% versus 100% (P<.001) and specificity 53.9% versus 44.0% (P<.001).
Conclusion: Interobserver agreement of unstructured clinical judgment for predicting clinically important cervical spine
injury is only fair, and the sensitivity is unacceptably low. The Canadian C-Spine rule was better at detecting clinically
important injuries with a sensitivity of 100%. Prospective validation has recently been completed and should permit
widespread use of the Canadian C-Spine rule.
[Ann Emerg Med. 2003;42:395-402.]

See related article, p. 403.

Introduction

More than 13 million patients with potential neck injuries are evaluated in US emergency departments (EDs) annually, of whom
approximately 30,000 have cervical spine injuries and 10,000 have spinal cord injuries. [1] [5] The results from 1 Canadian study
would suggest that 185,000 alert, stable adults with potential cervical spine injuries are treated in Canadian EDs annually, 0.9%
of whom have cervical spine fractures or dislocations. [6] Previous reports have suggested that clinical judgment is inadequate for
predicting cervical spine injuries. [7] [9] The American College of Surgeons Advanced Trauma Life Support course therefore
recommends that “… a c-spine film … be obtained on every patient sustaining an injury above the clavicle, and especially a
head injury.” [10] Previous reports have supported universal screening for cervical spine injuries, and approximately 97% of US
trauma centers have at one time routinely ordered cervical spine radiography for all trauma patients. [11] [13] A recent Canadian
study revealed institutional imaging rates ranging from 37% to 72% for alert, stable trauma patients at risk for cervical spine
injuries. [6]

Universal cervical spine radiography has yielded a positive test rate of less than 3% in most trauma series. [8] [12] [14] [25] We
estimate that 110,000 cervical spine radiography assessments are done each year in Canada on alert, stable, adult trauma
patients, of which 98% are normal. [6] The cost of inexpensive, high-volume tests may contribute more to rising health care costs
than more expensive “high technology” procedures. [26] [27] We estimate that more than CDN$30 million is spent annually in
Canada on outpatient cervical spine radiography. [28] Comprehensive radiography to clear the cervical spine requires a large
time commitment on the part of medical and hospital staff, distracting them from other responsibilities. [29] A recent survey
found that 98% of 300 Canadian emergency physicians would consider using a sensitive, reliable decision rule for the use of
cervical spine radiography, suggesting insight into the fallibility of current clinical judgment. [30] Such a rule would have the
potential to significantly reduce health care costs, improve efficiency, and improve patient care.

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 2/12

3/13/13 MD Consult - Print Previewer

The Canadian C-Spine and CT Head (CCC) Study is a prospective multiphase multicenter study. [31] [32] Phase I of the CCC

study involved derivation of a decision rule for cervical spine radiography by using 8,924 patients from 10 Canadian urban
teaching and community EDs. [33] Variables were assessed for their interobserver agreement and for their univariate association

with clinically important cervical spine injury. The final Canadian C-Spine rule, derived by multivariate recursive partitioning
of the strongest variables, was found to be 100% sensitive for detecting the 151 clinically important injuries (Figure 1).

Fig. 1. The Canadian C-Spine rule. MVC, motor vehicle crash.

To justify the time and effort involved in validating and disseminating a clinical decision rule, it is important to know whether
the rule improves on diagnostic accuracy of unstructured physician judgment alone. We sought to determine whether the
Canadian C-Spine rule would have performed better than physician judgment during Phase I of the CCC Study.

Methods

The multicenter, multiphase CCC Study was undertaken with reference to previously described guidelines for developing
clinical decision rules. [34] [35] This prospective observational cohort study was undertaken as part of phase I of the CCC Study.
It was carried out in 10 Canadian urban teaching and community EDs. Inclusion criteria were all ambulatory or immobilized
adult patients who (1) were hemodynamically stable (systolic blood pressure >90 mm Hg and respiratory rate between 10 and 24
breaths/min); (2) were alert (Glascow Coma Scale [GCS] score of 15); and (3) had either (a) neck pain from any mechanism of

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 3/12

3/13/13 MD Consult - Print Previewer

injury or (b) no neck pain but some visible injury above the clavicles, had not been ambulatory, and had experienced a
dangerous mechanism of injury. Exclusion criteria were age younger than 16 years, no neck pain or visible injury above the
clavicle, GCS score less than 15, unstable vital signs, time of injury more than 48 hours before assessment, penetrating trauma,
acute paralysis, known vertebral disease (eg, ankylosing spondylitis, rheumatoid arthritis, spinal stenosis, previous cervical
surgery), reassessment of the same injury, and pregnancy.

All attending staff physicians and emergency medicine residents were asked to carry out their normal assessment of each eligible
patient. After the clinical interaction, each physician completed a data form with standardized assessments of history and
physical findings. The form included the patient's hospital identification number, contact information, time and details of the
injury event, and potential decision rule clinical variables. Physicians were asked to prospectively estimate the probability that
the patient would have a clinically important cervical spine injury by circling 1 of the following: 0%, 1%, 2%, 3%, 4%, 5%,
10%, 20%, 30%, 40%, 50%, 75%, or 100%. Physicians were instructed to base this estimate on their judgment after considering
the facts obtained in the patient history and physical examination only, without assistance of a structured decision rule and
before radiographs, if any, were reviewed. When feasible, a second physician was asked to independently assess the patient and
fill out the study form to determine interobserver agreement. In cases in which interobserver data were obtained, each physician
was blinded to the findings and impressions of the other, and the forms were filled out prospectively before radiography was
undertaken.

Patients underwent plain radiography with or without flexion and extension views and computed tomography (CT) imaging at
the discretion of the treating physician after form completion. The decision to order films may not have been based solely on
physicians' stated judgment of likelihood of injury because physicians may adhere to guidelines instructing liberal use of
radiography or may agree with reports showing that judgment in these circumstances is poor. [7] [9] Patients who did not undergo

radiography participated in a structured telephone interview by a study nurse 14 days after discharge to determine patient
outcome and search for undetected injuries. This structured interview was previously validated on a random sample of 389
patients (including 66 with clinically important cervical spine injury) who had all undergone radiography. [33] Clinically

unimportant injuries were defined as isolated avulsion fracture of an osteophyte, isolated fracture of a transverse process not
involving body or facet joint, isolated fracture of a spinous process not involving the lamina, and isolated compression fracture
less than 25% of the vertebral body height. These definitions are based on a consensus opinion and strongly endorsed by a
formal survey of 129 Canadian academic emergency physicians, spinal surgeons, and neurosurgeons at 8 centers. [36]

Performance of unstructured physician judgment and the Canadian C-Spine rule were compared by using the areas under the
respective receiver operating characteristic (ROC) curves with 95% confidence intervals (CIs). Sensitivity and specificity of
unstructured physician judgment (for ≥0% probability) were also evaluated with 2 × 2 classification performance tables and
compared with those of the Canadian C-Spine rule (sensitivity 100% [95% CI 98% to 100%] and specificity 42.5% [95% CI
42% to 44%]). [33] The κ coefficient was calculated to measure the interobserver agreement on the subset of patients evaluated by

2 physicians. A reliable estimate was considered to have a κ higher then 0.6. Patients who could not be contacted or for whom
incomplete data were available were not included in the final analyses. Analyses were conducted with SAS (SAS Institute, Inc.,
Cary, NC) statistical software. The research ethics committees of the study hospitals approved the protocol without the need for
informed consent. Patients followed up had an opportunity to give verbal consent during the telephone interview by a study
nurse.

Results

This component of the CCC Study took place from October 1996 to April 1999. There were 6,265 patients enrolled by more than
200 physicians. Patient characteristics are summarized in Table 1.

Table 1. Characteristics of the 6,265 potential cervical spine injury patients.

Characteristic Representation in Study Population

Mean age, y (SD) 36.6 (16)

Range, y 16–97

Male sex, No. (%) 3,177 (50.6)

Mean time injury to assessment, h (SD) 3.5 (8.6)

Arrived by ambulance, No. (%) 3,282 (52.4)

Transfer from another institution, No. (%) 180 (2.9)

Ambulatory at any time, No. (%) 4,303 (68.7)

Upright position during examination, No. (%) 2,355 (37.6)

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 4/12

3/13/13 MD Consult - Print Previewer

Neck pain, No. (%) 5,744 (91.7)
Mechanism of injury, No. (%)
4,235 (67.6)
Motor vehicle crash 870 (13.9)
Fall 194 (3.1)
Pedestrian struck 213 (3.4)
Assault 209 (3.3)
Head struck by object 182 (2.9)
Sports 160 (2.6)
Bicycle 119 (1.9)
Axial load 48 (0.8)
Motorcycle 29 (0.5)
Motorized recreational vehicle
Other 6 (0.1)
Cervical spine radiography performed, No. (%) 4,344 (69.3)
CT scan performed, No. (%)
Cases followed up by telephone, No. (%) 236 (3.8)
Clinically important cervical spine injury, No. 1,956 (31.2)
(%)*
Clinically unimportant cervical spine injury, No. 64 (1.0)
(%)
Fracture, No. (%) 16 (0.3)
Dislocation, No. (%)
Ligamentous instability, No. (%) 76 (1.2)
Developed neurologic deficit, No. (%) 6 (0.1)
Stabilizing treatments, No. (%) 5 (0.1)
Internal fixation 5 (0.1)
Halo 69 (1.1)
Brace 14 (0.2)
Rigid collar 23 (0.4)
Admitted to hospital, No. (%) 5 (0.1)
27 (0.4)
* 437 (7.0)

* Some patients had >1 injury.

Assessments for 564 (9%) of the patients were carried out by resident physicians. Patients had been transferred from another
institution in 180 (3%) cases. The mean patient age was 36.5 years (range 16 to 97 years), and 50.6% were men. Sixty-four
patients (1.0%) had at least 1 clinically important cervical spine injury and 16 (0.3%) had clinically unimportant injuries.
Fractures occurred in 76 (1%) patients, dislocations in 6 (0.1%), and ligamentous injuries in 5 (0.1%). The most common
mechanism of injury was motor vehicle crash, followed by fall. Four hundred thirty-seven (7%) of all study patients were
admitted to the hospital, and 69 had stabilizing treatment (internal fixation, halo, rigid collar, or brace).

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 5/12

3/13/13 MD Consult - Print Previewer

The distribution of physician probabilities is given in Table 2.

Table 2. Distribution of physicians' predictions of injury and actual cervical spine injury.

Predicted Probability of Injury. Actual Injury?
%
No, No. (%) (N=6,213) Yes, No. (%) (N=64)

0 3,346 (53.8) 5 (7.8)

1 1,822 (29.3) 10 (15.6)

2 481 (7.7) 13 (20.3)

3 182 (2.9) 6 (9.4)

4 33 (0.5) 1 (1.6)

5 225 (3.6) 10 (15.6)

10 67 (1.1) 7 (10.9)

20 27 (0.4) 7 (10.9)

30 14 (0.2) 0

40 3 (<0.1) 0

50 13 (0.2) 4 (6.3)

75 0 1 (1.6)

100 0

The respective areas under the ROC curve for predicting cervical spine injury were physician judgment 0.85 (95% CI 0.80 to
0.89) and Canadian C-Spine rule 0.91 (95% CI 0.89 to 0.92; P<.05; Figure 2.

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 6/12

3/13/13 MD Consult - Print Previewer

Fig. 2. ROC curve for predicting cervical spine injury.

Physicians' stated judgment had a sensitivity of 92% (95% CI 82% to 96%) for predicting a probability of clinically important
fracture of greater than 0% and would have missed 5 cases (). For 89 cases, the interobserver κ for predicting a 0% probability of
important cervical spine injury according to physician judgment was only 0.46 (95% CI 0.28 to 0.65). For comparison purposes,
the Canadian C-Spine rule had a sensitivity of 100% (95% CI 94% to 100%) in 6,265 patients (Table 3).

Table 3. Classification of performance of physician judgment to predict at least 0% probability of clinically
important cervical spine injury.*

Cervical Spine Injury

Physician Judgment Yes No

Yes 59 2,864

No 5 3,337

*

* Sensitivity 92.2% (82%–96%); specificity 53.8% (53%–55%); κ 0.46 (0.28–0.65).

The specificities of physician judgment and the Canadian C-Spine rule for clinically important injury are 53.9% (95% CI 53% to
55%) and 44.0% (95% CI 43% to 45%), respectively (Tables 3 and 4).

Table 4. Classification of performance of the Canadian C-Spine rule for clinically important cervical spine
injury.*

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 7/12

3/13/13 MD Consult - Print Previewer

Clinically Important Injury

Decision Rule Yes No

Yes 64 3,475

No 0 2,726

*

*Sensitivity 100% (95% CI 94%–100%); specificity 44.0% (95% CI 43%–45%).

Discussion

Overall, physician judgment based on unstructured assessment alone is relatively sensitive for predicting zero risk of cervical
spine injury in adult blunt trauma patients. In this study, however, physicians' unstructured judgments did miss a small number
of important injuries. The recently validated Canadian C-Spine rule performed better than unstructured physician judgment in
predicting the absence of important cervical spine injury. The Canadian C-Spine rule has the potential to safely reduce the
number of radiographs ordered and should be followed in the treatment of adults sustaining blunt trauma to the head or neck.

The suboptimal sensitivity of physician judgment alone for detecting important cervical spine injuries, combined with the
potential for profound adverse outcomes when cervical spine injuries are missed, means that many patients with a low perceived
likelihood of injury will still undergo radiographic investigation. [1] [7] [10] [11] [13] [21] [37] This study would support this
practice because physicians estimated the probability of having an important injury at 0% for 5 patients who were eventually
found to have such an injury. Most Canadian and US emergency physicians disagree, however, with universal cervical spine
imaging advocated by some guidelines. [30] The high patient volume, brief patient-physician contact, uncertain follow-up, and
medicolegal risk associated with emergency medicine practice coupled with the lack of validated decision rules in this area have
historically left little alternative to the reluctant universal application of radiography. [31] [38] [40] Our study suggests that
physicians actually order more cervical spine radiographs than they would if they relied on their clinical judgment alone,
suggesting that physicians do not rely solely on their judgment. Physicians' current practice is sensitive at the expense of
specificity because of this overuse. We chose to compare the Canadian C-Spine rule to physicians' “likelihood of injury”
estimate rather than to their current ordering practice for precisely this reason. We believed it important to compare the Canadian
C-Spine rule with what physician judgment would offer if judgment alone were relied on to limit use.

In the derivation phase, the Canadian C-Spine rule for cervical spine radiography was 100% sensitive for detecting 151
clinically important injuries. [33] The current study has shown that the Canadian C-Spine rule is more accurate than physician
judgment. The derivation phase of the Canadian C-Spine rule would suggest that cervical spine radiography use rates by using
the rule would be approximately 60%. [33] According to this result, physicians using the Canadian C-Spine rule may confidently
reduce their cervical spine radiography rate significantly while avoiding appreciable risk of missing important cervical spine
injuries. Physicians did not order radiographs on 1,981 patients in this study, whereas the Canadian C-Spine rule would have
correctly predicted that an additional 745 patients would not have required imaging, a reduction of 17.5%. Avoiding
radiography on some patients with blunt neck trauma probably will shorten the time to “clearing” the cervical spine, which
would in turn reduce length of stay, lessen the burden of care for health care staff, facilitate mobilization of the patients for
investigation and treatment, and help ease the increasing congestion in many large EDs. The rule has recently undergone
prospective validation. [41] The US-based National Emergency X-Radiography Utilization Study (NEXUS) criteria were recently
found to have a sensitivity of 99.0% and specificity of 12.9% in a large validation study. [25] The relatively low specificity
means that adherence to this rule may fail to decrease ordering rates among physicians outside of the United States and in some
cases may actually increase use. A retrospective application of the NEXUS criteria to a CCC Study population revealed a
sensitivity of only 93%. [42] One of the reasons for this outcome may be the low interobserver reliability associated with certain
criteria in the US study. Our CCC Study data would suggest, for example, that such aspects as intoxication and distracting injury
are poorly defined and hard to assess, with interobserver reliabilities of 0.22 and 0.41, respectively. [33] Wide application of the
NEXUS rule should be preceded by further investigation of its sensitivity and interobserver reliability in a broad population.

Agreement of overall clinical impression according to unstructured physician assessment for risk of injury was poor. The κ value
measures agreement beyond that which would be expected by chance alone. In this study, the κ for estimation of the patient's
overall risk of important injury was only 0.46, which is not the case, however, for agreement on specific individual patient
variables. Individual components of the Canadian C-Spine rule were incorporated into the final decision rule only after they

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 8/12

3/13/13 MD Consult - Print Previewer

were found to have high interobserver agreement defined a priori as a κ value higher than 0.6. [33]

There are several potential limitations to this study. Some individuals may be concerned about the interpretation of “clinically
unimportant cervical spine injury.” We have, however, demonstrated good acceptance of our definition by Canadian academic
neurosurgeons, spine surgeons, and emergency physicians. The priority of diagnostic imaging for these trauma patients should
be to identify cervical spine injuries that require treatment and follow-up. “Clinically unimportant cervical spine injuries,”
according to the academic surgeons in our survey, require neither stabilizing treatment nor specialized follow-up and are
unlikely to be associated with long-term problems. For patients who do not require urgent specific treatment for neck injuries, we
are aware of no literature reports suggesting that emergency radiography is likely to predict the rare long-term complication,
such as a chronic pain.

Not all physicians in the study may have been aware of the explicit definitions used for clinically unimportant injury, nor were
they surveyed for individual understanding of the definitions used. Their predicted probability of clinically important injury
may not be based on the study definitions. We believe, however, that this difference would accurately reflect the reality that
individual physicians would have their own concept and threshold for unimportant injuries.

Some readers may be concerned that not all study patients underwent cervical spine radiography. We could not ethically insist
on universal radiography for all patients, because many Canadian physicians withhold radiography for patients who they feel
have a low risk of injury. Unimaged patients were only classified as having no clinically important injury if they satisfied all
criteria on the structured, validated, 14-day, telephone proxy outcome tool administered by a registered nurse.

Physicians were asked to estimate their predicted likelihood of clinically significant injury for each patient at study form
completion. This estimate may have knowingly or unknowingly been influenced by the completion of the form. The study
required only that physicians document their findings for each element of the rule, much as they would do on a patient medical
record, and did not involve application of the rule. The Canadian C-Spine rule is highly sensitive and specific, is based on
standard elements obtained during the initial patient interaction, and has high interobserver reliability. For these reasons, we
believe that any influence completion of the rule may have had on physicians' estimates would have narrowed the gap between
their unassisted judgment and the performance of the rule, meaning that eliminating this measurement bias would make the
improvement seen with the rule over unstructured judgment even more pronounced. Similarly, patients transferred for assessment
may have had previous cervical spine radiographs. Physicians were asked to complete the study forms before reviewing any
radiographs, and no breaches of protocol were reported. Any previous physician knowledge of radiography results in this small
percentage of patients (3%) would have falsely improved physician judgment and resulted in a conservative estimate of the
superiority of the Canadian C-Spine rule.

We acknowledge that not all eligible patients were enrolled in the study, which is not unusual for a clinical study. The
characteristics of patients not enrolled were similar to those of the patients who were enrolled.

The limitations of physicians' unstructured judgments have resulted in widespread universal use of cervical spine radiography in
stable, alert, adult trauma patients. In this substudy of the CCC Study, we have shown that the Canadian C-Spine rule for cervical
spine radiography is more accurate than physician judgment alone. We believe that use of a prospectively derived and validated,
reliable, and practical decision rule for the use of cervical spine radiography in these patients will reduce radiography rates
without compromising patient care. Prospective validation of the Canadian C-Spine rule has recently been completed. Further
study is needed to determine whether practice patterns can change with its implementation.

We thank the following for their assistance: Erica Battram, BA, Kim Bradbury, RN, Teresa Cacciotti, RN, Pamela Sheehan, RN,
Taryn MacKenzie, RN, Kathy Bowes, RN, Karen Code, RN, Virginia Blak-Genoway, RN, Debbie Karsh, RN, Sharon Mason, RN,
Percy MacKerricher, RN, Jan Buchanan, BSN (study nurses); My-Linh Tran and Emily Moen (data management); Irene Harris
(manuscript preparation); and all the physicians, nurses, and clerks at the study sites who voluntarily and patiently assisted with
case identification and data collection.

Author contributions:

IGS and GAW conceived the study, designed the trial, and obtained research funding. IGS, GAW, and KLV supervised the
conduct of the trial and data collection. GHG, RB, GB, DC, JD, MAE, IM, RDM, LM, MR, MS, JW, KLV, and CC undertook
recruitment of participating centers and patients and managed the data, including quality control. IGS, GAW, and VDM provided
statistical advice on study design and analyzed the data. HL provided clinical consultation. IGS chaired the data oversight
committee. GB drafted the manuscript, and all authors contributed substantially to its revision. IGS takes responsibility for the
paper as a whole.

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/4… 9/12

3/13/13 MD Consult - Print Previewer

References

1. Reid DC, Henderson R, Saboe L, et al. Etiology and clinical course of missed spine fractures. J Trauma. 1987;27:980-986. Abstract

2. McCaig L. National Hospital Ambulatory Medical Care Survey: 1992 emergency department summary. Adv Data. 1994;245:1-12.

3. Diliberti T, Lindsey RW. Evaluation of the cervical spine in the emergency setting: who does not need an X-ray? Orthopedics. 1992;15:179-183.

Citation

4. Bachulis BL, Long WB, Hynes GD, et al. Clinical indications for cervical spine radiographs in the traumatized patient. Am J Surg. 1987;153:473-

478. Abstract

5. McCaig LF. National Hospital Ambulatory Medical Care Survey: 2000 emergency department summary. Advance Data From Vital and Health
Statistics. Hyattsville, MD: National Center for Health Statistics; 2001.

6. Stiell IG, Vandemheen K, Laupacis A, et al. Variation in emergency department use of cervical spine radiography for alert, stable trauma patients.
Can Med Assoc J. 1997;156:1537-1544.

7. McKee TR, Tinkoff G, Rhodes M. Asymptomatic occult cervical spine fracture: case report and review of the literature. J Trauma. 1990;30:623-

626. Abstract

8. Roberge RJ, Wears RC, Kelly M, et al. Selective application of cervical spine radiography in alert victims of blunt trauma: a prospective study. J

Trauma. 1988;28:784-788. Abstract

9. Jacobs LM, Schwartz R. Prospective analysis of acute cervical spine injury: a methodology to predict injury. Ann Emerg Med. 1986;15:44-49.

Abstract

10. American College of Surgeons. Advance Trauma Life Support Instructor Manual. 5th ed. Chicago, IL: American College of Surgeons; 1993.

11. Mirvis SE, Diaconis JN, Chirico PA, et al. Protocol-driven radiologic evaluation of suspected cervical spine injury: efficacy study. Radiology.

1989;170(3, pt 1):831-834. Abstract

12. Davis JW, Phreaner DL, Hoyt DB, et al. The etiology of missed cervical spine injuries. J Trauma. 1993;34:342-346. Abstract

13. Woodring JH, Lee C. Limitations of cervical radiography in the evaluation of acute cervical trauma. J Trauma. 1993;34:32-39. Abstract

14. McNamara RM, Heine E, Esposito B. Cervical spine injury and radiography in alert, high-risk patients. J Emerg Med. 1990;8:177-182. Abstract

15. Fischer RP. Cervical radiographic evaluation of alert patients following blunt trauma. Ann Emerg Med. 1984;13:905-907. Abstract

16. Gbaanador GB, Fruin AH, Taylon C. Role of routine emergency cervical radiography in head trauma. Am J Surg. 1986;152:643-648. Abstract

17. Bayless P, Ray VG. Incidence of cervical spine injuries in association with blunt head trauma. Am J Emerg Med. 1989;7:139-142. Abstract

18. Neifeld GL, Keene JG, Hevesy G, et al. Cervical injury in head trauma. J Emerg Med. 1988;6:203-207. Abstract

19. Cadoux CG, White JD, Hedberg MC. High-yield roentgenographic criteria for cervical spine injuries. Ann Emerg Med. 1987;16:738-742.

Abstract

20. McNamara RM, O'Brien MC, Davidheiser S. Post-traumatic neck pain: a prospective and follow-up study. Ann Emerg Med. 1988;17:906-911.

Abstract

21. Kreipke DL, Gillespie KR, McCarthy MC, et al. Reliability of indications for cervical spine films in trauma patients. J Trauma. 1989;29:1438-1439.

Abstract

22. Hoffman JR, Schriger DL, Mower W, et al. Low-risk criteria for cervical-spine radiography in blunt trauma: a prospective study. Ann Emerg Med.

1992;21:1454-1460. Abstract

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/… 10/12

3/13/13 MD Consult - Print Previewer

23. Hills MW, Deane SA. Head injury and facial injury: is there an increased risk of cervical spine injury? J Trauma. 1993;34:549-553. Abstract

24. Lindsey RW, Doherty BJ, Watson AB. Efficacy of radiographic evaluation of the cervical spine in emergency situations. South Med J.

1993;86:1253-1255. Abstract

25. 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;343:94-99. Abstract

26. Moloney TW, Rogers DE. Medical technology: a different view of the contentious debate over costs. N Engl J Med. 1979;301:1413-1419.

Abstract

27. Angell M. Cost containment and the physician. JAMA. 1985;254:1203-1207. Abstract

28. Ontario Ministry of Health. The Ontario Statistical Reporting System 1993-94. Toronto, Ontario, Canada: Ontario Ministry of Health; 1994.

29. Vandemark RM. Radiology of the cervical spine in trauma patients: practice pitfalls and recommendations for improving efficiency and
communication. Am J Roentgenol. 1990;155:465-472.

30. Graham ID, Stiell IG, Laupacis A, et al. Emergency physicians' views on the use of computed tomography and cervical spine radiography. J Int
Soc Assessment Technol Health Care. 1998;14:31.

31. Stiell IG, Wells GA, McKnight RD, et al. The Canadian C-Spine rule for alert and stable trauma patients, Part I: background and rationale. Can J
Emerg Med. 2002;4:84-90.

32. Stiell IG, Wells GA, Vandemheen K, et al. The Canadian CT Head rule for patients with minor head injury. Lancet. 2001;357:1391-1396.

Abstract

33. Stiell IG, Wells GA, Vandemheen K, et al. The Canadian Cervical Spine Radiography rule for alert and stable trauma patients. JAMA.

2001;286:1841-1848. Abstract

34. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules: a review and suggested modifications of methodological standards. JAMA. 1997;277:488-

494. Abstract

35. Stiell IG, Wells GA. Methodologic standards for the development of clinical decision rules in emergency medicine. Ann Emerg Med. 1999;33:437-

447. Full Text

36. Stiell IG, Lesiuk H, Vandemheen K, et al. Obtaining consensus for a definition of “clinically important cervical spine injury” in the CC Study. Acad
Emerg Med. 1999;6:43-45.

37. Changaris D. Cervical spine films, cost, and algorithms. Am J Surg. 1987;153:478. Abstract

38. Long AE. Radiographic decision-making by the emergency physician. Emerg Med Clin North Am. 1985;3:437-446. Abstract

39. Feinstein AR. The “chagrin factor” and qualitative decision analysis. Arch Intern Med. 1985;145:1257-1259. Citation

40. Stiell IG, McDowell I, Nair RC, et al. Use of radiography in acute ankle injuries: physicians' attitudes and practice. CMAJ. 1992;147:1671-1678.

Abstract

41. Stiell IG, Clement C, Wells GA, et al. Multicentre prospective validation of the Canadian C-Spine rule [abstract]. Acad Emerg Med. 2002;9:359.

42. Stiell IG, McKnight R, Wells GA, et al. Application of the Nexus Low-Risk Criteria for cervical spine radiography in Canadian emergency
departments [abstract]. Acad Emerg Med. 2000;7:566.

Copyright © 2013 Elsevier Inc. All rights reserved. - www.mdconsult.com

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/… 11/12

3/13/13 MD Consult - Print Previewer

Bookmark URL: /das/journal/view/0/N/14044602?ja=386466&PAGE=1.html&issn=0196-0644&source=

Client IP Address: 146.189.194.69
DAS Host: isis

www.mdconsult.com.ezproxy.umassmed.edu/website/view/preview?printing=true&URL=http%3A//www.mdconsult.com.ezproxy.umassmed.edu/das/article/body/… 12/12