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

Specific questions in this review are based on Jaeschke RZ, Meade MO, Guyatt GH, Keenan SP, Cook DJ. How to use diagnostic test articles in the intensive care unit: diagnosing weanability using f/Vt. Crit Care Med. 1997;25(9):1514-2. [abstract]

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Validity of a Set of Clinical Criteria to Rule Out Injury to the Cervical Spine in Patients with Blunt Trauma.

Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI.

N Engl J Med. 2000 Jul 13;343(2):94-99. [abstract]

Reviewed by Al Torres, MD, MS, University of Illinois College of Medicine at Peoria

Review posted November 22, 2001

I. What is being studied?

Study objective:

To determine whether it is safe not to obtain cervical films if a blunt trauma patient has none of these 5 criteria:

1. the absence of tenderness at the posterior midline of the cervical spine
2. the absence of a focal neurologic deficit
3. a normal level of alertness
4. no evidence of intoxication
5. absence of clinically apparent pain that might distract the patient from the pain of a cervical-spine injury

Study design:

A prospective, observational study using 3 standard views of the cervical spine (cross-table lateral, anteroposterior, and open-mouth of the odontoid) as the gold standard for validation.

The patients investigated:

All patients with blunt trauma who underwent radiography of the cervical spine at one of the 21 participating US medical centers. Only one of the participating centers was a pediatric facility, which accounted for only 2.5% of the subjects being younger than 8 years of age.

The outcomes measured:

Radiographic cervical spine injury, deemed clinically significant versus clinically insignificant. The following list includes cervical spine injuries deemed clinically insignificant by the authors:

• Spinous process fracture
• Simple wedge-compression fracture without loss of 25% or more of vertebral body height
• Isolated avulsion without ligamentous injury
• Type I (Anderson-D'Alonzo) odontoid fracture
• End-plate fracture
• Osteophyte fracture, not including corner fracture or teardrop fracture
• Injury to trabecular bone
• Transverse-process fracture

It is debatable whether or not these injuries are not clinically significant in young children or infants.

II. Are the results of the study valid?

Primary questions:

1. Was there an independent, blind comparison with a reference standard?

Yes. The clinical criteria were assessed prior to obtaining the cervical spine radiographs and the radiologists interpreted the radiographs without knowledge of the criteria findings. If the radiographs were obtained prior to assessing the criteria, the patient was considered "unstable", which was equivalent to having a clinically significant injury, i.e., positive for one of the criteria.

2. Did the patient sample include an appropriate spectrum of patients to whom the diagnostic test will be applied in clinical practice?

Yes. There were 34,069 patients enrolled and 818 (2.4%) of those enrolled had radiographic findings consistent with a cervical spine injury. Only 12.6% of the total number studied, however met their criteria for low risk (none of the 5 criteria). Fifty-nine percent were male and the mean age for the entire sample was 37 years, with a range of 1 to 101 years. Since a small portion of the entire sample was less than 8 years old (2.5%), these criteria can only be applied to children older than this.

Secondary questions:

3. Did the results of the test being evaluated influence the decision to perform the reference standard?

Maybe. The authors acknowledged that some clinicians were probably already using similar criteria to screen blunt trauma patients to order cervical radiographs. However, the authors did check neurosurgery records and quality assurance logs three months after the completion of the study and no injuries were found to have been missed. Therefore, if this study is biased, they say it may be because clinicians were already avoiding cervical films based on their previous study results. We don't know the incidence of significant injury in the population of patients who did not get cervical films.

4. Were the methods for performing the test described in sufficient detail to permit replication?:

No, but the authors stated the guidelines available to the clinicians on how to use the criteria to exclude patients are available on request. The methods for performing the test were only that anyone with any history of blunt trauma gets neck films. They don't differentiate if some institutions already weren't getting neck films if they did not fulfill any of the 5 criteria.

III. What are the results?

1. Are the test's sensitivity, specificity, and likelihood ratios presented (or are the data necessary for their calculation provided)?

Yes. The following table contains the data presented.

 

Outcome with Clinical Criteria*	Radiographically Documented Injury
	Yes	No
Screening for any injury
Positive	810	28,950
Negative	8	4,301
Screening for clinically significant injury
Positive	576	29,184
Negative	2	4,307

 

 

*"Positive" indicates that the patient was considered to require imaging to detect a possible cervical-spine injury, and "negative" that the probability of such injury was considered so low that imaging was unnecessary.

The likelihood ratio (LR) a blunt trauma patient would have a cervical spine injury detected by radiographs if one or more of the screening criteria were present was 1.14 (95% CI, 1.13, 1.15). The LR a blunt trauma patient would have a clinically significant cervical spine injury given one or more of the criteria were present would be similar (1.14; 95% CI, 1.14, 1.15). Remember that a LR close to one does not significantly change the patient's post-test probability of having a disease/injury given a "positive" test result.

On the other hand, diagnostic criteria with high sensitivities are effective in ruling out the presence of a disorder. For example, the sensitivity of the criteria in "detecting" the lack of a cervical spine injury was 99%. The LR of finding radiographic evidence of a cervical spine injury if all the criteria are absent was 0.075 (95% CI, 0.038 to 0.15). The LR for a clinically significant cervical spine injury if all the criteria are absent was 0.03 (95% CI, 0.006, 0.11). Remember that a LR > 0.1 causes a significant decrease in the patient's post-test probability of having a disease/injury given a "negative" test result.

Given that we know the prevalence of injuries Ð 2.4% - with the negative LR we can calculate the post-test likelihood given a negative test Ð it is 0.2%. (The online calculator at http://www.healthcare.ubc.ca/calc/bayes.html is great for these calculations!)

Of the 8 patients who satisfied none of the five criteria but had radiographically documented cervical spine injury, two of these patients had injuries deemed clinically significant. One was an asymptomatic 54 year old man with a fracture of the anteroinferior portion of the second cervical vertebra. The other was a 57 year old man described as having pain in the right shoulder and tenderness of the paraspinous muscles and in the area of the right clavicle and scapula who suffered a fracture of the right lamina of the sixth cervical vertebra and right clavicle. It is unclear why this patient was negative for the criterion of "clinically apparent pain that might distract the patient from the pain of cervical-spine injury."

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?

Yes, but only if pediatric trauma specialists agree with the bone injuries categorized as clinically insignificant by the authors and if the guidelines used by the investigators are obtained and utilized. The likelihood that these criteria will reduce the total number of cervical spine radiographs obtained outside of pediatric trauma centers is unlikely due to the lack of comfort these healthcare professionals often have in dealing with injured children. Many pediatric trauma specialists already apply similar criteria and these data support this clinical approach. Despite the low incidence of injury and the LR being close to 1 for any of the 5 criteria being positive, the consequences of missing a clinically significant cervical spine injury in one patient may be catastrophic.

2. Are the results applicable to my patients?

Yes, if the patients are older than 8 years and can have the criteria assessed adequately. Spinal cord injury without radiographic abnormality occurs more frequently in children and needs to be taken into consideration.

3. Will the results change my management?

Yes. These criteria have held up to rigorous investigation and are probably useful in the older school-aged child and certainly the adolescent. In an older child or adolescent behaving appropriately in the trauma room and cooperating during the physical examination, it is probably not necessary to obtain radiographs of the cervical spine

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

Although cost was not discussed, reducing the number of radiographs received will cost less. There will be an associated minute decrease in radiation exposure. What are the costs of missing just one significant injury? The paper argues convincingly to justify these potentially small benefits. Even in the eight patients missed, none had a clinically significant bad outcome as a result.

 

Comments

Back to the EB Journal Club Index