Plain film radiography is readily available in all emergency centers; it is a reliable and quick method that can be performed with portable and fixed equipment. Plain radiography provides a very good overview of the extent and magnitude of injury and can make a definitive and specific diagnosis in certain spinal injuries. Plain radiography has been considered the standard initial “screening” examination used to evaluate patients with suspected spine trauma. In fact, plain film radiography remains the mainstay for evaluation of trauma patients worldwide who do not have ready access to CT.

Cross-table lateral radiographs have been used for years for the initial evaluation of the cervical spine trauma. However, the generally accepted opinion is that lateral radiographs are inadequate to exclude cervical spine injury because of the high number of false positives and false negatives, and because the reported predictive value of a study with negative findings is not sufficient for the study to be used as the only screening examination (12). An additional limitation is the frequent incomplete visualization of the cervicothoracic and cranio-cervical junctions in unconscious and uncooperative multiple-trauma victims, which can result in significant delay. Repeated exposures are frequently necessary before adequate radiographs are obtained in this subset of obtunded and uncooperative patients.

Overall, plain radiography remains accepted as the technique of choice for initial evaluation of an injured cervical spine. This practice is supported by the fact that if the results of an adequately exposed and properly positioned radiographic series of the spine are normal, then it is unlikely that CT will reveal a fracture. However, there is sufficient evidence that a significant number of fractures can be missed if the evaluation of the cervical spine relies exclusively on plain radiography. These claims are particularly valid for the subset of patients who meet multiple trauma criteria, for whom plain radiography has a limited value. Woodring and Lee (13) reported the limitations of plain radiography in detecting fractures, using a retrospective review of radiographs and CT scans in 216 patients with cervical spine fractures. They determined prospectively that plain radiography did not detect fractures in 23% of their patients and that the cervical spine injuries were unstable in 50% of the cases. Our own experience indicates that plain radiography can miss up to 57% of fractures (2, 3). In a series by Acheson et al. (14), only 47% of fractures were seen or suspected on initial screening radiographs when compared with those ultimately detected by CT.

Supine oblique views are no longer necessary in patients who are undergoing cervical CT examination. Oblique views, although useful in patients with unilateral locked facet, are most valuable in adding two more views of the cervicothoracic junction in patients with equivocal lateral or swimmer’s views that are not undergoing CT examination (low risk, obese short-necked patients). Both of these functions, however, can now be accomplished with the use of CT (1).

Flexion-extension radiographs are not very helpful in the acute setting because muscle spasm in acutely injured patients precludes an adequate examination. Insko et al. (15), in a review of 106 consecutive cases of blunt trauma evaluated with flexion and extension radiographs of the cervical spine obtained in the acute setting at a level-one trauma center, reported that when adequate motion was present on flexion and extension radiographs, the false-negative rate was zero. However, in the acute setting, 30% of the examinations were limited by inadequate motion. A higher percentage of injury (12.5%) was detected by subsequent cross-sectional imaging in these patients. Limited flexion and extension motion on physical examination should preclude the use of flexion and extension radiographs, as they are of limited diagnostic utility. Cross-sectional imaging may be warranted in this high-risk group of patients (15). Flexion-extension radiographs are best reserved for follow-up of symptomatic patients with suspected ligamentous instability, usually 7 to 10 days after muscle spasm has subsided; in these patients, however,
magnetic resonance imagery (MRI) is the procedure of choice. Flexion-extension radiographs are also helpful for ensuring that minor degrees of anterolisthesis or retrolisthesis in patients with cervical spondylosis are fixed deformities (1, 16, 17).

Some authors have suggested the use of passive flexion and extension imaging under fluoroscopic guidance for patients in whom one cannot obtain a reliable physical examination (5, 18, 19 and 20). Almost all patients studied in this manner have been normal, as would be expected from the extremely low pretest probability of unstable injury. The limited data available do not provide sufficient evidence to support routine use of this technique. A number of cervical injuries will not be detected by this method, including herniated intervertebral discs and extradural hemorrhage. These lesions may cause spinal cord compression that creates or worsens a neurologic deficit without evidence of overt subluxation on fluoroscopy. Davis et al. (21) performed fluoroscopic examinations on 301 patients. There were 297 true-negative examinations, two true-positive examinations (stable injuries), one false-negative examination, and one false-positive examination. The incidence of ligamentous injury identified by fluoroscopy in this study was two of 301 (0.7%). Unstable cervical spine ligamentous injuries were identified in only 0.02% of all trauma patients. Complications of this procedure have also been reported with one patient developing quadriplegia when fluoroscopic evaluation was performed (21).

The introduction of helical scanning has expanded the clinical applications of CT. The advantage of helical CT and MDCT over single-section CT is faster acquisition of a volumetric data set, which results in shorter examination time. Motion and misregistration artifacts are minimized, and high-quality reconstructed images can be obtained. In many instances, additional scanning after the initial acquisition is avoided by the ability to retrospectively generate overlapping interscan images using original data. These advantages are particularly relevant in acutely ill and multiple-trauma victims who require rapid and accurate imaging assessment. CT may reveal more fractures than plain films and may allow evaluation of the cervicothoracic and cranio-cervical junctions, areas traditionally poorly visualized on plain films.

With the development of MDCT, an increased tube output, and a reduction of the rotation time, it became possible to cover larger scan ranges in z-direction with a collimation of 0.5 to –4 mm and a reasonable acquisition time (22). As compared to single detector helical CT, MDCT results in improved image quality, such as decreased stair-step artifacts, when generating images in planes different from the one of data acquisition (multiplanar reformations [MPRs]) (23). The major improvement for generating MPRs, however, is the ability to achieve near isotropic voxels, as predicted by Kalender (24) in 1995. Isotropic imaging permits calculation of MPRs with an image quality close to that of the transverse images calculated from the raw data set (25, 26 and 27). On the other hand, these advances have lead to what Rubin (28) terms a data explosion. Processing this vast amount of data is now the challenge of multidetector-row CT technology (28, 29). To reduce the large number of images, Begemann et al. (29) retrospectively assessed the value of coronal and sagittal reformations alone in the diagnosis of acute vertebral fractures and evaluated whether it still remains necessary to examine the axial images. In 244 vertebral bodies, 70/70 fractured vertebrae were diagnosed on reviewing MPRs alone. There were no false positive cases. In 2/70 fractures, the anatomically exact diagnosis was complemented by examination of the axial images. Forty-two of 43 unstable fractures were correctly diagnosed on MPRs alone. With preferential MPR reading, the total number of images to be analyzed was significantly reduced (P < 0.01). Their preliminary results indicate that examination of the MPRs alone is a feasible approach for assessment of vertebral fractures that can also accurately classify them as either stable and unstable (provided the MPRs are done properly). The transverse images must be read complementarily in cases of complex fractures, in osteopenic patients, in cases of CT scans with artifacts, or if any questions arise about the exact type of the fracture.

Horizontal fractures that are oriented in the plane of the scan, such as transverse odontoid fractures, may not always be demonstrated by CT. However, helical and MDCT have been useful in overcoming this diagnostic problem secondary to overlapping imaging of contiguous segments. Furthermore, these injuries are more likely to be demonstrated in sagittal and coronal reformations.

What is the cost of CT versus plain radiography? Cost should not be based on charges but should be based on the actual cost to operate the equipment per hour (10). This includes the technologist’s time, film cost, time required for the study, and, most importantly, the accuracy of diagnosis (5, 30, 31 and 32). When these parameters are used, CT has been shown to be more cost effective than plain radiography (5). There is one additional “cost” to be considered with using CT: radiation exposure. It is generally acknowledged that helical CT, and especially multidetector CT examinations, carries a higher radiation dose than plain radiography (33). However, work is currently under way to determine methods of decreasing this radiation exposure, primary via lowering the milliamperage setting needed for diagnostic quality examinations.