97. The utility of the AP view of cervical spine radiography in nontraumatic cervical spinal disorders

Abstract

<h3>BACKGROUND CONTEXT</h3> While ordering the two-view cervical spine X-rays is a routine in evaluating patients with nontraumatic cervical spinal disorders, the anteroposterior (AP) view seems to provide relatively less information than the lateral view. Considering the radiation and cost of the examination, and given that most surgical patients would already have CT and/or MRI available, it is reasonable to examine the utility of the AP view in cervical spine radiography in nontraumatic conditions. <h3>PURPOSE</h3> This study aims to (a) examine how often the radiological reports of the cervical spine describe findings from an AP view and (b) survey senior spine surgeons nationwide regarding their utilization of the AP view in actual practice. <h3>STUDY DESIGN/SETTING</h3> Review of radiological reports using the technique of text-mining, followed by a survey of surgeons. <h3>PATIENT SAMPLE</h3> A total of 210 patients diagnosed with nontraumatic cervical spinal disorders at our hospital, and 24 senior spine surgeons nationwide. <h3>OUTCOME MEASURES</h3> Percentages of the radiological aspects of the cervical spine that are described in the radiological reports by view (AP vs lateral vs both views), percentages of the surgeons' responses. <h3>METHODS</h3> This study included two parts: a review of radiological reports and a survey of surgeons. In Part 1, radiological reports of the two-view cervical spine radiographs for patients diagnosed with nontraumatic cervical spinal disorders at a tertiary spine center were collected, excluding those with tumor or previous cervical spinal surgery. After the hospital's medical informatics service used programming algorithms to locate the first such radiography of each patient, the radiological reports were reviewed in two steps. Step 1 used the technique of text-mining to obtain 50 most frequently used words in the radiological reports so as to summarize the main radiological aspects of a cervical spine (eg, alignment, disc space, uncovertebral joint, etc.) Then, step 2 identified the source of each radiological aspect's description in the radiological report: Was it from the AP view, the lateral view, or both? For example, a sentence in the radiological report read, "Frontal view demonstrates no significant uncovertebral hypertrophy," and this indicates that this radiological aspect -the uncovertebral joint- was evaluated using the AP view. As another example, a report noted "disc space narrowing at C5-6 with prominent osteophyte." As disc space narrowing is inherently a lateral-view sign, this indicates that both the disc space and the osteophyte here in this report were described using the lateral view. In this way, each radiological aspect of the cervical spine was examined, recorded, and tallied in all of the radiological reports. Then in Part 2, a two-question survey asking surgeons whether they think the AP view is practically useful and why was distributed among a group of 37 senior spine surgeons nationwide, who were from a published survey study on a different topic. <h3>RESULTS</h3> A total of 210 radiological reports from 2006 and 2020 were eligible for inclusion. Together, they described 18 radiological aspects of the cervical spine. Of these 18 radiological aspects, 28% (5/18, eg, uncovertebral joint, transverse process) were described exclusively on the AP view, 50% (9/18, eg, lordosis, disc space) were exclusively on the lateral view, and 22% (4/18, eg, fracture, degenerative changes in general) could be from both views. These 18 radiological aspects varied in their frequency of presence in the radiological reports, ranging from disc degeneration (82%, 172/210) to cervical scoliosis (0.5%, 1/210). The AP view was used to describe uncovertebral joint (24%, 51/210), transverse process (3%, 6/210), lung apices (1%, 2/210), and some other signs. Three radiological aspects could be from both the AP and lateral views: alignment in general (81%, 171/210), fracture (53%, 112/210) and degenerative changes in general (26%, 54/210). The survey of surgeons obtained a response rate of 65% (24/37). Among the respondents, 63% (15/24) thought that the AP view is useful, and the reasons included showing deformity or alignment (47%, 7/15), showing facet osteoarthritis (27%, 4/15), showing clavicle or rib for spinal level localization (20%, 3/15) and showing the uncovertebral joints (20%, 3/15), among others. <h3>CONCLUSIONS</h3> For nontraumatic degenerative cervical spinal disorders, the AP view of the cervical spine radiography provides relatively less information than the lateral view. However, the radiological reports and surgeons' feedback suggest that the AP view is sometimes useful in actual practice. As such, the value of the AP view should be patient-specific and dependent upon whether the AP view is essential for the care of the specific patient. A granular practice guideline regarding when to and when not to order the AP view is needed, and warrants future research. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.