Limited Value of Plain Radiographs in Infant Torticollis

Abstract

OBJECTIVE. The purpose of this work was to assess the frequency of clinically relevant findings from plain films of infants evaluated for torticollis. PATIENTS AND METHODS. After institutional review board approval, radiology records were searched for infants 0 to 12 months of age who underwent plain film study for torticollis or “head tilt.” Infants evaluated for trauma or Down syndrome were excluded. All of the studies were reviewed, demographic data was recorded, and any additional imaging studies were examined. RESULTS. A total of 502 patients (189 girls and 313 boys) were identified with an average age of 0.37 ± 0.2 years. Head tilt was to the left in two thirds of patients. Ten patients had abnormal findings reported. Six of these proved normal on subsequent studies (3 suspected occipital-C1 fusions, 2 suspected cervical fusions, and 1 suspected hemivertebra). Four patients had true bony vertebral abnormalities including absent left C7 pedicle, multiple fusion anomalies from C4 to T2, C3 hemivertebra and thoracic spine anomalies, and C4 hypoplasia. This last patient had abnormal kyphosis on physical examination and demonstrated instability with dynamic testing. Twenty-five additional patients with normal plain films underwent spine computed tomography or magnetic resonance imaging; all were normal. CONCLUSIONS. The true-positive yield of plain films in nontraumatic infant torticollis was low (4 of 502). There were more false-positive than true-positive results. A common rationale for imaging is to exclude craniocervical or other unstable abnormalities that might contraindicate physical therapy, seen in only 1 of the 502 cases. Close physical examination could safely eliminate most patients sent for radiography.