Detection of incidental vertebral fractures on routine lateral chest radiographs

Purpose. This study sought to assess the usefulness of routine lateral chest radiographs for detecting unrecognised vertebral compression fractures. Materials and methods. We prospectively selected outpatients without symptoms, or risk factors for osteoporosis who underwent chest radiography for different clinical indications. Two independent reviewers with different levels of experience assessed the radiographs for vertebral deformities and graded them as mild, moderate and severe according to the semiquantitative Genant Index. The kappa statistic was used to evaluate interobserver agreement and verify the reproducibility of this method. The prevalence of vertebral fractures observed was compared with that recorded in the official radiology reports. Results. Our study involved 145 patients (73 men. 72 women: age range 50-86 years, mean age 67.5). Clinically relevant vertebral fractures were seen in 18/145 patients (12.4%). These were moderate in 13 patients and severe in five, and single in 12 patients and multiple in six. Interobserver agreement was very high (kappa=0.9). Only 11% of these fractures were recorded in the official reports. Conclusions. Lateral chest radiographs could be effective for assessing previously unknown vertebral compression fractures in individuals without clinical evidence or risk factors for osteoporosis.