Are plain radiographs sufficient to exclude cervical spine injuries in low-risk adults?