Objective: To determine whether a single Waters view (occipito-mental) radiograph could be substituted for a four-view sinus series to diagnose sinusitis, and to determine the inter- and intraobserver variabilities for sinus radiography.
Design: Radiographs were interpreted by radiologists blinded to the clinical history, and results were recorded on a standardized form.
Setting: Veterans Affairs Medical Center.
Participants: Staff attending radiologists, an attending radiologist with special training in skull radiology, and a senior radiology resident.
Measurements and main results: The agreement between the Waters view and the four-view sinus series was moderate to substantial (simple agreement = 75-84%, kappa = 0.5-0.68). However, agreement varied by sinus and, after correction for chance agreement, was substantial only for the maxillary sinuses (kappa = 0.72-0.87). Intraobserver agreement (kappa = 0.72-0.84) was superior to interobserver agreement (kappa = 0.49-0.59) for the four-view sinus series.
Conclusions: Substituting a single Waters view for a four-view sinus series may be an acceptable strategy for diagnosing maxillary sinusitis.