Marino, Michael J.; Weinstein, Jacqueline E.; Riley, Charles A.; Levy, Joshua M.; Emerson, Noah A.; McCoul, Edward D.
BackgroundThe purpose of this study was to develop and validate a radiographic metric for characterizing the degree of paranasal sinus pneumatization. A validated metric for the extent of sinus pneumatization that comprehensively includes the maxillary, ethmoid, frontal, and sphenoid cavities is not currently available.; MethodsA validation study was performed in which 5 independent reviewers evaluated 49 sinus computed tomography (CT) scans in coronal, sagittal, and axial orientations. Reviewers evaluated each scan, bilaterally, for 18 proposed dichotomous items as part of the Assessment of Pneumatization of the Paranasal Sinuses (APPS) metric. Evaluation of APPS items was independent of sinus opacification, which was simultaneously evaluated using the standard and validated Lund-Mackay scoring system. Interrater and intrarater reliability was assessed for each proposed APPS parameter and Lund-Mackay item using Fleiss kappa statistic.; ResultsNine parameters were included in the final APPS metric due to substantial interrater reliability ((mean) = 0.61, (range) = 0.41-0.81) and intrarater consistency ((mean) = 0.64, (range) = 0.53-0.77), variable radiographic presence, and unique contribution to the characterization of sinus pneumatization. Kappa values were also calculated for Lund-Mackay interrater reliability ((mean) = 0.58, (range) = 0.45-0.66) and intrarater consistency ((mean) = 0.71, (range) = 0.65-0.75). The final APPS metric has comparable interrater and intrarater reliability to Lund-Mackay scoring. APPS scores were normally distributed within the study group by Anderson-Darling normality test (p < 0.005).; ConclusionThe APPS score is the first comprehensive and validated metric for quantifying the degree of paranasal sinus pneumatization and anatomic variation. This has important potential utility in standardizing evaluation of sinus CT and researching the relationship of sinus pneumatization with clinical parameters.