A Comparison of Sinus Computed Tomography (CT) Staging Systems for Outcomes Research

Abstract

The success of outcomes research for sinus disease will depend in part on the utilization of a computed tomography (CT) staging system that meets rigorous statistical criteria. Such a staging system must demonstrate an even distribution of population by scale, a high inter-rater and intra-rater agreement, and a low percentage of unclassifiable scans. Four proposed sinus CT classification systems were evaluated for these parameters. Each system was used by four masked raters to rate 42 consecutive preoperative CT scans on two separate occasions. One of these systems is based on the site of disease rather than the degree of sinus opacification and gives a clear definition of sinus disease (2 mm or greater of mucosal thickening). This system had the highest inter-rater agreement (κ = 0.61, range = 0.38 to 0.61) and the lowest percentage of unclassifiable scans (0.6%, range = 0.6 to 14.2%). The statistical attributes of this particular sinus CT staging system support its use for outcomes research in sinus disease.