Sunnybrook and House-Brackmann Facial Grading Systems: intrarater repeatability and interrater agreement.

Abstract

To assess repeatability and agreement of the House-Brackmann (H-B FGS) and the Sunnybrook (SFGS) Facial Grading Systems. Eight video-recorded facial palsy patients were graded in 2 sittings by 28 doctors. Repeatability and agreement for SFGS were measured by intraclass correlation coefficient (ICC) and coefficient of repeatability (CR), and for H-B FGS by agreement percentage and kappa coefficients. Repeatability for SFGS was from good to excellent and for H-B FGS from fair to good. Agreement between doctors for SFGS was moderate by CR and nearly perfect by ICC. For H-B FGS, the agreement percentage was 50%, and generalized kappa indicated only fair agreement. SFGS was at least as good in repeatability as H-B FGS and showed more reliable results in agreement between doctors. While waiting for an ideal scale for facial grading, the usage of SFGS can be encouraged over H-B FGS.