Reliability of the sunnybrook facial grading system by novice users

Objectives: To evaluate the intrarater and inter-rater reliability of the Sunnybrook Facial Grading System (SFGS) by novice users. Design: Prospective interval study using one measurement instrument. Setting: Tertiary teaching hospital (Sunnybrook & Women's College Health Science Centre). Methods: Twenty-two patients with a wide spectrum of facial dysfunction recorded on videotapes were rated using the SFGS by eight novice observers independently in two different sittings separated by 3 weeks. The order of patients was randomized for the second sitting. Intraclass correlation coefficients were calculated for component scores and for total scores within and between raters. Results: The intrarater reliability coefficients for the eight raters ranged from .838 to .929. This largely overlaps with the data obtained in previous studies with expert raters. The inter-rater reliability for all eight raters at time I was .982 and for time 2 was .970. This is higher than what was previously obtained with expert raters. Conclusion: The SFGS is as reliable when applied by novice users as by expert users.