Clinical Sciences/Health Conditions
Myung Chul Yoo, MD
Myung Chul Yoo, M.D., Ph.D. Assistant professor
Department of Physical Medicine and Rehabilitation Kyung Hee University Medical Center Kyung Hee University School of Medicine
Dongdaemun-gu, Seoul-t'ukpyolsi, Republic of Korea
Jeong Wan Cho, PhD
Jeong Wan Cho, PT, PhD
Department of Physical Medicine & Rehabilitation, Kyung Hee University Medical Center, Seoul Republic of Korea
Dongdaemun-gu, Seoul-t'ukpyolsi, Republic of Korea
Reliable and consistent grading of facial palsy is crucial for clinical decisions and multicenter research. However, the conventional Sunnybrook (SB) scale shows variable interrater agreement across expertise levels and severity grades.To determine whether a modified Sunnybrook (mSB) scale with standardized visual criteria enhances rater agreement compared with the conventional SB scale, and to assess its performance by rater expertise and severity strata.
Design:
Prospective reliability study at an academic medical center. Facial photographs and short videos from 100 patients were assessed by eight raters (four novices; four experts > 5 years) across four sessions—SB(1), SB(2), mSB(1), and mSB(2). Raters were blinded and case order randomized each session. Facial palsy severity was graded using the conventional SB and mSB scales with standardized visual criteria for five voluntary movements (forehead wrinkle, gentle eye closure, open-mouth smile, snarl, lip pucker). Primary outcome was interrater reliability (ICC[A,1]); secondary outcomes included intrarater reliability, expert–novice ICC gap (ΔICC), subgroup ICCs by severity and House–Brackmann grade, and item-level agreement (weighted kappa).
Results:
For composite scores, experts showed higher interrater agreement with mSB (ICC = 0.910; 95% CI 0.892–0.926) than with SB (0.785; 0.694–0.858). Novices also improved (0.820; 0.768–0.859 vs 0.732; 0.629–0.819). Intrarater reliability for novices increased for both voluntary movement (0.83→0.92) and composite scores (0.84→0.92). The expert–novice ICC gap narrowed under mSB (ΔICC = 0.090; 95% CI 0.061–0.124) compared with SB. Item-level analysis showed the greatest gain in gentle eye closure (ΔICC +0.26) and highest reliability for open-mouth smile.
Conclusion:
The mSB enhanced interrater and intrarater reliability for both experts and novices, reducing expertise-dependent variability. These findings support mSB as a reliable and standardized tool for clinical practice and multicenter research in facial palsy assessment.
