Sunnybrook facial grading system: reliability and criteria for grading

Area of Science:

• Clinical research methodology
• Facial movement analysis
• Quantitative outcome assessment

Background:

• Clinical research demands precision, particularly in multicenter trials.
• The Sunnybrook Facial Grading System (SB) is widely used but has ambiguities.
• Improving the reliability of facial grading systems is crucial for objective assessment.

Purpose of the Study:

• To develop specific grading criteria for the Sunnybrook Facial Grading System (SB).
• To evaluate the impact of these criteria on intra-rater and inter-rater reliability.
• To test the hypothesis that specific criteria enhance reliability, even for inexperienced observers.

Main Methods:

• Prospective study design to test the hypothesis.
• Facial video recordings of 30 subjects with facial paralysis were assessed.
• Two independent raters evaluated videos using the SB system, with and without specific grading criteria.

Main Results:

• The Sunnybrook Facial Grading System (SB) demonstrated good initial reliability (ICC=0.890).
• Specific grading criteria significantly improved inter-rater reliability (ICC=0.927).
• Score variability was highest in the midrange, particularly for brow rising and lip puckering movements.

Conclusions:

• This study introduces the first reported specific criteria for the Sunnybrook Facial Grading System (SB).
• The findings provide an in-depth analysis of where score variances typically occur within the SB system.
• Implementing specific criteria enhances the precision and reliability of facial movement assessments.