Facial grimace during speech in cleft lip and palate: a proposal for classification.

Purpose: To investigate the effectiveness of a proposal for classification of facial grimace (FG) and its correlation with objective evaluation of velopharyngeal closure (VPC).

Methods: Twenty individuals with repaired cleft lip and palate underwent velopharyngeal area measurement by means of rhinomanometry and speech sample recording. The FG was rated in two steps, by three speech-language pathologists. First the evaluators rated the FG using their own criteria as: 1= absent FG; 2=mild; 3=moderate; 4=severe. Subsequently, they were submitted to a training session that established the following FG rating criteria: 1=absent FG; 2=movement only of the nose or upper third of the face; 3=strong movement of the nose or upper third of the face; 4=movement of the nose and upper third of the face. The evaluators rated the FG using the established criteria. Intra- and inter-rater agreement were calculated using weighted Kappa coefficient. Correlation between the two stage ratings with the VPC was calculated by Spearman's correlation coefficient.

Results: In the first stage inter-rater agreement ranged from fair to substantial; in the second stage, from substantial to almost perfect. Intra-rater agreement ranged from moderate to almost perfect in the first stage, and from moderate to substantial in the second stage. The correlation between FG and velopharyngeal area was positive and significant in both stages.

Conclusion: The proposed FG judgement proved to be effective in determining the symptom and reliable in diagnosing the severity of velopharyngeal dysfunction. The significant correlation between perceptual and instrumental methods suggests that FG can be used in predicting VPC.