The Rhinocleft®: A Comprehensive Perceptual Assessment of Velopharyngeal Function for Children

Abstract

Introduction The perceptual speech assessment conducted by a Speech Pathologist (SP) is the gold standard for the clinical evaluation of velopharyngeal function. This research aimed to develop an Australian English speech sample and perceptual speech assessment tool, collectively known as the Rhinocleft® Perceptual Cleft Palate Speech Assessment, or “Rhinocleft®”, and to evaluate the tool’s reliability and validity. Methods The Rhinocleft® was developed based upon speech sampling recommendations from an international working group and guidelines from an international speech assessment competencies report. The content and face validity of the tool were established using SPs with experience in the perceptual evaluation of velopharyngeal function. The construct, criterion and diagnostic validity, as well as the reliability of the Rhinocleft® were determined using mean visual-analogue scale ratings from a second expert group of SPs who listened to audio recordings of speech elicited from 40 children aged between 4 and 14 years. Each recording included a spontaneous speech sample as well as short sentence stimuli from both the Rhinocleft® and the Great Ormond Street Speech Assessment (GOS.SP.ASS). Results Ratings using a visual-analogue scale (VAS) demonstrated higher reliability estimates than ordinal scales. The range of estimates for inter-rater (ICC = .93 - 96), intra-rater (ICC = .83 - .97) and test-retest (rs (39) = .92 - .95) reliability was good to excellent for global and specific measures of speech performance. The expert SPs’ content and face validity indices were 90.6% and 99.2%, respectively. Spearman’s rho statistic ranges for VAS ratings established convergent, (rs (39) = -.72 - .98), divergent (rs (39) = -.25 - .06), and predictive, (rs (39) = -.49 - .87), validity, but concurrent validity was only established for global measures of speech performance (rs (39) range = .88 - .91), not specific measures (rs (39) range = -.15 - .07). The sensitivity (95.65%) and specificity (100%) of the Rhinocleft® was also determined. There were no clinically meaningful differences in the reliability and validity results when comparing the Rhinocleft® short sentence speech sample and the GOS.SP.ASS speech stimuli. Conclusion The high reliability and validity estimates have shown that the Rhinocleft® is a valid and reliable tool for the perceptual assessment of velopharyngeal function in Australian English-speaking children.