Reliability results of perceptual ratings of resonance, nasal airflow and speech acceptability in patients with cleft palate by Ugandan speech-language pathologists following a two-day workshop

Introduction: To meet the scarcity of training and/or educational initiatives in Uganda regarding state-of-the-art speech diagnosis and treatment in patients with cleft palate, a workshop was organized for all interested speech-language pathologists and health care workers. Objective: To evaluate and compare the inter- and intra-rater reliability of the perceptual evaluation of hypernasality, hyponasality, nasal emission, nasal turbulence and speech acceptability before and after a two-day workshop in Ugandan speech-language pathologists. Methods: On the first day, perceptual speech evaluation was discussed and practiced. Perceptual exercises included individual ratings of a specific speech variable followed by a group discussion and consensus listening exercises in listener pairs. The second day focused on speech treatment. Ten Ugandan speech-language pathologists rated speech samples of Ugandan patients with a CP +/- L before and immediately following the two-day workshop. Inter- and intra-rater reliability of the perceptual ratings of hypernasality, hyponasality, nasal emission, nasal turbulence and speech acceptability were determined by means of the absolute percentage of agreement. The Wilcoxon signed-rank test was used to compare results at both time points. Results: Overall inter- and intra-rater reliability improved when observing the absolute percentage agreement. However, median agreement results only showed enhanced reliability for hypernasality, hyponasality and nasal turbulence whereas (limited) deteriorated reliability was observed for nasal emission and speech acceptability. Regarding inter-rater reliability only the median percentage agreement for hyponasality exceeded 50% (median: 56.3%). Overall, better results were found for intra-rater reliability, with only weak results for speech acceptability. Discussion and conclusion: Training positively affected reliability results. Nonetheless, this improvement was not achieved for each speech variable and reliability improvements were modest. Findings indicate the need for continued training and the search for the most effective training paradigm and feedback techniques, whilst taking into account clinical relevance and practical considerations