Evaluation of assessment scales of hypernasality: The Temple St. Scale of Nasality (TSS) vs. the Visual Analogue Scale (VAS) in terms of inter-rater reliability in the trained and untrained listener

Background: Hypernasality is a distinctive speech problem often associated with cleft palate and velopharyngeal dysfunction (VPD; Sweeney 2011). Assessment of hypernasal speech is essential; speech has long been considered as one of the primary outcome measures of palatal surgery (Grunwell et al 1993; McWilliams et al 1990). The TSS provides a standardised speech sample and five point descriptive ordinal scale suitable for rating hypernasality. Clinicians wishing to utilize this tool may avail of a training course. It has been suggested that the standardised speech sample provided by the TSS might be utilized in conjunction with a visual analogue scale (VAS), and that the role in training in the use of both scales might be explored. Objectives: Aim 1: To compare the inter-rater reliability of the descriptive ordinal scale provided within the TSS with that of a VAS. Aim 2: To compare inter-rater reliability for both scales for trained and untrained listeners. Methods: Trained and untrained listeners rated the speech of 10 individuals with cleft palate and VPD, once according to the TSS and again 4 weeks later via a VAS. Results: Intra-class correlation coefficients revealed very good levels of inter-rater reliability for both tools, with slightly higher levels of inter-rater reliability demonstrated by the VAS (Altman 1991). Trained listeners demonstrated higher levels of inter-rater reliability on the TSS than on the VAS, where they had been trained on the TSS and not the VAS. Untrained listeners displayed higher levels of inter-rater reliability than trained listeners on both scales. Untrained listeners were more reliable utilizing the VAS than when using the TSS. Conclusions: Both tools can be used to rate hypernasality with confidence in terms of inter-rater reliability. The results of this pilot study give support to Cheng’s (2006) unpublished study demonstrating the success of a VAS in rating hypernasality. The role of training in enhancing the reliability of the rating of hypernasal speech remains unclear and requires further examination.