Objective: Due to the multidimensional nature of resonance disorders, multivariate diagnostic assessment is advisable. The nasality severity index (NSI) is based on this point of view. Because of the influence of personal and environmental variables on the current NSI, this study aims to refine this index.
Design: Prospective case-control study.
Setting: Tertiary university hospital.
Patients: Forty-two patients with cleft lip and palate and 50 children without resonance disorders were tested.
Interventions: Resonance was investigated by perceptual as well as objective measurements. A Nasometer was used to score nasalance, and spectral speech characteristics of a sustained sound /i:/ were determined, among which the voice low tone to high tone ratio (VLHR). Binary logistic regression analysis was performed to calculate the optimal index to discriminate patients from control children. Additionally, the validity of the index was determined based on data from an independent patient and control group.
Results: The NSI 2.0, a weighted linear combination of three variables, can be obtained using the equation NSI 2.0 = 13.20 − (.0824 × nasalance /u:/ [%]) − (.260 × nasalance oral text [%]) − (.242 × VLHR 4.47*F0 [dB]). The NSI has a sensitivity of 92% and a specificity of 100%. Moreover, it has excellent validity (sensitivity 88%, specificity 89%).
Conclusions: The NSI 2.0 discriminates patients from control children with high sensitivity, specificity, and validity. This multiparametric method can offer a more powerful approach in the assessment and treatment planning of individuals with hypernasality.