Evaluation of the "disfluency profile" as a predictor of persistent stuttering in children who have just started to stutter

According to Yairi and Ambrose (2005), for 95% of people who begin to stutter the onset occurs before the age of 4 years, most of the beginnings occur in the third year of life, and 90% of children who start to stutter recover spontaneously, usually no later than the fourth year post-onset. If symptoms persist beyond this term, treatment becomes difficult, so it is important to submit the subjects that tend to persist to early treatment.
The project «Phonetic indexes predictive of chronic stuttering in preschool children» (Dr. Claudio Zmarich, CNR-RSTL n. 995) started in 2008, and aimed to identify some behavioral and clinical indexes that predict persistency. 40 families where at least one component stuttered at that time, for as many children, participated. All children were first audio and video recorded when they were 24-month-old, and as soon as a child showed the first symptoms of stuttering, he was addressed to the «Centro Medico di Foniatria» in Padua for receiving a diagnosis and for being evaluated for speech and language abilities and for ruling out other diseases. Subsequently, the child began to be recorded at home every 3 months up to 16-22 months, to collect data on phonetic development, speech-associated
attitude and severity of stuttering. Here we present data about 12 children which started to stutter. The experimental design is to predict the persistency based on the value of the «disfluency profile» at 9-15 months post-onset (counts of repetitions of part of words or monosyllabic words, as well as disrhythmic phonations, Yairi
and Ambrose, 2005). At the end of the observation period this measure was compared with the SSI-3 score (Riley, 1994). To better determine the final clinical outcome of each child, a reference was also made to a telephone structured interview to parents after an average of 43.2 months from the onset of stuttering, which determined that 3 children had become persistent stuttererers (S) and 9 had recovered spontaneously (NS).
In order to be considered a good clinical marker, the disfluency profile must simultaneously exhibit a level of sensitivity (accuracy in detecting S) and specificity (accuracy in detecting NS) of at least 80%. Comparing the stage of 9-15 months post-onset to the final outcome, you get 2S and 5NS, including 4 false positive and 1 false negative (sensitivity 66.3%, specificity 55.5%). A half year later, the predictive value increases: you get3S and 6NS, with 3 false positives and 0 false negatives (sensitivity 100%, specificity 66.6%). For the stage of 16-22 months we also have the SSI-3 scores: applying the criterion of cut-off used by Howell and Davis (2011), and comparing the results to final outcome, you get 3S and 5NS, including 4 false positives and 0 false negatives (sensitivity 100%, specificity 55.5%).
In conclusion, the predictive value of the disfluency profile at 9-15 months post-onset is lower than the clinical standards, but increases in the following semester, and it is a better and more «user-friendly» predictor than the SSI-3.

Publication type: 
Contributo in atti di convegno
Author or Creator: 
Zmarich Claudio
Bernardini Simona
Colantonio Daniel
Lenoci Giovanni
Natarelli Giulia
Pisciotta Caterina
2nd International Conference On Stuttering, pp. 207–209, Roma (I), 15-17/10/2015
Resource Identifier: 
ISTC Author: 
Claudio Zmarich's picture
Real name: