Focus on early intervention: Stuttering test could be used to screen all schoolchildren

A new model developed by Wellcome Trust researchers to predict the persistence of stuttering could be used to screen all children at school age, new research in the Journal of Fluency Disorders suggests.

Stuttering - also known as stammering - is thought to affect one in 20 children under the age of five, and onset generally occurs around the age of three. Around one child in 100 is likely to still be stuttering in their teenage years.

Screening for communication problems at key stages, including school entry, was suggested by the government in its Every Child Matters initiative, which was launched ten years ago.

Professor Peter Howell and colleagues from the Division of Psychology and Language Sciences at UCL (University College London) previously developed a model that allowed them to accurately predict which children at the age of eight would continue to stutter in their teenage years. In this new study, Professor Howell used the model to show that it is possible to screen children for stuttering when they enter primary school, around the age of five.

Professor Howell says: "For a screening tool to be used it effectively, it needs to meet the rigorous standards for accurately identifying children who stutter separately from children who are fluent. We found that this method can do just that.

"If we can identify children at risk of stuttering, then we can offer appropriate interventions to help them early on. Primary school is a key time in a child's development and any help in tackling potential communication problems could make a big difference to the child's life."

Professor Howell developed the model working with 222 children who stutter and 103 fluent children, and he and his colleagues validated their findings in 272 children who stutter and 25 fluent children, all between ages five and 18. He used the Stuttering Severity Instrument Version 3 (SSI-3, a standardised test involving measurement of speech symptoms, their durations and physical symptoms accompanying stuttering such as tics), which they had previously been shown to be the only reliable tool for predicting persistence of stuttering.

He found that their model performs extremely well with five-year-olds. At this age, over 96 per cent of children who stutter and over 83 per cent of fluent children would be correctly identified. Although this would mean that some children who do not stutter would be incorrectly classified, Professor Howell and colleagues believe that some of these fluent children may suffer from other communication problems.