Red flags for speech impairment: who should we refer to speech therapy?

Speech disorders are a common presenting concern to paediatricians. Yet evidence to guide detection and speech sound therapy referral of cases at risk for persistent speech disorder is limited, with no normative English speech data published for over two decades. Here we describe speech development in a large, contemporaneous normative sample, to help clinicians identify errors of disordered speech from developmental errors (which are of lesser concern). We also examine differences in speech development over 20 years.

Children aged 2 years 0 months-12 years 11 months were recruited from 24 sites (9 schools, 15 childcare/kindergarten settings) from New South Wales and Victoria in Australia. Speech sound acquisition and accuracy were documented, including whole word accuracy and type of sound error patterns categorised (developmental vs disordered).

1179 participants (53% male) were assessed. Speech performance was highly variable up to 6 years of age. By 7 years, 90% of participants could produce all sounds. There were negligible differences in word accuracy and error pattern performance between 8 and 12 years. Common disordered errors included transpositions (eg,'efelant' for elephant), vowel errors and backing (eg,'glack' for black). Compared with historical norms, sounds were acquired at a slower rate, and resolution of some common developmental errors was also slower.

We found a high prevalence and variability of errors up to 6 years which explains why so many families seek support for speech development in the preschool years. Furthermore, our data suggest speech may be mastered more slowly nowadays compared with two decades ago. We provide an assessment tool with age-referenced normative cut-points and identify 'red flag' speech errors to guide data-driven referral of children at risk for persistent disorder to speech and language therapy.