Long-Term Speech Outcomes in Moderate-to-Severe Childhood Speech Sound Disorder: A Systematic Review.

The evidence base is all observational. Included studies were cohort studies, case series and case-control studies (NHMRC level III-2, III-3 and IV); no randomised trials were available, so the review cannot show that treatment changes long-term outcome. Studies defined severity in incompatible ways: one called a child severe below the 10th percentile on the Goldman-Fristoe Test of Articulation and a percentage of consonants correct under 90%, another called a child severe at a percentage of consonants correct under 50%. 'Persistent' therefore means different things across the review. Follow-up timing was variable. Most studies compared preschool with early primary school, roughly two years apart with wide age ranges, and only one cohort study stated an exact interval between assessments, so whether difficulties resolve later than school age is largely unknown. Twenty-one papers came from only 15 unique studies because of overlapping samples from the same children, so some findings are reported more than once and the effective evidence base is smaller than the paper count suggests. Almost every child had received or was receiving speech therapy, and the studies did not describe the type, duration or frequency of that therapy, so the contribution of therapy to any change in speech cannot be separated from development. No cohort of children with dysarthria or a fluency disorder met inclusion, so the review says nothing about those groups. Quality appraisal flagged small sample sizes, non-standardised speech tasks at baseline, and confounding factors such as prior speech therapy.

The work was funded by the National Health and Medical Research Council (grants 1116976 and 1160893). The authors declare no conflicts of interest. No commercial sponsor is reported.

Do not read this review as evidence that speech therapy works or does not work. All the included studies were observational, and almost every child was receiving or had received speech therapy that the studies did not describe in type, duration or frequency, so therapy's contribution to any change in speech cannot be separated from development. The persistence figures also cannot be used as a prognosis for an individual child, because each study set its own threshold for what counted as severe and as resolved.